Home

RK06/RK07 Disk Drive User`s Manual

1. 3 1 Heat RENS mh a 3 1 ii 1 tA 1 1 2 WwW WW G9 C9 3 2 1 3 2 2 3 2 3 3 2 4 3 2 5 3 2 6 4 CONTENTS CONT Page ACOUSUCS M 3 1 asa 3 1 Radiated Emissions u E SH DE OD o eens 3 2 Vibration Mechanical 5 3 2 3 2 SITE PREPARATION 3 5 qaya 3 5 Gun dT 3 5 Power ecce et HA IE RS Ue ae eredi scat d 3 5 POOR LOR GI o od 3 5 Installation Constraints cusa vi poet oU sd Pl vade vie II BEL C Cono 3 5 Fire and Safety Precautions CI PH ERE DER MEA 3 5 INSTALLATION UNPACKING ANDINSPECTION c dula 4 1 SAFETY PRECAUTIONS far cor MAN eO enean tI EA 4 6 INSTALLATION PROCEDURE 4 6 RK611 Controller una HR E 4 6 oen 4 10 INSTALLATION CHECKS AND 4 15 Controller Power Supply Voltage Checks
2. 4 18 RK07 Head Assembly 4 19 RK611 RK06 or RK611 RK07 Dual Access Cable Diagram 4 27 RK611 RK06 or RK611 RK07 Cable Diagram for Testing Dual Access 4 28 RK06 or RK07 Disk Drive Power and Control Switches 6 2 Typical Disk Drive with Eid Released re 6 5 Typical Disk CartrIdg a a atascas esas tisse enn vas 6 6 Engagihg Lock a S awasqa 6 7 Releasing Protection CG Over a E 6 8 Positioning for Loads u aa cus at 6 9 Cartridge Locked in Shroud o ar Ge et tee dut 6 10 Installing Protection 5 cui t voe hice 6 11 Engaging Lock Release ep ORE dU vice e eU UE 6 12 Device Registers ooo eher ond doce es ae QU 7 3 Control Status Register Lulu A a ES 7 7 Word Count Resistet a 7 7 Bus Address 7 7 DiskAddress RegISt er o 7 7 Control Status
3. Veo dpi hs 6 1 RUN STOP Switch with STOP Indicator eee 6 1 UNIT SELECT Switch with READY Indicator 6 3 FAUEFPIndicdtOP uso eco ande det 6 3 WRITE PROT Switch with PROTECT Indicator 6 3 ACCESS A ENABLE Switch with SELECT Indicator 6 3 Single Access Drive A x vestris lave oe ea ete E ob vu Ves 6 3 Dual Access Drive and B a 6 4 ACCESS B ENABLE Switch with SELECT Indicator 6 4 Single Access Drive B uico Le 6 4 Dual Access Drive and 6 4 OPERA TING PROCEDURES isi nee Ferr once CO GE EE SPERA e ERN URS 6 4 RKO6K and RKO7K Cartridge Loading eene 6 4 RKO06K and RKO7 Cartridge Unloading eene 6 12 7 7 1 7 2 7 2 1 7 2 2 7 2 3 7 2 4 7 2 5 7 2 6 7 2 7 2 8 7 2 9 7 2 10 CONTENTS CONT Page PROGRAMMER S GUIDE INTRODUCTION aeq deua 7 1 DEVICE REGISTERS S 7 2 Control Status Register RKCS1 7 2 Word Count Register RKWC 7 6 Bus Address Register RK BA 7 6 Disk Addre
4. CERR an error burst has been detected that exceeds 11 bits in length negating the value of the information contained in the ECC Position and Pattern registers Under these conditions the program must initiate a data recovery routine to provide a sequence of successive rereads Figure 7 32 ASSUME READ 1 MULTISECTOR READ COMMAND SECTORS 1 TO 10 2 SECTOR 5 HAS A READ ERROR READ NO ERROR TRANSFER NO DONE ECC PROCESS TO SECTOR 5 YES ERROR CORRECTABLE YES RELOAD DESIRED SECTOR ADDRESS TO SECTOR 1 RE INITIATE GO BIT RE INITIATE A GO BIT DESIRED SECTOR ADDRESS REG ERROR NOT ECC EQUAL 6 RECOVERABLE NO LOSE A REVOLUTION ONLY APPLICABLE IF ERROR IS UNCORRECTABLE AFTER FIRST PASS 11 4231 Figure 7 32 Data Error Flow 7 31 Any of 16 rereads at track centerline could result in recovery of data with or without ECC correction cycle Typically this could occur if an error producing material e g dirt specks either disappeared from that area of the disk surface or diminished to a point that allowed the data to become ECC correctable If error data is recovered or corrected during any one of 12 offset 400 800 1200 for RK06 or 200 400 600 for RKO7 attempts two rereads per offset position it may indicate that the data on the entire track has been generated by a misaligned head Under these condi
5. 7 25 Start Spindle Command a repris tUe e VY ve x RE DEVE 7 25 Kecalibrate cO Rt E e UM cis 7 26 Offset Command asna m etr rbv Eat da iq REN INE 7 26 Offset Command d or E RI ae aie uie te 7 27 Seek Tr 7 27 Data Handling Commands daar depen den 7 28 Maintenance Clock MCLE Timing cos erro pre Er E FORE PE icine 7 31 Error Correction Process de osse upto Ra E 7 35 Example of 8 Bit Error Correction Process 2 4 0 7 36 Data Error FOW T 7 37 a ER 8 2 RK611 RK06 or RK611 RK07 Dual Access System Configuration 9 RKO06 or RK07 Disk Drive Power and Control Switches 9 2 Drive Status Bits Contained in and B 22 022 9 7 TABLES Title Page RK611 Controller Performance Specifications 1 5 06 Disk Drive Performance Specifications 0000 1 6 RK 07 Disk Drive Performance Specifications sss 1 8 RK611 RK06 Disk Subsystem 21 1 1 1 11 RK611 RKO07 Disk Subsystem Options ess
6. ERROR REGISTER RKER 15 14 13 12 10 09 ATTENTION SUMMARY AND OFFSET 777456 yess 07 DESIRED CYLINDER REGISTER UNIBUS R W ADDRESS 777460 UNUSED 777462 DATA BUFFER R W DB DB DB DB DB DB DB DB DB 15 8 7 0 1 RD WRT a p ECC POSITION REGISTER RKECPS 08 07 05 04 03 02 01 00 EKEEPT ECC PATTERN REGISTER 777472 10 09 08 07 06 05 04 13 12 08 07 06 EPT EPT 03 01 MAINTENANCE REGISTER 2 RKMR2 RO 777474 MAINTENANCE REGISTER 3 R 777476 1 4199 Figure 7 1 RK611 Device Registers 7 3 Interrupt Enable Bit 6 When Interrupt Enable IE is set the controller will be allowed to interrupt the processor under any of the following conditions When Controller Ready R DY bit 7 in RKCSI is set upon completion of a command When any drive sets an associated Attention ATNO ATN7 bit 8 15 in RKAS OF with the set condition of Controller Ready RDY When any drive or the controller indicates the presence of an error by the setting of Controller Error bit 15 in RKCSI In addition via program control an interrupt can be forced by the simultaneous setting of the Inter rupt Enable IE and Controller Ready
7. 82 RK06 EA RK06 EB 06 RK06 ED RK06 FA RK06 FB RK06 FC RK06 FD RK611 EA RK611 EB RK611 EC RK611 ED RK611 FA RK611 FB RK611 FC RK611 FD RKO6C RK611 C RK06K DC RK06K AC RKO6K EF 70 12292 08 70 12292 15 70 12292 25 70 12292 40 Table 1 4 RK611 RK06 Disk Subsystem Options Description Single access RK06 H969 cabinet 115 Vac 60 Hz Single access RK06 H969 cabinet 230 Vac 60 Hz Single access RK06 H969 cabinet 115 Vac 50 Hz Single access RK06 H969 cabinet 230 50 Hz Dual access RK06 H969 cabinet 115 Vac 60 Hz Dual access RK06 H969 cabinet 230 Vac 60 Hz Dual access RK06 H969 cabinet 115 Vac 50 Hz Dual access RK06 H969 cabinet 230 Vac 50 Hz RKO06 EA Drive with RK611 Controller RKO06 EB Drive with RK611 Controller RK06 EC Drive with RK611 Controller RKO06 ED Drive with RK611 Controller 06 Drive with two RK611 Controllers RK06 FB Drive with two RK611 Controllers RK06 FC Drive with two RK611 Controllers RKO06 FD Drive with two RK611 Controllers Dual access kit Dual access kit plus one Unibus controller Drive data cartridge Drive alignment cartridge Drive error free data cartridge 8 ft cable 15 ft cable 25 ft cable 40 ft cable Option RKO7 EA RKO7 EB 07 RK07 ED 07 07 RK07 FC RK07 FD RK711 EA RK711 EB RK711 EC RK711 ED RK711 FA RK711 FB RK711 FC RK711 FD RKO7C RK611 C RK07K DC RK
8. Bit 9 When DMD is set MERD is used to simulate encoded serial data Maintenance Encoded Write Data MEWD Bit 10 The MEWD bit is a read only bit that is nor mally unconditionally set by the controller A reset condition of the bit when read in conjunction with the setting of DMD indicates that the encoding logic has functioned properly during the simulated execution of a write command Precompensation Advance PCA Bit 11 Precompensation Advance is a read only bit that is set to indicate that write precompensation logic within the controller has determined that the pulse required to record the current data bit must be advanced in time to ensure the reliability of the bit during subsequent reads Precompensation Delay PCD Bit 12 Precompensation Delay is a read only bit that is set to indicate that write precompensation logic within the controller has determined that the pulse required to record the current data bit must be delayed in time to ensure the reliability of the bit during sub sequent reads ECC Word ECCW Bit 13 ECC Word is a read only bit that is normally set in the controller A reset condition of the bit indicates that the timing logic is reflecting the ECC field within the sector Write Gate WRT GT Bit 15 Write Gate is a read only bit that is set to indicate that Write Gate has been asserted by the controller Read Gate RD GT Bit 14 Read Gate is a read only bit that is set to indicate th
9. Directs Drive To Generates Control Signals 1 Select Drive 2 Pack Acknowledge 1 Position R W Heads 1 Polled Attention 3 Drive Clear 2 Select Acknowledge Non Data 4 Unload 2 Read or Record Data 3 Multiple Drive Select Handling 5 Start Spindle 4 Sector Index Pulses 6 Recalibrate 3 Send Fault Error or Status 5 Write Clock 7 Offset Information 6 Strobe 8 Seek Header 9 Read Header Performs Parallel to Serial Write Generates Status Messages Handling 10 Write Header and Serial to Parallel Read Data Eight 16 Bit Words 11 Read Data Conversions for Direct Memory Access Data Data 12 Write Data Transfers Handling 13 Write Check 11 4195 Figure 2 4 Basic RK06 RKO7 Disk Subsystem 2 5 Address octal 777440 777442 777444 777446 777450 777452 777454 777456 777460 777462 777464 777466 777470 777472 777474 777476 Table 2 1 RK611 Unibus Addressable Registers Control Status Register 1 RKCS1 Word Count Register RK WC Bus Address Register RK BA Disk Address Register RKDA Control Status Register RKCS2 Drive Status Register RK DS Error Register RKER Attention Summary Offset Register RKAS OF Desired Cylinder Register RKDC Data Buffer Register RKDB Maintenance Register 1 RKMR1 ECC Position Register RK ECPS ECC Pattern Register RK ECPT Maintenance Regist
10. Drive response problem Drive is not available dual access only Any of 21 drive error conditions Data transfer attempt to from bad sector Header Check indicates header information Table 7 1 Combined Error CERR Cont Error Indicator Bit Condition Operation Incomplete OPI RKER bit 13 Desired header cannot be found DTC Parity Error DTC RKCSI bit 13 Parity error in message from drive PAR Write Check Error WCE RKCS2 bit 14 Write Check indicates data from disk did not match data from memory Nonexistent Memory RKCS2 bit 11 No SSYN within 10 us of MSYN as NEM sertion Data Late Error DLTERR RKCS2 bit 15 Data late to from Silo Drive Timing Error DTE RKER bit 12 Write Clock loss during write data loss during read Data Check DCK RKER bit 15 Data error detected by ECC Multiple Drive Select MDS RKCS2 bit 9 More than one drive simultaneously selected Error Correction Hard ECH RKER bit 6 Data error uncorrectable using ECC Controller Time Out CTO RKCSI bit 11 Command possibly not executed 7 2 2 Word Count Register RKWC The RKWC register Figure 7 3 is loaded with the two s complement of the number of data words to be transferred to or from main memory The register is incremented by 1 after each bus cycle and accommodates a maximum transfer of 65 536 words The RKWC register can only be cleared by writing all zeros via program control 7 2 3 Bus Address Register RKBA The
11. MANUAL SELECT VERIFY ENTER HEAD NO 0 2 2 HEADS POSITIONED AT CYLINDER 365 OCT HEAD 2 SELECTED ENTER HEAD NO 0 2 2 4 22 4 4 4 3 8 Head Alignment Using Field Text Box Remove the rear cover from the RK06 or RKO7 Remove power from the RK06 or RKO7 and from the Field Test Box Connect the head alignment cable from the Field Test Box to the RK06 or RKO7 read write board Attach the appropriate drive interface cable to the Field Test Box Apply power to RK06 or RKO7 and to the Field Test Box Install an RKO6K or RKO7K alignment cartridge depending on the drive and WRITE LOCK the drive NOTE Prior to beginning head alignment procedures power should be applied to the drive for two hours the car tridge should be mounted and spinning for at least one half hour To WRITE LOCK the drive set the SAFETY SWITCH S2 on the RK06 M7729 board or the RK07 M7906 board to the MAINT position This action will also prevent the heads from unloading if a Servo Unsafe condition occurs Mount the head alignment fixture on the appropriate head Figures 4 10 and 4 11 CAUTION Make sure that the carriage does not tip beyond the point where the heads will fall out of the cam tower Torque the head alignment fixture mounting screw to 5 in lb Loosen the head mounting screw and then tighten it to 2 in lb Turn both vertical screws of the head alignment fixture fully CCW Select the desired head and load t
12. Write Check Error WCE Bit 14 Write Check Error is a read only error bit that is set to indicate that a data word read from the disk during the execution of a Write Check command did not compare with the corresponding data word contained in main memory If a write check error is detected and the BAI bit is not set the RKBA register will contain the memory address of the next data word location mismatched word address plus two Data Late Error DLT Bit 15 Data Late Error is a read only error bit that is set to indicate the following During the execution of a Write command the Silo was empty when the disk required a data word During the execution of a Read or a Write Check command the Silo was full when the disk provided the next data word This error bit can be forced set under the following conditions By loading the Data Buffer register RK DB when Input Ready IR bit 6 is reset By reading the Data Buffer register RK DB when Output Ready bit 7 is reset NOTE When RKCS2 is loaded under program control the controller Sector Index pulse separator logic is in itialized This action causes the controller to ignore the next sector or index pulse from the selected drive Therefore unnecessary loading of RKCS2 can in crease latency time by an amount equivalent to 1 5 sectors 7 9 7 2 6 Drive Status Register RKDS The RKDS register Figure 7 7 is a read only register that is used to stor
13. 7 8 Drive Status Register scan 7 10 Drive Error Register ios tec rue ripa oes vet od cu d OUR ER OUR 7 13 Attention Summary Offset Register 7 15 Desired Cylinder Keglstet iin areas oo EDU I ae paio Ne Fas EUR 7 17 Butler Repistei o k dite adu ics Con Rcs ma 7 17 Maintenance Register er o rd son boda eda rax UU ARMS 7 17 ECC Position co ossa rto o Vete tom UE 7 19 ECC Pattern rode arces Mo Edu Paix ce huma veta kuwa 7 19 Maintenance Register 2 Controller to 3 7 20 Maintenance Register 2 Drive to 7 21 Maintenance Register 3 Controller to 41 7 22 Maintenance Register 3 Drive to Controller 7 22 Select Drive Command ooo CH 7 24 Pack Acknowledge Command aao ao og WR OY E YE d RR 7 24 Figure No 7 21 7 22 7 23 7 24 7 25 7 26 7 27 7 28 7 29 7 30 7 31 7 32 8 1 9 1 9 2 9 3 Table No FIGURES CONT Title Page Drive Clear OMAN tant vn RA Ran rau at 7 24 Unload Commando odo eo prr
14. CABLE GROUND 25 CABLE GROUND 3 LG SCALE THIS HARDWARE 1 FOR THE PURPOSE OF FASTENING A GROUND STRAP IN A MULTIPLE DRIVE SYSTEM 969 CABINET POWER CORD CABLE DRIVE BUS 25 FT CABLE GROUND 3 FT LG POWER CORD CABLE DRIVE BUS 25 FT LG THIS GROUND STRAP 15 NOT IF BOTH CONTROLLERS ARE MOUNTED IN THE SAME EXPANDER BOX OR CAB 11 4983 Figure 4 9 Dual Access Disk Subsystem H969 Cabinet 4 14 10 For the last drive connect a drive bus terminator PN 70 12293 0 0 to I O 12 Figure 4 7 NOTE For the dual controller cabling configuration Figure 4 7 the installation of the drive bus cable should fol low the above sequence The drive connectors used are J3 B IN and J4 B OUT A second drive bus terminator PN 70 12293 0 0 must be connected to 74 OUT of the last drive The second controller may be mounted in the same expander box as the first or in another cabinet altogether If the same box is used a second transition connector is mounted in the cabinet to accommodate the second drive bus cable 11 Ensure that circuit breaker is OFF Connect the power cable to a power outlet whose voltage and frequency meet the requirements of the drive model Figure 4 6 12 Insert the proper drive select plug into the unit select jack of the operator control panel Figure 4 6 Table 4 5 lists the part numbers for eac
15. PCD for proper controller operation In this manner the data paths existing between main memory the Silo and the encoder separator logic can be exercised and evaluated In addition since both read and write data originates in main memory and is effectively returned directly or by monitor a software comparison can be made to validate the integrity of the data paths 7 29 Controller Modifications In Diagnostic mode operation of controller is modified by disabling the generation of certain error and control signals while simulating and monitoring the generation of other signals as follows Generation of these error conditions is inhibited 1 Controller Time Out CTO bit 11 in RKCS1 2 Programming Error PGE bit 10 in RKCS2 3 Unit Field Error UFE bit 13 in RKCS2 4 Drive Timing Error DTE bit 12 in RKER Generation of these signals is simulated 1 Index Pulse MIND 2 Sector Pulse MSP 3 Data Control Clock MCLK 4 Read Data MERD 5 Even Parity PAT on Messages and can only be done in real time These signals are monitored 1 Precompensation Advance PCA 2 Precompensation Delay PCD 3 Timing in ECC area of sector ECCW 4 Write Data MEWD 5 Read Gate RD 6 Write Gate WRT Timing Considerations The Maintenance Clock MCLK bit bit 8 in can be alternately set and reset by repeatedly loading 1 to provide program controlled clock Pulses ge
16. 4 16 Disk Drive Power Supply Voltage Checks a 4 16 Head Load Unload Checks a de o re e e YN IE 4 16 Head Alignment Checks and Cartridge Care 4 17 Head Alignment Using Program 4 18 Sample u san e i ELO PREIS S TE DIS 4 21 Head Alignment Using the Field Text 4 23 Response to Abnormal 4 24 Clearing an Unsafe Condition eene 4 24 Removing a Cartridge When Drive Power is 4 4 25 Detecting Head to Disk Interference seen 4 25 POWER CONVERSION 4 25 DUAL ACCESS DRIVE INSTALLATION AND ACCEPTANCE 4 26 Dual Access Installation Procedure eee eene 4 26 Dual Access Acceptance 20 2 0000 4 26 Dual Access Acceptance Procedure 4 26 Dual Access Diagnostics oce tret pde pase ated 4 26 RK611 RK06 and RK07 4 26 RK06 and RKO7 Performance 4 26 5 CA CA CA CA CA CA Q b WKH NN 1
17. X Wo X l4 ae pq pow dL uh b L P P de xe s DRIVE SELECT CODE E SEPT ud Pip STATUS PARITY lt FORMAT GE ul A1 SELECT CODE REV RTZ PRES a A2 2 DRIVE SELECT CODE CYLINDER DIFFERENCE OFFSET VALUE 5 PARITY 5 DRIVE SELECT CODE DRIVE SERIAL NUMBER 0 RESVD FOR INV AC PRTY am cd _ WR WRITE SEEK LIM Bi RESVD FOR SECTOR 8 NO GATE HEAD MULT INDEX TRIBIT SERVO No SERVO Parity ADD L MESS ERR WRITE S NO FAULT HD SEL ERR RESVD FOR CYLINDER ADDRESS ADD L MESS o ul gt RESVD FOR DECODED B ADD L MESS SECTOR COUNT HEAD ADDRESS B DRIVE TO CONTROLLER TRANSMISSIONS NOTE THESE BITS ARE USED ONLY ON THE RKQ7 11 5328 Figure 9 3 Drive Status Bits Contained in and B Messages 9 7 RK06 RK07 DISK DRIVE Reader s Comments USER S MANUAL Your comments and suggestions will help us in our continuous effort to improve the quality and usefulness of our publications What is your general reaction to this manual In your judgment is it complete accurate well organized well written etc Is it easy to use What features most useful What faults or errors have you found in the manual Does this manual satis
18. 2 Remove RK611 from its shipping container Install the RK611 backplane into the designated expander box and secure with the screws provided in the accessories bag Install power harnesses 2 using expander box instructions NOTE For RK07 installations the M7901 must be above ETCH E If not FCO no M7901 003 must be added Step 3 is FCO no M7901 003 If the M7901 is above ETCH E skip step 3 For all RK06 installations skip step 3 If the subsystem contains any RK07 Disk Drives verify that the RK611 Controller has been modified to operate with RK07 Drives as well as with RK06 Drives To make this veri fication ensure that the following changes have been made to the M7901 module On Side 2 the etch at pin E11 11 has been cut On Side 2 the etch at pin E12 5 has been cut On Side 1 a wire has been added connecting E11 10 to E11 11 On Side 1 a wire has been added connecting E12 5 to E17 2 PED 4 6 Table 4 2 RK611 Control Switch Configurations for Desired Unibus Address and Interrupt Vector Switch Section Switch Section Unibus Bit 51 1 51 2 5 1 3 5 1 4 5 1 5 5 1 6 5 1 7 5 1 8 For standard address of 777440 these switches should be in the ON position all others should be in the OFF position Note that this address overlaps the Look Ahead Register in the Disk Subsystem For the standard vector of 210 these switches should be in the OFF position all others sho
19. 5 2 1 1 5 2 1 2 5 2 2 1 5 2 2 2 5 2 3 5 2 4 5 2 4 1 5 2 4 2 6 6 1 6 2 6 2 1 6 2 2 6 2 3 6 2 4 6 2 5 6 2 6 6 2 6 1 6 2 6 2 6 2 7 6 2 7 1 6 2 7 2 6 3 6 3 1 6 3 2 CONTENTS CONT Page FIELD ACCEPTANCE TEST PROCEDURE ERROR DEPINITIONS RAT ES eene oov 5 CA ERAI eie 5 1 Hard POEs a 5 3 lu aaa s He ae 5 3 Cartridge BEEOES y PU cuba S 5 3 Seek NE 5 3 DIAGNOSTIC TESTING e eis oai 5 3 RKO Controleren SEP ES ERN QU OP KIC 5 3 General ese m a tsi ep ue Gee ros tua OVE UA 5 3 Procedure e err 5 7 RK06 Or RK07 Disk Drive 5 8 Generally maa du 5 8 Procedure isa a upa 5 8 Interchangeability T CSU 5 9 Data Reliability 5 9 General a 5 9 Procedur Ganne n 5 9 OPERATOR S GUIDE INTRODUCTION 6 1 OPERATOR CONTROLS INDICATORS 6 1 Power ON OFF Circuit Breaker
20. OPTIONS P uw tema Doe c 1 4 SUBSYSTEM DESCRIPTION INTRODUCTION L eres 2 1 RK06 AND RK07 DISK 2 1 RK06K AND RK07K DISK CARTRIDGES 11 2 3 RK611 DISK DRIVE CONTROLLER eremi enini 2 3 Command Control so uir 2 3 Data Control ea pani atahay wu Resa DIR 2 3 Controller us 2 3 Silo M moryxy una Sama 2 7 RK611 RK06 AND RK611 RK07SUBSYSTEMS 2 8 Controller Drive Interface 2 8 SUDsystem Commalide Mee Pant 2 12 Data PormattiDg pta addat 2 12 Servo Surface oot vo ex d WE D E 2 13 Data Error Control is ai a asa a a ha au uh tov D yl a Tana ERES 2 13 Bad Sector Elle i an au 2 14 Data Field SE QA AA 2 15 Track Foral o ooo res 2 15 SITE CONSIDERATIONS AND PREPARATIONS ENVIRONMENTAL 3 1 General NE 3 1 Temperature i saa k w 3 1 Relative ode
21. When both messages are defined by the equivalent value Message is sent to the RK MR2 register as Message is simultaneously sent to the RKMR3 register The select bits are cleared by conventional initialization or by loading a com mand other than Select Drive into RKCSI Parity Test PAT Bit 4 When the Parity Test bit is set the controller is forced to generate even parity for the messages sent to the drive on the Message A and Message B lines Similarly the con troller makes an even parity check on the status messages returned from the drive Diagnostic Mode DMD Bit 5 When the Diagnostic mode bit is set the controller is effectively disconnected from all of the drives and placed under the control of specific bits in the register 1 MSP MIND MCLK Maintenance Sector Pulse Bit 6 MSP Bit 6 When MSP is set in conjunction with the setting of the bit the receipt of a sector pulse from a drive is simulated Maintenance Index MIND Bit 7 When MIND is set in conjunction with the setting of the DMD bit the receipt of an index pulse from a drive is simulated Maintenance Clock MCLK Bit 8 When DMD is set MCLK replaces the internal controller clock derived from the data separator The toggling set reset etc of MCLK provides the clock pulses required to step through a controller command including the simulated reading or writing of data Maintenance Encoded Read Data
22. drive is seized by the other port At this time an initialize will a Clear the port s attention bit b Cancel the port request if one is pending 9 8 GETTING DRIVE STATUS IN DUAL ACCESS MODE If the drive is seized by one of the ports that port s controller may interrogate any of the eight status multiplexers 0 BO 1 2 2 3 B3 within the drive If the drive cannot be seized busy with another port only status multiplexers AO and BO may be read through the unseized port This situation causes the Combined Error bit bit 15 in RKCS to be asserted in the controller of the unseized port Refer to Figure 9 3 for the data bits contained in drive status messages AO through B3 9 8 1 Status Bits Pertinent to the Dual Port Available at the Unibus Interface 1 Drive available DRA This bit is set to the port that has access to the drive and reset to the other port Read only 2 Volume Valid VV This bit is used to indicate when a disk cartridge may have been changed There is a Volume Valid bit for each port 3 Ready RD Indicates that the selected drive is up to speed and its heads are settled over the specified cylinder This bit will always be reset to a port not having access to the drive 9 6 SSS SSS Se 7 44 ps ee CONTROL CLOCK JUUU uud To n Ts Ta Ts Te T Tg Tg is 115 STROBE
23. drives Detects 1 head misalignment 2 positioner lat eral misalignment 3 spindle cartridge interface runout 4 improper levels of write current and 5 incorrect addressing of read write heads Prints summary of compatibility test results for each drive Reports unexpected errors as they occur Provides a capability for entering editing saving recalling and executing test programs designed by the user Provides an interactive command set for use in en tering storing retrieving editing and executing tests NOTE In the following steps repair or replace the failing module rerun the program that detected the failure then return to step 1 1 Load diagnostic DZR6A run two passes with no errors 2 Load DZR6B run two passes with no errors 3 Load DZR6C run two passes with no errors 4 Load DZR6D run two passes with no errors 5 7 6 5 2 2 2 Procedure Load DZR6E run two passes with no errors Load DZR6K run two passes with no errors NOTE Any failures occurring during program DZR6K may not necessarily be limited to the RK611 Controller Thus it should only be run on a known good drive CAUTION Diagnostic program DZR6K modifies headers on certain tracks Failure to halt the program either during the End of Pass printout or by typing CTRL C may leave the pack formatted incorrectly and cause inaccurate error reporting in subsequent testing If the formatting is le
24. requested When the status is obtained a Clear command is executed to clear the Attention signal Headers are read and compared with the desired disk address until the correct sector is found Transfer of data through the Silo data buffer to memory is initiated When the sector data transfer is complete the ECC logic is checked to ensure that the data read from the disk was error free If a data error occurred the ECC correction algorithm is initiated to determine whether the error is correctable when finished the command is terminated to allow software to apply the correction information Assuming no data errors the word count in RK WC is checked if nonzero the data transfer operation is repeated into the next sector The word count is checked at the end of each sector until it reaches zero at which time the command is terminated by setting the Ready bit Figure 7 28 7 27 MESSAGE DR SEL CODE SEEK FORMAT TRACK ADD PARITY FROM RKCS2 BIT BIT FROM RKDA ODD BITS 0 2 X 0 1 BITS 8 10 00 01 03 04 05 06 07 08 09 10 11 12 13 14 15 MESSAGE B 5 0 4 065 0 6 DC8 X NE DESIRED CYLINDER ADD PARITY FROM RKDC BITS 0 9 X 0 1 11 4226 a First Message 00 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 DR SEL CODE FROM RKCS2 X 0 1 MESSAGE PARITY ODD 11 4227 b Second Message Figure 7 28 Data Handling Commands Write Data 23 The following sequence is executed entir
25. Error DRERR Bad Sector Error BSERR Header VRC Error HVRC Operation Incomplete OPI DTC Parity Error DTC PAR Write Check Error WCE Nonexistent Memory NEM Data Late Error DLTERR Drive Timing Error DTE Data Check DCK Multiple Drive Select M DS Error Correction Hard ECH Controller Time Out CTO Seek Incomplete SKI RK611 RK06 or RK611 RK07 Disk Subsystem Diagnostic Errors Cont Indicator RKCS2 bit 13 RKCS2 bit 12 RKDS bit 0 Message BO T FAULT RKER bit 7 RKER bit 8 RKER bit 13 RKCSI bit 13 RKCS2 bit 14 RKCS2 bit 11 RKCS2 bit 15 RKER bit 12 RKER bit 15 RKCS2 bit 9 RKER bit 6 RKCSI bit 11 RXER Bit 1 Bit Condition Error in 16 bit data word from main memory Drive response problem Drive is not available dual access only Any of 21 drive error conditions Data transfer attempt to from bad sector Header check indicates bad header information Desired header cannot be found Parity error in message from drive Write check indicates data from disk did not match data from memory No SSYN within 10 us of assertion Data late to from silo Write clock loss during write data loss during read Data error detected by ECC than onedrive simultaneously selected Data error uncorrectable using ECC Command possibly not executed Seek operation did not complete 5 2 E
26. F Maximum temperature rate of change 20 C hour 36 F hour 10 to 90 for a maximum wet bulb temperature of 28 C 82 F Relative humidity Maximum altitude 2440 m 8000 ft Bit transfer rate unbuffered nominal 4 30 M s cell width 232 5 ns Latency Rotational frequency 2400 rpm 2 5 Average 12 5 ms 1 2 rotation 2 5 Maximum 25 0 ms Seek Times Average 38 ms Maximum 75 ms Start Stop times 60 seconds maximum 30 seconds nominal Model Designations RK06 EA 90 128 Vac 60 0 5 Hz RK06 EB 180 264 Vac 60 0 5 Hz RK06 EC 90 128 Vac 50 0 5 Hz RK06 ED 180 264 Vac 50 0 5 Hz NOTE Models RK06 FA through FD are the dual access models corresponding to models EA through ED re spectively Table 1 2 06 Disk Drive Performance Specifications Cont Characteristics Specifications RKO06K DC Cartridge Temperature stabilization time If the cartridge and drive are at approximately the Operating temperature range Allowable relative humidity for operation Storage temperature range for recorded disks Allowable relative humidity for storage same temperature 30 min if there is some differ ence up to two hours depending on the difference 16 C to 49 C 60 F to 120 F 8 to 90 for a maximum wet bulb reading of 28 C 82 F 40 C to 65 C 40 to 150 8 to 80 for a maximum wet bulb reading of 28 C 82 F Table 1 3 RK07 Disk Drive Performance Specifications Chara
27. M7706 BOARDS USED WITH DUAL ACCESS ONLY OTHERWISE ONLY ONE IS USED AS FOLLOWS IN SLOT 2 WHEN B ACCESS IS SELECTED IN SLOT 3 WHEN A ACCESS IS SELECTED 2 DUAL PORT LOGIC USED ONLY WITH DUAL ACCESS OPTION DC POWER J2 PORT A out J4 PORT ioe OUT J1 PORT A IN J3 PORT N IN INPUT OUTPUT CONNECTORS CP 2456 Figure 4 2 06 Disk Drive Logic Module Locations 4 4 TRACK POSITION SENSOR M7908 SERVO ANALOG PCB M7906 MAINT 0 SAFETY ADJ SW 2 MAINT JRun SERVO RELEASE SW 1 SERVO CONTROL LOGIC M7907 OO O O O 0 29 28 27 26 25 24 23 22 21 20 OFFSET 2 21 RTZ LD FWD DRIVE CONTROL LOGIC M7705 HDS UNLD C BRUSHES DUAL PORT DRIVE CONTROL LOGIC 7730 SEE NOTE 2 4 OU INV B PORT ADDR SEIZED REQUESTED INV SW INTERFACE amp TIMING LOGIC M7706 SEE NOTE 1 ee C M DRIVE SELECT A INTERFACE amp TIMING LOGIC M7706 SEE NOTE 1 2 DRIVE SELECT B CABLE BOARD 1 POWER CONN NOTES 2 1 TWO M7706 BOARDS USED WITH DUAL ACCESS ONLY OTHERWISE PORT ONLY ONE IS USED AS FOLLOWS A IN SLOT 2 WHEN ACCESS tS SELECTED OUT IN SLOT 3 WHEN A ACCESS IS SELECTED 2 DUAL PORT LOGIC USED ONLY WITH DUAL ACCESS OPTION PO
28. Message 0 7 has been set in the drive which enables the setting of CERR bit 15 in RKCS1 to indicate one or more of the following AC low error Message BO T6 Nonexecutable function Message 0 8 Controller to drive message parity error Message 0 79 Seek incomplete Message 0 10 Invalid address Message 0 5 e Seek and no motion Message 1 12 Limit detected on seek Message B1 T13 Servo unsafe Message 1 14 Write lock error Message 11 Speed loss error Message 12 Drive off track Message 0 13 09 In addition to setting CSA bit and asserting Attention these drive fault conditions will cause the heads to unloaded h Bead He unsafe Message BO T14 Sector error Message 1 4 Write current no write gate Message 5 Write gate no transitions Message 6 Head fault Message 1 7 Multiple head select Message 1 8 Index error Message B1 T9 Tribit error Message B1 T10 Servo signal error Message B1 T11 NOTE A Drive Status Change DSC is also indicated by the setting Message A0 T14 of bit 14 in REMR2 Status Valid SVAL Bit 15 Status Valid is a read only bit that is set to indicate that the bits in both the Drive Status RKDS and Error RKER registers have been updated for the selected drive via receipt of a complete status message The bit is cleared by conve
29. RK611 data interface contains all of the logic required to transfer data 16 or 18 bit words between main memory and a selected drive Since Unibus transfers require parallel data flow while transfers to and from a disk require serial data flow the data interface also contains parallel to serial memory to disk and serial to parallel disk to memory conversion logic In addition a bidirectional read write data line is used to accommodate the exchange of serial data between the RK611 data interface and a selected drive Finally the data interface logic contains a 66 word data buffer Silo that is used to compensate for the timing differences that exist between the average Unibus data transfer rate 4 3 us word and the burst transfer rate of the disk 3 7 us word 2 4 3 Controller Registers The controller contains 15 Unibus addressable registers Table 2 1 Nine of the 15 registers provide temporary storage for the manipulation of status and or error information This information coupled with the inherent diagnostic capabilities of the controller allows a programmer to create diagnostics that can be extremely effective as a fault isolation aid R READ WRITE SURFACE READ WRITE ga SURFACE R W R SERVO SURFACE TOP DISK BOTTOM DISK READ WRITE 2 SURFACE R W HEAD ORIENTATION 3 IDENTICAL READ WRITE 1 SERVO SURFACE SURFACES 0 1 2 CYL 410 411 PRERECORDED CYLINDE
30. Sector Pulse MSP bit pit 6 in RKMRI to simulate the generation of a sector pulse 6 Toggle MCLK in conjunction with the Read Data bit bit 9 in RK MR1 to simu late the transfer of 128 of the 255 Header Preamble Zeros 7 Check for the setting of the Read Gate RDGT signal bit 15 in RKMRI 8 Toggle MCLK in conjunction with MERD to simulate the transfer of the remaining 127 Header Preamble Zeros concluding with the generation of a single One bit to reflect the SYNC bit 9 Simulate the transfer of the three header words 10 Simulate the transfer of the Gap bits 64 zeros 7 32 11 12 If header compare is correct following simulation of 128 of the 255 Data Preamble Zeros the RDGT signal bit will again be set and the remaining sector bits must be generated as follows a Remaining eight words of Data Preamble 127 zeros 1 One bit b 256 words of 16 18 bit data 4096 4608 bits c Two words of ECC 32 bits d One word of Postamble 16 bits e Seven words of Gap 112 zeros for 16 bit data or six words of Gap 96 zeros for 18 bit data both require approximately six NPR cycles 6 X 37 or 40 data clock cycles to reach the end of a sector If the word count in RKWC is not equal to zero indicating that another sector must be read return to step 5 of the sequence and continue simulation However if the word count is equal to zero 37 additional NPR cycles are required to complet
31. THERE ARE TWO MODES OF OPERATION MANUAL MODE ALLOWS SELECTION OF DRIVES AND HEADS BY TTY INPUT AND AUTO MODE ALLOWS DRIVES AND HEADS TO BE SELECTED BY OFF ON OPERATION OF DRIVE PORT SELECT SWITCHES IN EITHER MODE UP TO 5 MINUTES OF SEEK EXERCISES MAY BE REQUESTED FOR EACH DRIVE ALSO IN EITHER MODE A VERIFY OPERATION ALLOWS HEAD SELECTION WITHOUT THE UNLOADING AND LOADING OF THE DRIVE BY THE PROGRAM WHICH OTHERWISE OCCURS TO ALLOW MANIPULATION OF THE ALIGNMENT TOOL TO RESTART EITHER MODE TYPE 2 TO RESTART ALIGNMENT AID TYPE TO SELECT NEW DRIVES IN MANUAL MODE TYPE C FOR HEAD ALIGNMENT PROCEDURE REFER TO FIELD TEST BOX RK06 07TA RK06 07TB OPERATOR S MANUAL MANUAL OR AUTO MODE M OR A M MANUAL SELECT MODE ENTER DRIVE NO 0 7 0 DRIVE SER NO ALIGN VERIFY OR EXERCISE A V OR E A 4 21 MANUAL SELECT ALIGNMENT ENTER HEAD NO 0 2 0 lt R gt WHEN READY R HEADS POSITIONED AT CYLINDER 365 OCT HEAD 0 SELECTED ENTER HEAD NO 0 2 1 lt R gt WHEN READY R HEADS POSITIONED AT CYLINDER 365 OCT HEAD 1 SELECTED ENTER HEAD NO 0 2 2 TYPE lt R gt WHEN READY R HEADS POSITIONED AT CYLINDER 365 OCT HEAD 2 SELECTED ENTER HEAD NO 0 2 2 ALIGN VERIFY OR EXERCISE E TYPE lt R gt WHEN READY R RANDOM SEEK EXERCISES IN PROGRESS ON DRIVE 0 Z ALIGN VERIFY OR EXERCISE A V OR E V w
32. bits 10 15 are zero 7 2 10 Data Buffer Register RKDB The RKDB register Figure 7 11 can be read or written via program control Reading from the register empties the Silo while writing into the register fills the Silo Both the RKDB register and the Silo are cleared by conventional initialization INIT CCLR SCLR 15 14 18 12 n 10 09 08 06 05 04 03 02 0 00 DB DB DB DB DB DB DB DB DB DB DB DB 5 14 13 12 10 09 07 05 04 02 11 4209 Figure 7 11 Data Buffer Register 7 2 11 Maintenance Register RKMR1 The RKMRI register Figure 7 12 can be read or written via program control and is primarily used to provide an operational analysis of the subsystem i e RK611 Controller and the selected drive This includes drive message analysis and parity testing and the initiation of a Diagnostic mode DMD of operation in which the controller can be isolated and exercised 5 14 2 0 09 08 07 06 05 04 02 00 RD 5 5 5 ECCW PCD PCA MCLK MSP pat E be ee READ ONLY RO RK06 OR RKO7 11 4210 Figure 7 12 Maintenance Register 1 7 17 Message Select MS0 MS3 Bits 0 3 RK06 RKO7 Message Select MSO and 1 bits define one of the four pairs of 16 bit status messages AO A3 and BO B3 that can be simultaneously delivered to the controller on the Message A and Message B lines
33. disabled in the drive to prevent any erroneous drive selection 2 5 2 Subsystem Commands The 13 subsystem commands Figure 2 4 can be divided into two basic groups One group nondata handling is concerned with the various operational requirements of the drives while the second group data or header handling is concerned with the transfer of data and header information to or from a selected device 2 5 3 Data Formatting When the tracks of a disk are formatted via separate Write Header commands the following informa tion is recorded to identify each sector in relation to its radial and rotational position on a disk Figure 2 3 1 Header Preamble field contains 16 words consisting of 255 Zero bits and a single One bit SYNC that defines the start of valid information 2 The Header field consists of two words a Word contains the cylinder address which defines the radial location of the sector in relation to the 411 for the RK06 or 815 for the RK07 addressable cylinders on the disk b Word 2 contains the sector address which defines the sector in relation to the number 20 or 22 of sectors per track and the track address which defines one of the three R W heads and therefore the appropriate R W surface in relation to the disks a Format bit that is a zero to reflect 16 bit data or one to reflect 18 bit data words Finally the Good Sector flag GS FLAG bits are set to indicate that the sector is
34. from the servo track 8 Servo Signal Error Message B1 T11 indicates that no signals have been detected from the servo head In addition to the setting of the UNS bit in the RKER register the occurrence of any one of these fault conditions allows the CDA bit to be set in the RKDS register the Attention signal to be asserted for the RKAS OF register the DSC bit Message A0 T14 to be set in RKMR2 and the Fault bit Mes sage 7 to be set in RKMR3 Data Check DCK Bit 15 Data Check is a read only bit that is set to indicate that a data error was detected when the current sector was read 7 2 8 Attention Summary Offset Register RKAS OF The RKAS OF register Figure 7 9 can be read or written via program control and as such is used to store the head offset value required by an Offset command and the current condition of the Attention signal line that is monitored for each drive 11 10 08 07 01 15 14 6 04 02 7 READ ONLY 11 4207 Figure 7 9 Attention Summary Offset Register 7 15 Offset OF0 OF7 Bits 0 7 The Offset field 7 defines both the magnitude 5 and direction OF7 of head movement in relation to the centerline of a track As shown below each binary increment of the offset value excluding OF6 produces a move of 25 microinches for an RK06 or 12 5 microinches for an RKO7 in a positive or negative direction A positive offset 0 provides head mot
35. initiated without disabling intradrive cabling The logic and power board cages can be extended following removal of the rear panel of an RK06 or 7 Disk Drive to facilitate emergency maintenance Mean time to repair MTTR for emergency maintenance is improved by inclusion of light emitting diodes LEDs on the front edge of certain logic boards The LEDs provide a service engineer with an indication of the functionality of certain key logic areas For the implementation of cartridge alignment procedures a protection switch is available to the service engineer to provide write protection for the cartridge and control of head velocity to ensure alignment integrity If an RK06 or RKO7 Drive loses ac power neither the drive nor cartridge will be harmed data will be lost and the format will remain intact When dual access RK06 or RKO7 Disk Drives are installed the logic allows dual access func tionality to be tested utilizing only one of the two available controllers This feature provides enhanced availability for multiprocessor systems by freeing one controller for normal dual access operations 1 2 4 Reference Documentation These documents supplement the RK06 RK07 User s Manual Number Title EK RK067 TD 001 RK06 RKO7 Disk Drive Technical Description Manual EK RK611 TM 002 RK611 Controller Technical Manual EK RK067 SV 001 RK06 RKO7 Disk Subsystem Service Manual EK RK06 IP 001 RK06 Illustrated Parts Breakdown E
36. lid Check the five output voltages at the logic backplane Backplane Location Voltage BO8A2 5 Vdc 3 D01E2 4 25 10 C08D2 15 Vdc 5 08 2 15 5 08 2 5 3 If any voltage is not within tolerance refer to the 06 RK07 Disk Subsystem Service Manual for adjustment and or repair procedures 4 43 Head Load Unload Checks The following head load unload checks are performed during the POWER ON sequence to ensure that operational status can be attained 1 2 With power install a scratch cartridge into the drive Depress the appropriate access select switch either A or B Press the RUN STOP switch After approximately one second the spindle will begin to rotate and the RUN STOP light will extinguish NOTE Power must be applied to the controller or to the Field Test Box if used during performance of the load unload checks 4 16 When disk reaches nominal speed after approximately 30 to 60 seconds the heads will load onto the disk surfaces and begin a slow movement toward the inner limit After reach ing this limit they then return to cylinder zero When head motion has stopped the READY light on the operator control panel should be lit The FAULT light should not be lit Press the RUN STOP switch again The READY light should extinguish and the heads move to their fully retracted position When the drive has stopped the RUN STOP in dicator will light The drive
37. motor should stop in approximately 20 seconds after the RUN STOP switch is released 4 4 4 Head Alignment Checks and Cartridge Care Several principles should be understood before attempting any alignment procedures A familiarity with these principles will hopefully discourage any shortcuts leading to possible damage or im proper alignment 1 10 Whenever head alignment exceeds 200 microinches maximum heads should ligned to within 50 microinches Power must be applied to the drive for two hours prior to performing any head alignment procedures The cartridge should be mounted and spinning for a minimum of one half hour prior to beginning head alignment Head alignment techniques assume that the drive being aligned is a properly working device If the device is not working properly do not attempt alignment To protect the alignment pack Write Protect should always be set when the alignment pack is in the machine The SAFETY SWITCH on RK06 module M7729 or RK07 module M7906 near the D end forces Write Lock and disables the Servo Unsafe circuit It is important to remove this special maintenance switch from its maintenance mode to re enable the Servo Unsafe cir cuit Linear forces applied to the carriage during adjustments may cause the heads to unload and cause damage if the SAFETY SWITCH is not set to the MAINT position Lateral force on a carriage can cause the heads to tilt and possibly crash T
38. non existent drive error in the port controller 9 2 2 Port B Mode With only the enabling switch depressed the drive will respond to Any attempt by Port A to issue a command to the drive will generate a nonexistent drive error the Port A controller 9 2 3 Dual Access Mode With Port A and Port B enable switches depressed the drive is in the programmable mode In this mode the drive is initially available to both ports Once the drive has been seized by one of the ports that port has full control over the drive and can issue any legal command A port seizes a drive if the controller selects the drive through that port and the drive was available 93 DUAL ACCESS LOGIC CIRCUITS For the dual access option the drive houses two additional modules an M7730 dual port board and an M7706 interface and timing module The two interface and timing modules interface with each controller The dual port board contains logic for controlling the access from both ports and acts as a multiplexer for commands and timing from each port to the remainder of the drive electronics The dual port option also uses a different backplane than the single port option 9 2 The logic for the dual port option is located the M7730 module and consists of 1 Arbitration Circuits 2 Timing Counters and Decoders 3 Auto Release Timer 4 Multiplexers 9 3 1 Arbitration Circuits These circuits determine and control the state of t
39. received an invalid head or cylinder address A Seek and No Motion Message B1 T12 bit has been set in the drive indicating that a Seek command was received by the drive but no track crossing pulses were detected within 10 ms A Limit Detected on Seek Message B1 T13 bit has been set in the drive indicating that an inner limit or an outer limit was detected during a Seek operation If this fault occurs the heads are unloaded A Servo Unsafe Message B1 T14 bit has been set in the drive indicating a servo runaway condition If this fault occurs the servo actuator is disconnected from the servo drive and connected to a battery that provides the necessary power to drive the heads to the home posi tion addition to setting the CSA bit and asserting Attention these drive fault conditions will cause the heads to be unloaded 7 12 15 14 13 12 11 10 09 06 07 06 05 04 03 01 00 i M READ ONLY 11 4206 Figure 7 8 Drive Error Register In addition to the setting of the SKI bit Message B0 T10 in the register the occurrence of any one of these faults allows the CDA bit bit 14 to be set in the RKDS register the Attention signal to be asserted for the RKAS OF register the DSC bit Message A0 T14 to be set in the RK MR2 register and the Fault bit Message 7 to be set for the RKMR3 register Nonexecutable Function Bit 2 Nonexecutable F
40. selected drive becomes true a drive status is requested When the status is obtained a Clear command is executed to clear the Attention signal The RK611 then waits for Index from the drive When Index is detected zeros are written until sector pulse is detected Then the header preamble including sync 1 and the three header words are written The all zero gap the data preamble including sync 1 and all zero data ECC postamble and end of sector gap are writ ten This is repeated in each successive sector until Index is encountered again and the command is terminated The Ready bit is set NOTE All three words of the header including the check word are prepared by software and treated as data by the controller Only one complete track can be formatted at a time Either 20 or 22 headers 60 or 66 words will be written depending on the state of the format bit RKCSI bit 12 at the beginning of the command RKWC must be set to 60 or 66 de pending on format Figure 7 28 Write Check 31 The following sequence is executed entirely by the RK611 A seek to the cylinder in RKDC command is performed When the Attention signal from the selected drive becomes true a drive status 1s requested When the status is obtained a Clear command is executed to clear the Attention signal The selected drive provides data as in a Read command and data is obtained from memory as in a Write command The data are compared on a word for word basis
41. the positioning of the heads to a desired cylinder and the reading of 32 headers the desired header could not be found This error can result from any one of the following 1 Head mispositioning 2 Incorrect head selection 3 Read channel failure 4 Improper pack formatting Drive Unsafe UNS Bit 14 Drive Unsafe is a read only bit that is set Message BO T14 to indicate that any one of the following Read Write Unsafe conditions each of which causes the heads to be unloaded has been detected 1 Sector Error Message BI T5 indicates that Write Gate was asserted in the drive in coinci dence with the trailing edge of a sector pulse 2 Write Current No Write Gate Message B1 T5 indicates that write current has been de tected in the drive without the assertion of Write Gate 3 Write Gate No Transitions Message 1 6 indicates that the drive received the assertion of Write Gate while no encoded write data has been received 4 Head Fault Message B1 T7 indicates that an electrical imbalance exists in the write signals that have been fed to the head which could cause erroneous data to be recorded 5 Multiple Head Select Message B1 T8 indicates that more than one read write head has been selected simultaneously 6 Index Error Message 1 9 indicates that the once per revolution index pulse is either mispositioned or is missing 7 Tribit Error Message B1 T10 indicates that three successive tribits were missing
42. total number for all drives of WORDS READ 65K is less than 967 allow the subsystem to run for an additional 15 minutes and again obtain the statistics Repeat until the total WORDS READ 65K is greater than 967 If the total number of SOFT DATA ERRORS is zero or one accept the subsystem If it is three or more the failing component drive or controller must be fixed If there were two SOFT DATA ERRORS continue running the subsystem for an additional hour At this point obtain statistics and continue running until the total WORDS READY 6SK is greater than 1934 If the total SOFT DATA ER RORS is still two accept the subsystem If the total is three or more the failing component drive or controller must be fixed and acceptance restarted 5 9 6 OPERATOR S GUIDE 6 1 INTRODUCTION The following material describes the function of each of the external controls available to the user of an 06 or RKO7 Disk Drive and provides procedural information for subsystem operation 6 2 OPERATOR CONTROLS INDICATORS The 06 RKO7 Disk Drive contains the following power and control switches and indicators Figure 6 1 Rear Panel Circuit Breaker AC power ON OFF switch with POWER indicator Front Panel Control Panel Combination RUN STOP switch with a STOP indicator UNIT SELECT switch with a READY indicator FAULT indicator WRITE PROT switch with a PROTECT indicator ACCESS A ENABLE switch with a SE
43. type of drive and WRITE LOCK the drive NOTE Prior to begining head alignment procedures power should be applied to the drive for two hours the car tridge should be mounted and spinning for at least one half hour To WRITE LOCK the drive set the SAFETY SWITCH S2 on the RK06 M7729 board or the RK07 M7906 board to the MAINT position This action will also prevent the heads from unloading if a Servo Unsafe condition occurs 12 13 14 16 17 18 19 Respond to program questions 1 through 5 as illustrated in sample printout graph 4 4 4 2 Mount the head alignment fixture on the appropriate head Figures 4 10 and 4 11 CAUTION Make sure that the carriage does not tip beyond the point where the heads will fall out of the cam tower Torque the head alignment fixture mounting screw to 5 1 Loosen the head mounting screw and then torque it to 2 in Ib Turn both vertical screws of the head alignment fixture fully CCW Load the heads by typing the letter R Set the torque wrench to 5 in lb Tighten the forward vertical screw to move the head forward or the rear screw to move the head toward the rear Since these screws cause oposite actions the screw not in use must be turned fully CCW Adjust until a red LED on the meter is ON steady state b The meter on the Field Test box reads 0 50 microinches NOTE If the needle deflects to the right of zero move the head toward spindle if t
44. update time alloted to a drive an ATN bit will be set by the assertion of an Attention Line signal if the Drive Status Change flip flop is set Thus the clearing of the flip flop within the drive will clear the associated ATN bit The condition of the Drive Status Change flip flop is also noted in the controller by the conditioning Message A0 T14 of the DSC bit bit 14 in the RKMR2 register 7 2 9 Desired Cylinder Register RKDC The RKDC register Figure 7 10 can be read or written via program control and is used to store the address of the desired cylinder Following an initial load the value in the RKDC register will be incremented by one whenever the track address TAO TA2 value in the RKDA register overflows during a data transfer When the RKDC register is incremented and the RK WC register is not equal to zero a single cylinder seek is initiated by the controller 15 14 13 12 11 10 09 08 07 Figure 7 10 Desired Cylinder Register 06 0 5 04 03 01 00 C DC DC DC DC DC 4 3 2 1 0 11 4208 Desired Cylinder DC0 DC9 Bits 0 9 For an RK06 Drive valid cylinder addresses range from outer edge of the disk 0 to the center 4100 requiring nine DC DC8 address bits to define the range 000 6323 For an RKO7 Drive valid cylinder addresses range from the outer edge of the disk 0 to the center 81510 requiring ten DCO DC9 address bits to define the range 000 1456 Spares Bits 10 15 Spare
45. valid for the reading or writing of data 3 The Header Check field consists of one word that is an exclusive OR sum of header words 1 and 2 The check word is generated by software and extracted during read operations to verify the integrity of both the Header field and the check word 2 12 4 Header field contains four words consisting of 64 Zero bits which provide time required by the controller to complete a header check and accommodate mechanical tolerances within the drive 5 The Data Preamble field contains 16 words consisting of 255 Zero bits and a single One bit SYNC that defines the start of the Data field gt 6 Data field accommodates 256 16 bit data words 4096 bits or 18 bit data words 4608 bits depending on the format The extra length required by 256 18 bit words is com pensated by the reduction of the maximum number 20 of sectors allowed per track 7 Error Correction Code ECC field contains 2 word correction code that is generated by the controller during a write operation by the total bit configuration of the 256 word Data field The ECC field is read during read operations to verify the integrity of both the data and the ECC words 8 The Data Postamble field consists of one word of Zero bits which allows the write amplifier to be disabled when writing during an unused portion of the sector 9 The Sector Tolerance Gap contains six or seven words depending on t
46. value 00 2119 for a 22 sector format 16 bit data words The Sector Address is incremented by one when the sector has been transferred 7 6 13 12 11 10 09 08 07 06 05 04 03 01 15 14 51 CCLR DI 11 4200 Figure 7 2 Control Status Register 1 15 00 WC wc WC wc WC WC wc WC we WC 15 14 13 12 11 10 o9 08 07 06 05 04 03 02 01 00 11 4201 Figure 7 3 Word Count Register 15 00 15 14 13 12 11 10 09 08 07 06 05 04 03 01 11 4202 Figure 7 4 Bus Address Register 15 14 13 12 11 1 01 0 09 08 07 06 05 04 03 02 o TA TA TA SA SA SA SA 11 4203 Figure 7 5 Disk Address Register Spares Bits 5 7 Spare bits 5 7 are 0 Track Address 0 2 Bits 8 10 The Track Address bits are configured 0 2 to select the appropriate read write head associated with the desired track The Track Address bits are incremented by one after the last sector 23s or 25s on the track has been transferred and the Sector Address has been reset to zero Similarly if transfers continue beyond the last sector of the last track 2 of a given cylinder the Track Address is reset to zero and the Cylinder Address in is incremented by one In this manner subsequent sectors tracks and cylinders can be consecutively transferred until the word count eq
47. 02 00 E PARITY DRIVE SERIAL NUMBER SELECTED ADDRESS 11 4214 Figure 7 16 Maintenance Register 2 Drive to Controller 7 21 MR3 RESERVED CYLINDER ADDRESS OFFSETS RESERVED MESSAGE PARITY FOR ADDL REQUEST MESSAGE 11 4216 Figure 7 17 Maintenance Register 3 Controller to Drive 15 14 13 12 11 10 09 08 07 06 05 04 05 01 00 WRITE EXECUTE parity UNSAFE prive WRITE c p FUNCTION INVALID OFF LOCK PARITY FAULT ADDRESS TRACK ERROR ERROR BI TRIBIT WRT GATE SECTOR UNSAFE ERROR amp NO ERROR MESSAGE 1 0 MOTION TRANSITIONS gt DETECT SIGNAL ERROR FAULT 8 NO WRT GATE ON SEEK ERROR 15 14 13 12 11 10 09 06 05 04 01 00 9 8 7 6 5 4 3 2 t 0 T 2 2 2 2 2 2 2 2 2 2 ENS 512 256 128 64 32 16 8 4 2 1 PARITY CYLINDER ADDRESS BITS MESSAGE 10 15 14 13 12 11 10 09 08 07 06 05 04 02 Ol 4 2 1 2 2 2 E i 1 w 16 8 4 2 1 DECODED SECTOR MESSAGE LD HEAD ADDRESS COUNT 11 4215 Figure 7 18 Maintenance Register 3 Drive to Controller 7 22 Table 7 2 Disk Commands Command Function Code Octal F4 2 Fi GO Nondata Handling Select Drive 0 0 0 0 1 01 Pack Acknowledge 0 0 0 1 1 03 Drive Clear 0 0 1 0 1 05 Unload 0 0 1 1 1 07 Start Spindle 0 1 0 0 1 11 Recalibrate 0 1 0 1 1 13 Offset 0 1 1 0 1 15 Seek 0 1 1 1 1 17 Data or Hea
48. 06 Drives It has a maximum for matted storage capacity of 6 3 million 18 bit words or 6 9 million 16 bit words The RK07 Disk Drive uses a similar cartridge the RK07K Disk Cartridge as the storage medium It also can be inter changed between any of the available RK07 Drives It has a maximum formatted capacity of 12 6 million 18 bit words 13 8 million 16 bit words Note that an RK06K cartridge cannot be used on an RKO7 Drive an RKO7K cartridge on an 6 Drive Each dual platter pack provides three recording surfaces that are serviced by three read write heads and a servo surface that is monitored by a single read head The recording surfaces provide three data tracks per cylinder with a maximum of 411 cylinders per RK06 disk pack or 815 cylinders per RK07 disk pack Each track can accommodate either 20 18 bit word or 22 16 bit word sectors while each sector can accommodate 256 data words The servo surface provides positioning signals for the heads and timing signals for data transfer synchronization 1 2 high density recording capability required to achieve maximum storage capacities is provided by the use of the modified frequency modulation MFM encoding technique In addition each RK06 and 07 Disk Drive and the controller contain phase locked oscillator PLO circuitry which main tains a constant bit density if a disk is subjected to minor speed variations For head positioning the maximum seek time in t
49. 115 V 50Hz 6 3 A rms at 230 V 50 Hz 0 80 minimum Less than 90 V rms for one cycle or more 2 7 m 9 ft NEMA 5 15 for 120 Vac nominal models NEMA 6 15P for 240 Vac nominal models 10 C to 40 C 50 F to 104 F 20 C hour 36 F hour 10 to 90 for a maximum wet bulb temperature of 28 C 82 F 2440 m 8000 ft 4 30 M s 232 5 ns 2400 rpm 2 596 12 5 ms 1 2 rotation 2 590 25 0 ms Table 1 3 RK07 Disk Drive Performance Specifications Cont Characteristics Specifications Seek times Average 36 5 ms Maximum 71 0 ms Start Stop times 60 seconds maximum 30 seconds nominal Model Designations RKO7 EA 90 128 Vac 60 0 5 Hz RKO07 EB 180 264 Vac 60 0 5 Hz RKO7 EC 90 128 Vac 50 0 5 Hz RK07 ED 180 264 Vac 50 0 5 Hz NOTE Models RK07 FA through FD are the dual access models corresponding to models EA through ED re spectively RK07K DC Cartridge If the cartridge and drive are at approximately the same temperature 30 min if there is some differ ence up to two hours depending on the difference Temperature stabilization time Operating temperature range 16 Cto49 60 Cto 120 8 to 90 for a maximum wet bulb reading of 28 C 82 F Allowable relative humidity for operation Storage temperature range for recorded disks 40 C to 65 C 40 F to 150 Allowable relative humidity 890 to 8096 for a maximum wet bulb reading of 28 for storage
50. 3 15 are 0 7 2 13 ECC Pattern Register RKECPT The Error Correction Code ECC Pattern register Figure 7 14 is read only register that contains 11 bit error correction pattern from the ECC polynominal generator The pattern is valid if the error is ECC correctable Error Pattern 10 Bits 0 10 The Error Pattern bits are read only bits that provide an 11 bit correction pattern for an error burst that does not exceed 11 bits in length and is therefore ECC correctable Spares Bits 11 15 Spare bits 11 15 are 0 15 14 13 12 11 10 09 08 07 06 05 04 03 01 00 o 0 10 09 08 07 06 05 04 03 02 01 V ESS READ ONLY RO 11 4212 Figure 7 14 ECC Pattern Register 7 19 7 2 14 Maintenance Register 2 2 RKMR2 is a read only register that is used for Message A controller to drive transmissions eon mands as well as Message AO A3 drive to controller transmissions status Controller to Drive Message A For controller to drive transmissions Figure 7 15 the 16 bit Mes sage A shift register assembles disk commands along with an appropriate odd parity bit from various device registers and command logic within the controller As the bits are assembled in the register they are serially transmitted to the drives in coincidence with the transmission of Messa
51. 4 20 16 3 1 40 100000 000200 000100 000001 040000 100000 040000 000100 UNIT DSKADR BUSADR WCNT CYLADR FUNCTN RK6ACT PROPRO OFFSET WORD WORD WORD WORD WORD WORD BYTE WORD RKCS1 base Unibus address RKCS2 RKDS offset from RKCSI RKER offset from RKCSI RKDC offset from RKCSI offset from RKCSI Basic Pack Acknowledge function Basic Select Drive function Subsystem Clear Status Valid Drive Ready Volume Valid Drive Available Drive Unsafe Controller Error Controller Clear Drive Interrupt Interrupt Enable Desired unit in bits 0 2 Track sector address Low order 16 bits of Unibus address Two s complement word count Desired cylinder address Desired function IE A16 A17 GO 1 if RK06 currently active 1 if doing Seek function Value to load into Offset register 7 39 RK06 5 RK6ACT Is the RK06 currently active BNE RK06 Wait for it to become inactive MOV RKCS2 R2 Point to CS2 register MOV SCLR 2 Issue a Subsystem Clear MOV UNIT R2 Select the desired unit MOV DSKADR R2 sLoad RKDA MOV BUSADR R2 RKBA MOV WCNT R2 RK WC MOV PACACK R2 Issue Pack ACK function 15 TSTB R2 15 us for the serial message BPL 1 RKDS R2 R1 Get Drive Status register COM Complement bits BIT SVAL DRDY VV DRA R1 Can drive accept
52. 7 Provide closed cabinets for disk storage 8 Clean and vacuum subfloor areas and air conditioning systems just prior to installation 9 Place impregnated mats at each entrance to reduce the amount of dust tracked in from other areas The disk drive air circulation system Figure 3 2 is designed to supply clean air for l Disk temperature control 2 Cleaning of the shroud area by low pressure 3 4 inch of water at 70 cfm of air 3 Card cage cooling 4 Power suply power amplifier assembly cooling To ensure that exhausted air moves freely floor air conditioning outlets should not be placed at the rear of the drive HUMAN HAIR 004 DIA NC N FINGERPRINT SMUDGE SMOKE PARTICLE 250 4 FLYING 5 HEIGHT ANE 2 ZZ Figure 3 1 Relationship of Disk Head Disk and Contaminants NN 11 4978 3 3 CARTIDGE CFM AIR HOSE ABSOLUTE FILTER CARD CAGE ASSY lt ee 2 29 t E E 4 PREFILTER NOT SHOWN ly DUAL BLOWER POWER SUPPLY POWER 70 CFM AIR DUCT AMPLIFIER ASSY BELLOWS 11 4982 Figure 3 2 RK06 RK0 7 Disk Drive Air Supply and Distribution Subsystem 3 4 3 2 SITE PREPARATION 3 2 1 Space Provision should be made for service clearances of 916 mm 36 in at the front and rear of the RK06 or 07 disk drive Space should also be made available in the system environment for storage of disk cartridge
53. A Seek or any other head motion command will be ignored by a drive if the Drive Ready DRDY bit is not asserted in the RKDS register when the command is initiated Dual Access Operation The programmer should be aware of a problem that can occur only with the Dual Access Option See Paragraph 9 4 3 7 44 8 CUSTOMER MAINTENANCE GUIDE 8 1 INTRODUCTION User maintenance procedures limited to the care and cleaning external of the RKO6K or RK07K Disk Cartridge and the cleaning of the drive spindle assemblies 8 2 PROFESSIONAL CARTRIDGE CLEANING Professional cartridge cleaning is recommended every 6 months or whenever practicable Complete cartridge cleaning procedures must be performed by either qualified DIGITAL Field Service per sonnel in emergency or troubleshooting situations or by a professional cleaning service Application of cleaning procedures to the actual recording surfaces by unqualified personnel will void not only the warranty on the serviced cartridge but the warranty for any drive on which the cartridge is oper ated 8 3 USER CARTRIDGE CLEANING The outer sides of a completely assembled cartridge may be cleaned by the user with a lint free wiper dampened with a solution of 9 percent water and 91 percent isopropyl alcohol However the cartridge must not be saturated and all excess solvent must be removed with a dry wiper This is necessary to prevent solvent from entering the seams of the assemb
54. A releases the drive while Port B is still in the process of issuing a release command the Port B controller will find the attention asserted The assertion of attention should have been prevented by the release of Port B This race condition can be avoided by not allowing a port to release its port request to an unavailable drive 9 4 4 Auto Release Timer At the initiation of every command to the drive by the port that has access a one second release timer is started The exception to this is the release command This timer is restarted with every new command issued by the controller having access to the drive If a port accessing the drive does not release the drive and the one second release timer times out the other port s port request flip flop will be examined If it is reset the drive will go back to the neutral state If the other port s port request flip flop had been set the following events take place 1 The drive becomes deselected from the port that had access 2 The drive is effectively seized by the other port 3 other port s attention is asserted signifying that the drive is now available 95 START STOP AND WRITE PROTECT IN DUAL ACCESS When ac power is first applied to 6 RK07 the drive is initially in the neutral state The volume valid bit for each port is cleared A pack acknowledge command is required for each port The drive will not seek or write through a port unless that port has its Vol
55. E A DR SEL CODE FORMAT PARITY FROM RKCS2 BIT ODD BITS 0 2 X 0 1 0 1 00 02 04 05 06 08 09 10 fl 12 13 14 15 MESSAGE ICM OFFSET DISTANCE VALUE FORCED PARITY 4 9 AND DIRECTION BIT 10 SET ODD FROM RKAS OF BITS 0 6 X 0 1 11 4223 Figure 7 25 Offset Command 7 26 05 RKAS OF REG ICIEIGES DIRECTION BIT 25 12 5 in 0 25 in or 50g In 200 or 100p 400 in 20 800 or 400 1 RK 6 RKO7 equals head motion toward spindle 11 4224 Figure 7 26 Offset Command Coding Seek 17 Figure 7 27 This command directs the selected drive to relocate the heads over a new cylinder The new cylinder address is derived from the Desired Cylinder register RKDC When Messages A and B are sent a return status is requested When the seek is completed a drive Attention signal is asserted for polling 01 DO 07 DR aes py FORMAT PARITY FRON Diese BIT ODD 0 1 00 01 2 03 04 05 06 07 08 09 10 11 12 13 14 15 MESSAGE CYLINDER ADDRESS PARITY RKDC BITS 0 000 X 0 1 11 4225 Figure 7 27 Seek Command Read Data 21 The following sequence is executed entirely by the RK611 A Seek to the cylinder in RKDC is performed When the Attention signal from the selected drive becomes true a drive status is
56. EK RK067 UG 001 RKO6 RKO7 Disk Drive User s Manual digital equipment corporation maynard massachusetts First Edition February 1978 _ Copyright 1978 by Digital Equipment Corporation The material in this manual is for informational purposes and is subject to change without notice Digital Equipment Corporation assumes no respon sibility for any errors which may appear in this manual Printed in U S A This document was set on DIGITAL s DECset 8000 computerized typesetting system The following are trademarks of Digital Equipment Corporation Maynard Massachusetts DEC DECtape PDP DECCOMM DECUS RSTS DECsystem 10 DIGITAL TYPESET 8 DECSYSTEM 20 MASSBUS TYPESET 11 UNIBUS CHAPTER 1 1 1 1 2 1 2 1 1 2 2 1 2 3 1 2 4 1 3 1 4 2 2 5 6 2 3 2 UUU 99 4 1 ud zs l CONTENTS Page INTRODUCTION dO i E 1 1 GENERALINEORMATION ua 1 2 1 2 Diagnostic and Error Detection Correction Capabilities 1 3 Maintenance Featufes a rae abad tsi 1 3 Reference Documentation vta 1 4 RK611 RK06 AND RK611 RK07 DISK SUBSYSTEM PERFORMANCE SPECIEICA TLIONS tae pa esr 1 4
57. Humidity control is important in a data storage system as static electricity can cause errors in any CPU with memory Either disk subsystem is designed to operate efficiently within a relative humidity range of 10 to 90 with a maximum wet bulb temperature of 28 C 82 F and a minimum dewpoint of 2 C 36 F The nonoperating relative humidity range is from 10 to 95 with a maximum wet bulb temperature of 46 C 115 F The relative humidity range for the disk cartridge is 8 to 90 with a maximum wet bulb temperature of 28 82 3 1 4 Heat Dissipation Heat dissipation of the 06 or is 1500 Btu hr nominal and 1700 Btu hr maximum By adding this figure to the total heat dissipation for the other system components and then adjusting the result to compensate for such factors as the number of personnel heat radiation from adjoining areas sun exposure through windows system efficiency etc the approximate cooling requirements for the sys tem can be determined It is advisable to allow a safety margin of at least 25 percent above maximum estimated requirements 3 1 5 Acoustics Most computer sites require at least some degree of acoustic treatment however the RK06 or RK07 Disk Subsystem will not contribute unduly to the overall acoustic problem Acoustic materials should neither produce nor harbor dust 3 1 6 Altitude Computer system operation at high altitudes can result in heat dissipation problems The maxim
58. K RK07 IP 001 RK07 Illustrated Parts Breakdown 13 RK611 RK06 AND RK611 RK07 DISK SUBSYSTEM PERFORMANCE SPECIFICATIONS Tables 1 1 1 2 and 1 3 are performance specification summaries for the RK611 Controller the RK06 Disk Drive and the RK07 Disk Drive respectively 1 4 OPTIONS Table 1 4 lists the several options currently available for the RK611 RK06 Disk Subsystem As in dicated each drive is contained in an H969 dedicated cabinet and contains all the necessary cabling as well as an RKO6K DC data cartridge E models are single access and models are dual access Similarly E model RK611 Controllers are single units and F model RK611 Controllers consist of two controllers The RK611 C is a dual access controller for the Disk Subsystem and a kit to convert an RKOGE to an RKO6F Table 1 5 lists the several options currently available for the RK611 RK07 Disk Subsystem As in dicated each drive is contained in H969 dedicated cabinet and contains all the necessary cabling as well as an RK07 DC data cartridge E models are single access and F models are dual access Similarly E model RK711 Controllers are single units and F model RK711 Controllers consist of two controllers The RK711 C is a dual access controller for the Disk subsystem and a kit to convert an to an Table 1 1 RK611 Controller Performance Specifications Characteristics Specifications Required mounting space Occupies one double sy
59. LECT A indicator ACCESS B ENABLE switch with a SELECT B indicator 6 2 1 Power ON OFF Circuit Breaker When the three wire plug is inserted into an outlet ac power is applied to the rear panel circuit breaker on an 06 or RK07 Disk Drive and the red indicator lamp is illuminated When the circuit breaker is switched to the ON position ac power is applied to the drive and the blower motor is energized 6 2 2 RUN STOP Switch with STOP Indicator The STOP indicator is illuminated during STOP conditions when the heads and brushes are home the spindle is completely stopped and the spindle motor is not energized When the alternating action pushbutton is initially pressed RUN under these conditions the spindle motor is energized the STOP indicator is extinguished and the heads are loaded When these initial operational conditions are met STOP indicator extinguished and heads loading a RUN condition is entered However if the heads do not load due to a malfunction the spindle motor is immediately de energized and the STOP indicator remains illuminated If the pushbutton is released STOP when the heads are loaded the heads will be unloaded and the spindle motor de energized When these conditions are met heads home and spindle stopped the STOP indicator is again illuminated 6 1 If the RUN STOP switch is in RUN position and drive has been cycled down heads unloaded and spindle completely stopped due to an Unload co
60. O7K AC RKO7K EF 70 12292 08 70 12292 15 70 12292 25 70 12292 40 RK6 7 TA RK6 7 TB RKO7U RK6 7 TU Table 1 5 RK611 RK07 Disk Subsystem Options Description Single access 07 H969 cabinet 115 Vac 60 Hz Single access RK07 H969 cabinet 230 Vac 60 Hz Single access 7 H969 cabinet 115 Vac 50 Hz Single access RK07 H969 cabinet 230 Vac 50 Hz Dual access RK07 H969 cabinet 115 Vac 60 Hz Dual access 07 H969 cabinet 230 Vac 60 Hz Dual access RK07 H969 cabinet 115 Vac 50 Hz Dual access RK07 H969 cabinet 230 Vac 50 Hz RKO07 EA drive with RK611 controller RKO7 EB drive with RK611 controller RK07 EC drive with RK611 controller RKO07 ED drive with RK611 controller RKO07 FA drive with two RK611 controllers RKO07 FB drive with two RK611 controllers RKO7 FC drive with two RK611 controllers RKO07 FD drive with two RK611 controllers Dual access kit Dual access kit plus one Unibus controller Drive data cartridge Drive alignment cartridge Drive error free cartridge 8 ft cable 15 ft cable 25 ft cable 40 ft cable RK06 RK07 Field Test Box 115 Vac 50 Hz or 60 Hz RK06 RKO7 Field Test 230 Vac 50 Hz or 60 Hz Upgrade kit to convert RK06 to RK07 Upgrade kit to convert RK06 TA TB to RK6 7 TA TB 2 SUBSYSTEM DESCRIPTION 2 1 INTRODUCTION The following subsections provide separate drive controller and subsystem descriptions that are in tended to give an overview of the RK611 RK06
61. ORT OPERA TION us anan REIR EMEN 9 3 Seizing a Drive the Neutral State Ashes 9 3 Attempting to Seize an Unavailable 9 3 Caution Issuing a Release Command to an Unavailable Drive 9 4 Te et D coded teet e oceanus 9 4 START STOP AND WRITE PROTECT IN DUAL 9 4 ERROR HANDLING IN DUAL ACCESS 9 5 Notifying the Controller of an 4 5 o eei o 9 5 Clearing the Drive Errors and Resetting the Attention 9 5 Multiple Drive Select and CTD Parity Errors in the Dual Access Configuration 9 5 INITIALIZE INEDUA TA CCESS Sapa Q moses 9 5 GETTING DRIVE STATUS IN DUAL ACCESS 9 6 Status Bits Pertinent to the Dual Port Available at the Unibus Interface 9 6 FIGURES Title Page RK611 RK06 RK611 RK07 Dual Access System Configuration 1 2 X H969 Standalone Disk Drive 2 2 Read Write Heads and Data Track Orientation 2 4 2 5 Basic RK06 or Disk Subsystem a 2 5 Data Interface u
62. RDY bits in RKCSI Finally Interrupt Enable bit 6 can be reset via program control as well as by conventional initialization INIT CCLR SCLR Controller Ready RDY Bit 7 Controller Ready RDY is effectively a read only bit However the bit can be externally set via conventional initialization INIT CCLR SCLR or internally set upon completion of a command The R DY bit is reset when GO bit 0 in RKCSI is set Extended Bus Address BA16 17 Bits 8 9 The Extended Bus Address bits reflect Unibus upper address bits 16 and 17 and as such are an extension of the 16 bit RKBA register which contains the memory address required for the current data transfer Controller Drive Type CDT Bit 10 This bit specifies the type of drive that will be selected by the controller To specify RK06 Disk Drives the bit must be reset For an RK07 Drive the bit must be 1 Controller Time Out CTO Bit 11 Controller Time Out is a read only error bit that is set to indicate that GO bit 0 in RKCSI has been set for approximately 800 ms Since this interval exceeds the time required to execute the longest possible drive operation i e a Seek from cylinder 410 to cylinder 0 followed by a 65k word data transfer the set condition of this bit indicates that the last command has not been completed due to a malfunction Controller Format CFMT Bit 12 The Controller Format bit specifies the number of sectors for each track in the selected d
63. RIPED BERG r S EDGE CONN S3 SN REF v i X STRIPED STRIPED EDGE NC TRANSITION BRACKET CABLE ROUTE FOR HORIZONTAL RIGHT CAB MOUNTED CONTROLLER BA11 K UPRIGHT REF SEE NOTE 2 CABLE ROUTE FOR VERTICAL MOUNTED CONTROLLER BA11 F NOTES 1 FOR INSTALLATION IN BA11 F EXPANSION BOX a CUT THE WRAP AND UNFOLD POWER HARNESS b ADD EXTENSION POWER CABLE TO SHORTEN THE HARNESS 2 WHEN INSTALLED IN BA11 K EXPANSION BOX THE CABLE SHOULD TURNED 90 AND ROUTED UP OUT OF THE BOX AS SHOWN 7 3 THE ALTERNATE BRACKET MOUNTING POSITION MAY USED 5 1 WITH ANY EXPANSION BOX SIMPLIFY INSTALLATION WHEN THE ALTERNATE MOUNTING POSITION 15 USED THE BERG CONNECTOR IN THE TRANSITION BRACKET MUST BE INVERTED SO THAT THE CABLE FROM THE 6 OR RK07 DRIVE WILL HANG IN A DOWNWARD POSITION AS SHOWN lt gt RK611 CONTROLLER 11 4984 Figure 4 5 RK611 I O Connector Details 4 9 4 3 2 RK06 RKO7 Disk Drive To install the disk drive 1 Roll the drive cabinet to its designated location Level cabinet by lowering four leveling feet attached to the cabinet until all weight is removed from the casters Remove the front and rear access covers from the drive Verify that the model number listed on the serial tag Figure 4 6 corresponds to the site power requirements Refer to the voltage ranges given in Fig
64. RKBA register Figure 7 4 is initially loaded with the low order 16 bits of the Unibus address that will reflect the main memory start location for a data transfer With the low order bit 0 always forced to 0 the RKBA register content is combined with high order bits 8 and 9 of the RKCSI register BA16 17 to form a complete even numbered word address Following each data transfer bus cycle the register is incremented by two to select the next even numbered location 7 2 4 Disk Address Register RKDA For Read Write and Write Check commands the RKDA register Figure 7 5 is initially loaded to define the desired sector 1 of 20 or 22 and track 1 of 3 read write heads on the selected drive from or to which the first block of a data transfer will be initiated If the word count value indicates that a block of more than 256 data words is to be transferred the Sector Address bits of the RKDA register will be incremented to select the next consecutive sector and the next track if necessary until a word count overflow indicates that data transfers are completed or an error occurs In either case com pletion of the command is indicated by the setting of the Controller Ready RDY bit in RKCS1 and an increment of the RKDA register to the next sector location Sector Address Bits 0 4 SA0 SA4 Sector Address bits 0 4 are configured 00 23 to select a value 00 191 for a 20 sector format 18 bit data words or configured 00 25 to select a
65. RS SERVO TRACKS cYLO READ WRITE SURFACES SERVO SURFACE TIMING amp POSITION RKO6K DISK CARTRIDGE 3IDENTICAL READ WRITE 1SERVO SURFACES SURFACE 0 1 2 CYL 814 815 CYLINDERS 815 PRERECORDED SERVO TRACKS CYLO READ WRITE SURFACES SERVO SURFACE TIMING amp POSITION RK07K DISK CARTRIDGE ADDRESSING 1 CYLINDER ADDRESS DEFINES 1 OF 411 TRACKS FOR RKO6K OR 1 OF 815 TRACKS FOR RKO7K 2 TRACK ADDRESS R W HEAD DEFINES ONE R W SURFACE 3 SECTOR ADDRESS DEFINES 1 OF 20 OR 22 SECTORS TRACK MA 0922 Figure 2 2 Read Write Heads and Data Track Orientation 2 4 at SECTOR FORMAT 304 305 WORDS NUMBER OF WORDS 16 SECTOR TOLERANCE GAP 22sec TRACK 112 BITS AMBLE 20 96 BITS DATA WORDS 16 BIT WORDS 4096 BITS 18 WORDS 4608 BITS HEADER PREAMBLE DATA PREAMBLE Cylinder 000 6325 for the RK 6 or 000 14564 for the 2 Track head 0 1 2 3 Sector 00 23 0 00 25 4 Format t 18 bit 0 16 bit data EXCLUSIVE OR SUM OF WORDS 1 AND 2 5 Good sector both bits set for good sector 41 4354 Figure 2 3 Sector Formats PDP 11 MAIN MEMORY FBEH RK 6 OR CENTRAL DISK DISK DRIVE S PROCESSOR CONTROLLER Central Processor Disk Controller Disk Drive Delivers Commands 13 Interprets Commands Stores and Retrieves Data To Controller
66. RT JT 1 OUT J1 PORT IN J3 B IN INPUT OUTPUT CONNECTORS MA 0923 Figure 4 3 RK07 Disk Drive Logic Module Locations 4 5 42 SAFETY PRECAUTIONS When performing the installation procedures in Paragraph 4 3 observe the following precautions to avoid injury to personnel or damage to the equipment 1 Keep fingers and hands of area between cartridge and carriage while drive power is ON Always remove ac power cables when it is necessary to work inside the drive CAUTION Always remove wristwatch before servicing the drive linear motor electromagnetic fields are destructive to watches Use only nonmagnetic tools near the recording area of the disk cartridge Use care when working in the linear motor area with ferrous materials Keep read write heads away from the electromagnetic area Avoid touching or blowing breath on read write heads Skin acids can etch and ruin heads breath can cause condensation deposits that disfigure the gliding surface Never move the carriage assembly forward manually without a spinning disk in place or the head separator tool installed Never remove or change modules without shutting down all internal drive power 4 3 INSTALLATION PROCEDURE The procedures for installing the RK611 Controller and the RK06 or RK07 Disk Drive are covered in the following paragraphs 4 3 1 RK611 Controller To install the RK611 Controller in a PDP 11 expander box 1
67. T MESSAGE DESEL SEEK JRECAL START DRIVE DRIVE SELECT CODE MED voL RESVD LINE A RELSE COMM COMM SPINDLE CLEAR ORMAT or VAL CODE MESSAGE RESVD FOR ADD L MESS CYLINDER ADDRESS OFFSETS RESVD PARITY LINE B NOTEQ These bits are used only on the RKO7 a Controller to Drive Transmissions MESS REQ Figure 2 7 Controller Drive Serial 16 Bit Message Formats 2 9 CONTROL CLOCK STATUS MESS AO lt 2 A1 e A2 5 o ul z A3 DRIVE SELECT CODE DRIVE SERIAL NUMBER 0 WR DR RESVD FOR INV AC SEEK SPEED RD WR i FAULT PRTY K OFF a WR CN SEEK LIM Bt MESSIO RESVD FOR 5 amp NO GATE HEAD MULT INDEX TRIBIT SERVO 5 panrry J pup ADDL MESS ERR WRITE 8 FAULT SEL ERR ERR SIG ERRIMOTION SEEK UNSAFE 55 10 RESVD FOR B2 CYLINDER ADDRESS RESVDI PARITY 2 0 0 mess Z MESS ID RESVD FOR DECODED B3 1 1 ADD L MESS SECTOR COUNT HEAD ADDRESS RESVD PARITY NOTEO These bits are used only on the RK07 n 4198 b Drive to Controller Transmissions Figure 2 7 Controller Drive Serial 16 Bit Message Formats Controller to Drive Line When the CTD line is asserted the transmission of Message line A and B information from the controller to the drive is enabled The CTD si
68. TIONS RATES The RK611 has five registers that display the various error conditions that can exist in the Disk subsystem Table 5 1 lists the various error indications and gives an explanation of their significance Errors are broken down into five categories Control Cartridge Seek Hard errors Soft errors These classifications aid in identification of various errors encountered in acceptance testing only 5 1 1 Control Errors A control error indicates an operational aberration in the execution of command and control func tions Any control error detected by a diagnostic program during acceptance testing indicates a hard ware malfunction that must be corrected before testing can continue Table 5 1 RK611 RK06 and RK611 RK07 Disk Subsystem Diagnostic Errors Error Indicator Bit Error Type Programming Error RKCS2 bit 10 Register written except Control PGE CCLR SCLR with GO set Illegal Function RKER bit 0 Illegal command in low order Control ILF 5 bits of RKCSI Format Error RKER bit 4 or 5 Format or drive type error Control FPER Cylinder Overflow RKER bit 9 Cylinder address exceeded Control COE Invalid Disk RKER bit 10 Invalid cylinder or track Control Address IDAE address detected Unit Field Error RKCS2 bit 8 Drive selection problem Control UFE 5 1 5 1 Unibus Parity Error UPE Nonexistent Drive NED Drive Available Interrupt DAINT Drive
69. XC 115 VAC 50 HZ RKO6 XD OR RKO7 XD 230 VAC 50 HZ 11 4980 Figure 4 6 Serial Tag Location and Model Identifications Tag Visible with Front Cover Removed 4 11 DC POWER CABLE CNP1 CNP2 CNP3 CNP4 CNP5 CNP6 CABLE BOARD M7706 J2 A OUT AND TERMINATOR J4 B OUT J1 A IN AND J3 B IN 7897 14 Figure 4 7 RK06 RK07 I O Connector Details 4 12 CABLE CHART TO REMARKS CONNECTION J2A OUT NEXT DRIVE J4B OUT DRIVE NEXT DRIVE DRIVE TERMINATED SEE NOTE CABLE GROUND 25 FT VIEW C C POWER SUPPLY REF CABLE GROUND 3 FT LG CABLE DRIVE BUS 25 FT THE TOTAL LENGTH OF ALL DRIVE BUS CABLES ATTACHED TO AN RK611 CONTROLLER MUST NOT EXCEED 100 FEET CABLE GROUND 25 THIS GROUND STRAP IS NOT REQ D iF BOTH CONTROLL ERS ARE MOUNTED IN THE SAME EXPANDER BOX OR CAB FOR INSTALLATION OF 4 TRANSITION BRACKET A AND CABLE ORIENTATION RK611 CONTROLLER REFER TO FIGURE 4 4 EXPANDER BOX USED FOR MOUNTING 4 CONTROLLER 15 LOCATED SEPARATE CAB FROM DRIVE AND CAN BE POSITIONED AT ANY VERTICAL HEIGHT WITHIN THAT CAB 3 CHASSIS REF 7 47 IF THIS IS THE LAST DRIVE A SERIES A TERMINATOR DEC PART NO 70 12293 00 15 INSERTED IN THESE CONNECTIONS SEE CABLE CHART VIEW A A RKO6 OR RK07 DISK DRIVE WITH REAR SKIN REMOVED AND CARD NEST OPEN 11 4985 Figure 4 8 Dual Access Disk Subsystem H969 Cabinet 4 13
70. a d DU A 2 7 Interface 2 8 Controller Drive Serial 16 Bit Message 2 9 Figure No 3 1 3 2 4 1 4 2 4 3 4 4 4 5 4 6 4 7 4 8 4 9 4 10 4 11 4 12 4 13 6 1 6 2 6 3 6 4 6 5 6 6 6 7 6 8 6 9 7 1 7 2 7 3 7 4 7 5 7 6 71 7 8 7 9 7 10 7 11 7 13 7 14 7 15 7 16 7 11 7 18 7 19 7 20 FIGURES CONT Title Page Relationship of Disk Head Disk and 5 2 2 3 3 RK06 RKO0 7 Disk Drive Air Supply and Distribution Subsystem 3 4 RK06 RK07 Shipping Package 0 4 3 RK06 Disk Drive Logic Module Locations 4 4 4 RK07 Disk Drive Logic Module Locations a 4 5 RK611 Controller Backplane from Module Side 4 8 RK611T1I O Connector Details a 4 9 Serial Tag Location and Model Identifications Tag Visible with Eront Cover Removed EE OR CUYO 4 11 RK06 RK071 O Connector Details teer etit ent teet ano 4 12 Dual Access Disk Subsystem H969 2 200 4 13 Dual Access Disk Subsystem H969 00 4 14 RK06 Head Assembly
71. a recovery routine con sists of a sequence of 28 rereads during which 16 ECC cycles are attempted at track center and 12 ECC cycles are attempted in the six offset positions two attempts per offset position 4096 2048 1024 512 256 128 64 32 16 08 04 02 ECC POSITION REGISTER 2055 1 8 2 9 1 4 FIRST BIT OF ERROR BURST 15 BIT 2055 BIT BIT BIT BIT BIT 01 16 32 2055 2065 4049 4065 DATA STREAM 16 BIT WORD WORD ERROR WORD WORD WORD ASSUMED 1 2 BURST 254 256 ERROR BURST ONE BIT IN ERROR BURST CAUSES DATA FIELD BIT TO 0 0 0 2 4 1 1 COMPLEMENT ERROR BIT ERROR BIT ERROR BIT CONTENTS OF ECC CONTENTS OF 44 BIT FIELD IN DATA BLOCK AFTER CORRECTION NOTE Bits 0 3 and 8 or ECC pattern register are assumed in error 11 4229 Figure 7 30 Error Correction Process 7 35 7 5 1 Correctable Error DCK CERR Figure 7 31 provides an example of the ECC correction process that is initiated by the controller when a correctable data error is detected In the example an 11 bit error burst bits 20 30 is detected in word 2 bits 17 32 of a recorded data field The 11 bit error burst containing eight erroneous data bits 20 27 is delivered to main memory Under ECC control the RK611 loads the ECC Position RKECPS register with a value that is equal to the location of the first bit of the error burst 20 0 In addition the ECC Pattern RK ECPT register is loaded
72. a selected drive if the RUN STOP switch is in a RUN state This is accomplished by the setting of the Start Spindle bit in Message A T6 When Messages A and B are sent a return status is requested 02 SEL pou FORMAT PARITY FROM RKCS2 SPINDLE BIT BITS 0 1 ix on 01 MESSAGE B PARITY ODD 11 4221 Figure 7 23 Start Spindle Command Recalibrate 13 Figure 7 24 This command is used to relocate the heads to cylinder zero address of the outermost cylinder on the disk and to clear the drive s Cylinder Address register This is accom plished by setting the Recalibrate bit in Message A T5 When Messages A and B are sent a return status is requested 7 25 55 DR SEL CODE RECALIBRATE FORMAT PARITY FROM RKCS2 BIT BIT ODD BITS 2 0 1 0 1 MESSAGE Figure 7 24 Recalibrate Command Offset 15 Figures 7 25 and 7 26 This command directs the selected drive to position its heads a specified distance from the track centerline and to inhibit the recording logic The distance in micro inches and the direction toward or away from spindle that the heads are offset are determined by the value loaded into the Offset register RKAS OF This is accomplished by configuring the Cylinder Address Offset bits in Message B T4 T12 When Messages A and B are sent a return status is requested NOTE Notice in Figures 7 25 and 7 26 that message bits are a reversal of register bits MESSAG
73. ad only bit that is set Message 0 10 to indicate that the selected drive is in Offset mode Drive AC Low ACLO Bit 3 Drive AC Low is a read only bit that is set Message BO T6 to indicate that ac power in the selected drive is low or has been interrupted Following detection of an ac low condition the heads will unload when the next Sector pulse occurs This action allows the transfer of the current sector to be completed before the heads are unloaded If ac power is restored the error condition can be reset within the drive by a Unibus Initialize INIT Subsystem Clear SCLR power up or RUN switch reset Speed Loss SPLS Bit 4 Speed Loss is a read only bit that is set Message 0 12 to indicate that spindle speed is unacceptably low Drive Off Track DROT Bit 5 Drive Off Track is a read only bit that is set Message 0 13 to indicate that the Write Gate signal has been asserted while the heads are not properly positioned over the track centerline 7 10 Volume Valid VV Bit 6 Volume Valid is a read only bit that is set Message 0 6 to indicate that the Volume Valid flip flop has been set in the selected drive by a Pack Acknowledge command The set condition of the bit ensures the program that the cartridge and the unit number plug have not been changed since the last command was issued to the drive and power has not been removed The bit is reset when the cartridge the unit number plug or ac powe
74. address 310 has been detected with the receipt of a Seek command In addition to setting the IDAE bit either of these faults will set the SKI bit in the RKER register the CDA bit in the RKDS register while asserting an Attention signal for RKAS OF the DSC bit in RKMR2 and the Fault bit in RKMR3 2 The controller detected an illegal desired cylinder DCO DC9 value 41110 through 81410 in the RKDC register during the initiation of a command if the drive as defined by bit 10 of 51 is an RK06 These cylinder values are allowed only on the RKO7 3 controller detected an illegal track address TAO T A2 value 310 through 70 in the RKDA register during the initiation of a command Write Lock Error WLE Bit 11 Write Lock Error is a read only bit that is set Message BO T11 to indicate that the drive received the assertion of the Write Gate signal when it was in Write Protect mode The occurrence of this fault allows the CDA bit to be set in RKDS the Attention signal to be asserted for RKAS OF the DSC bit to be set in 2 and the Fault bit to be set in RKMR3 Drive Timing Error DTE Bit 12 Drive Timing Error is a read only bit that is set to indicate the following 1 The loss of Write Clock by the controller during a write 2 The loss of encoded read data by the controller during a read Operation Incomplete OPI Bit 13 Operation Incomplete is a read only bit that is set to indicate that following
75. al bit 15 in Toggle MCLK in conjunction with MERD to simulate the writing of the remaining 127 Header Preamble Zeros concluding with the generation of a single One bit to reflect the SYNC bit Simulate the reading of the three header words Simulate the Gap bits 64 zeros If header compare is correct the Write Gate WRTGT signal bit bit 14 in RKMR1 will be set and the remaining sector bits may be written as follows a 16 words of Data Preamble 255 zeros 1 One bit b 256 words of 16 or 18 bit data 4096 or 4608 bits c Two words of ECC 32 bits d One word of Postamble 16 bits If the number of data words to be written exceeds the length of one sector return to step 5 of the sequence and continue otherwise write simulation is complete 7 5 ERROR DETECTION AND CORRECTION When a write data command Write is executed a 32 bit Error Correction Code ECC is generated by the data and following the recording of the data field is written in the ECC field When a read data command Read Write Check is executed the ECC field is also read to verify the integrity of the recorded data If a read error is detected the read function of the current command is disabled and the ECC logic attempts to define both the position of the first error bit within the data field and the pattern of the error over an 11 bit range If the data error is confined to no more than an 11 bit error burst the error is correctable an
76. an unsafe condition and unloads the heads If a controller has seized a drive and multiple drive select is detected on the other port the drive detects the condition but does not set fault and unsafe and does not unload the heads The port detecting multiple drive select reports it to the controller regardless of whether it had access to the drive Controller to drive CTD parity errors are detected and stored separately for each port Unlike all other errors the CTD parity error flip flop is cleared with a Drive Clear regardless of whether or not that controller has seized the drive The Attention in the port is raised when a CTD parity error occurs 9 7 INITIALIZE IN DUAL ACCESS The RK06 or 7 will honor an initialize issued separately from either port The effect the in itialize has depends upon the availability of the drive to the port issuing the command The three drive conditions are described 1 The drive is seized by a port At this time an initialize will a Clear that port s attention bit b Clear the error flip flops in the drive c Cause the port to release the drive If the other port s port request flip flop is cleared the drive will be returned to neutral If the other port s port request flip flop is set the drive will become seized by the other port 2 The drive is in neutral At this time an initialize will Clear the port s attention bit b Clear the error flip flops in the drive 3
77. and RK611 RK07 Disk Drive Subsystems capa bilities 2 2 06 AND RK07 DISK DRIVES The RK06 and RKO7 are moving head disk drives that function as random access mass storage devices Figure 2 1 The drives perform under the direction of the RK611 Disk Controller and contain all of the response circuitry required to perform those operations for device status reporting and the storage of data The storage medium designed for the RK06 Drive is a dual platter assembly defined as the RKO6K Disk Cartridge which has a maximum formatted storage capacity of 6 3 million 18 bit words or 6 9 million 16 bit words Similarly the RK07 Drive uses a dual platter assembly the RK07K Disk Car tridge as the storage medium It has a maximum formatted storage capacity of 12 6 million 18 bit words 13 8 million 18 bit words As stated before R K06K cartridge may only be used in RK06 Drive and an RKO7K cartridge will only operate in an RKO7 Drive RK06 and RK07 Controls and Indicators The following manual controls and indicators are located on the front of each RK06 and RK07 Disk Drive RUN STOP pushbutton switch with STOP indicator UNIT SELECT plug with READY indicator Drive FAULT indicator WRITE PROT pushbutton switch with PROTECT indicator ACCESS A pushbutton switch with SELECT indicator ACCESS B pushbutton switch with SELECT indicator The switches are selectively used refer to Chapter 6 to place an associated drive on or off
78. ant in all facets of a computer system it is particularly crucial in the case of a device such as the RK06 or RK07 Disk Subsystem Disk cartridges are not sealed units while loading and are extremely vulnerable to dirt Even such minute obstructions as smoke particles fingerprint smudges or dust specks can cause head crashes and catastrophic destruction of heads and or disk surfaces as shown in Figure 3 1 The RK06 7 is capable of operating in an ambient atmosphere containing not more than one million particles of 0 5 micron or larger per cubic foot of air During site preparation there are a number of steps that may be taken to enhance subsequent clean liness 1 Seal all windows in the vicinity of the Disk Subsystem 2 partitions are to be installed consider floor to ceiling walls which minimize the flow of dust 3 If painting is to be done in advance of installation select paint for walls ceilings and floors that will not tend to flake or powder excessively Waterbound distemper is generally unsat isfactory in this respect 4 Check that the flow of air from the air conditioning system will tend to carry lint etc away from the Disk subsystem location Provide filtration to inhibit dust and other particulants 5 Select acoustical material that will neither produce nor harbor dust 6 Avoid glass fiber tiles that could produce abrasive particles and floor coverings that tend to crack or crumble 3 2
79. ardless of whether they are stationary or in motion the servo signals provide the generation of Sector and Index once per revolution pulses within the drive The Sec tor Index pulses define the position of each sector for the controller providing rotational coordinates for read and write operations Finally within the controller the clock derived from the servo signals is used to synchronize the operation of a phased locked oscillator PLO Synchronization of the PLO circuitry ensures a con stant recording density regardless of minor variations in the rotational speed of the disks 2 5 5 Data Error Control The RK611 Controller contains both data error detection and correction logic and a data error recov ery capability When properly supported by appropriate software procedures these controller features minimize the disruptive effect of data errors from the disks Data Error Detection The following defines the under which data errors can be detected by the controller e Data bit errors indicated by the Error Correction Code ECC logic The controller has detected a mismatch resulting from a comparison of header data For data error correction purposes each sector contains a 32 bit Error Correction Code ECC The code is used with a combination of hardware and software procedures to provide the following Detection and correction of Data field error bursts up to 11 bits in length Detection without correction o
80. at Read Gate has been asserted by the controller 7 18 7 2 12 ECC Position Register RKECPS The Error Correction Code ECC Position register Figure 7 13 is a read only register that is used store the error position value which results from the successful execution of an ECC correction se quence When an ECC correction sequence is not in progress the RKECPS register contains one of two possible values 004066 0050663 If the CFMT bit bit 12 in RKCSI is reset reflecting a 22 sector format 16 bit data words the lesser value is contained in the register However if the CFMT bit is set reflecting a 20 sector format 18 bit data words the greater value is contained in the register 13 12 1 10 09 08 07 06 05 04 03 02 01 15 14 00 ECPS 2 12 10 09 08 07 06 05 04 03 02 01 00 X READ ONLY RO 11 4211 Figure 7 13 ECC Position Register Error Position EP00 EP12 Bit 0 12 The Error Position bits are read only bits that define the start bit location of an error burst containing from 1 to 11 error bits within a 256 word Data field follow ing the completion of an ECC correction sequence The position is valid if the error is ECC correc table If the detected error is not correctable using ECC the register contains a constant 010040 Spares Bits 13 15 Spare bits 1
81. bit is internally set CERR by an error that is followed by an external set CCLR to initialize the controller bit 15 will be cleared However since only controller errors will be initialized by CCLR any error originating in a drive will remain set in the drive CAUTION When using BIC instruction on the RKCSI regis ter ensure that a 1 is set in bit 15 of the mask If this is not done and CERR is set a CCLR will occur and the RK611 will be cleared For example to clear the interrupt Enable IE bit bit 6 in RKCS1 the fol lowing instruction format is recommended BIC 100100 RKCS1 Table 7 1 Combined Error CERR Error Programming Error PGE Illegal Function ILF Format Error FTER Cylinder Overflow COE Invalid Disk Address IDAE Unit Field Error UFE UNIBUS Parity Error UPE Nonexistent Drive NED Drive Available Interrupt DAINT Drive Error DRERR Bad Sector Error BSERR Header VRC Error HVRC Indicator Bit RKCS2 bit 10 RKER bit 0 RKER bit 4 or 5 RKER bit 9 RKER bit 10 RKCS2 bit 8 RKCS2 bit 13 RKCS2 bit 12 RKDS bit 0 Message 0 7 FAULT RKER bit 7 RKER bit 8 7 5 Condition Register written except CCLR SCLR with GO set Illegal command in low order 5 bits of RKCSI Format or Drive Type error Cylinder address exceeded Invalid cylinder or track address de tected Drive selection problem Error in 16 bit data word from main memory
82. c Programs Used in RK611 RK06 RK611 RK07 Disk Subsystem Field Acceptance Tests Program MAINDEC 11 DZR6A RK611 Diskless Controller Diagnostic Part 1 MAINDEC 11 DZR6B RK611 Diskless Controller Part 2 MAINDEC 11 DZR6C RK611 Diskless Controller Diagnostic Part 3 MAINDEC 11 DZR6D RK611 Diskless Controller Diagnostic Part 4 Function Reads and writes every RK611 register Tests the interrupt mechanism Tests the SILO loading logic Note No drive is required for the execution of this program Tests the loading of drive bus messages by execu ting Class A commands Some Part tests execute commands that are partially maintenance mode and partially at normal speed in order to fool the controller and force errors no drive is required Tests the loading of the drive bus message shift reg ister for Class B commands Tests index and sector pulse detection Tests SILO and NPR transfers from memory in 16 and 18 bit mode Tests nonexistent memory and Unibus parity error detection Tests Read and Write MFM loopback Tests Class B instruction errors no drive is re quired Tests loading of drive bus message shift registers for Class C commands Tests header generation for search operations Tests write data NPR transfers to SILO Tests header recognition Tests cylinder track and sector increment after successful header search Tests detection of all header type errors Tests ECC g
83. capacity formatted Cylinders cartridge Tracks cylinder Tracks cartridge Sectors track Words sector Bits word Bits sector Bits track Bits surface Bits pack Bits inch inner track Tracks inch Electrical Voltage Input Power Single access drives Dual access drives Start current Single access drives Dual access drives Power factor Specifications Dual platter magnetic cartridge DEC RK06K DC 355 mm 14 in nominal 3 read write 1 servo 18 bit Word 16 bit word 411 411 3 3 1233 1233 20 22 256 256 18 16 4608 4096 92 160 90 112 37 88 37 04 113 63 111 11M 4040 4040 192 3 192 3 See model designations in Section 1 5 500 W maximum at 60 Hz 450 W nominal at 60 Hz 550 W maximum at 50 Hz 500 W nominal at 50 Hz 550 W maximum at 60 Hz 500 W nominal at 60 Hz 600 W maximum at 50 Hz 550 W nominal at 50 Hz 10 5 Arms at 115 V 60 Hz 5 3 Arms at 230 V 60 Hz 11 0 A rms at 115 V 50 Hz 5 5 A rms at 230 V 50 Hz 11 3 Arms at 115 V 60 Hz 6 1 Arms at 230 V 60 Hz 11 8 Arms at 115 50 Hz 6 3 Arms at 230 V 50 Hz 0 80 minimum Table 1 2 RK06 Disk Drive Performance Specifications Cont Characteristics Specifications AC low detection Less than 90 V rms for one cycle or more Power cord Length 2 7 m 9 ft Plug type NEMA 5 15P for 120 Vac nominal models NEMA 6 15P for 240 Vac nominal models Operating environment Ambient temperature 10 C to 40 C 50 F to 104
84. cteristics Storage Type Medium Disk diameter Magnetic heads Recording capacity formatted Cylinders cartridge Tracks cylinder Tracks cartridge Sectors track Words sector Bits word Bits sector Bits surface Bits pack Bits inch inner track Tracks inch Electrical Voltage Input power Single access drives Specifications Dual platter magnetic cartridge DEC RK07K DC 355 mm 14 in nominal 3 read write 1 servo 18 bit word 16 bit word 815 815 3 3 2445 2445 20 22 256 256 18 16 4608 4096 73 43M 75 11 M 220 32M 225 33M 4040 4040 384 6 384 6 See model designations in Section 1 5 500 W maximum at 60 Hz 450 W nominal at 60 Hz 550 W maximum at 50 Hz 500 W nominal at 50 Hz 1 8 Table 1 3 RK07 Disk Drive Performance Specificatons Cont Characteristics Dual access drives Start current Single access drives Dual access drives Power factor AC low detection Power cord Length Plug type Operating environment Ambient temperature Maximum temperature rate of change Relative humidity Maximum altitude Bit transfer rate unbuffered nominal Bit cell width Latency Rotational frequency Average Maximum Specifications 550 W maximum at 60 Hz 500 W nominal at 60 Hz 600 W maximum at 50 Hz 550 W nominal at 50 Hz 10 5 Armsat 115 V 60 Hz 5 3 Arms at 230 V 60 Hz 11 0A rmsat 115 V 50 Hz 5 5 rms at 230 V 50 Hz 11 3A rmsat 115 V 60 Hz 6 1 Armsat 230 V 60 Hz 11 8A rmsat
85. d routine will transfer the content of the Maintenance registers RKMR2 into a previously defined buffer its address is in RO Maintenance Register Read R2 Address of RKCS1 CLR R3 1965 4100000 R2 Clear controller MOV UNIT RKCS2 R2 Reselect unit MOV R3 RKMR Select drive serial message 1 R2 Issue select drive function TSTB R2 Wait for the serial message BPL 2 MOV 2 2 Save content of MR2 MOV RKMR3 R2 RO Save content of MR3 INC R3 Increment drive serial message count CMP R3 44 Done four messages yet BNE 196 loop RETURN 7 7 PROGRAMMING CONSIDERATIONS The following provides a selection of programming information for the reader s consideration Flagging a Bad Sector Since a Write Header command cannot be used to selectively write individual sectors an entire track is written when the command is executed Therefore if a programmer desires to flag a bad sector all currently recorded sector and data information from the desired track must be extracted and preserved elsewhere for rewriting along with the new information Unit Selection When the Controller Clear CCLR bit bit 15 in RKCSI is asserted via the Unibus the Drive Select 080 082 bits bits 0 2 in RKCS2 are cleared Therefore before the next command is initiated unit selection must be reconfigured 7 43 Start Spindle Command If a Start Spindle command is initia
86. d the following indicators are set l 2 3 Data Check error bit bit 15 in Controller Error CERR bit bit 15 in RKCSI Ready RDY bit bit 7 in RKCS1 However if the detected data error exceeds 11 bits the error cannot be corrected and the following indicators are set 1 2 6 Data Check error 15 Controller Error CERR bit 15 RKCSI Ready RDY bit 7 in RKCSI 7 34 Thus within less than one complete disk revolution program can determine if the error is table DCK CERR or uncorrectable ECH DCK CERR via a check of the indicator bits If the error is correctable the ECC Position RK ECPS register can be examined to determine the location of the first error bit within the error field while the ECC Pattern RKECPT register can be examined to determine which of the 11 error burst bits must be corrected in main memory In this regard correction results from the exclusive ORing of the 11 bits contained in the Pattern register with the 11 bits defined by the error burst Figure 7 30 However if the error is detected as uncorrectable the program can reread the error sector during subsequent disk revolutions utilizing both ECC and head offset operations if necessary to improve the chances for an eventual correction and or recovery Thus a standard dat
87. der Handling Read Data 1 0 0 0 1 21 Write Data 1 0 0 1 1 23 Read Header 1 0 1 0 1 25 Write Header 1 0 1 1 1 27 Write Check 1 1 0 0 1 31 Select Drive 01 Figure 7 19 This command is basically used to select a drive and obtain the return of status information This is accomplished by the setting of the Drive Select bits of Message A 0 2 However the command can also function as a release control to allow one of the dual ports of the selected drive to be freed for access by a second controller The latter is accomplished by the setting of the Release bit of Message A T3 When Messages A and B are sent return status messages AO and BO are requested However Select Drive is the only command with the ability to request MS1 MSO any one of the three pairs of additional message types A1 A3 and 1 3 Pack Acknowledge 03 Figure 7 20 This command is used to set the Volume Valid flip flop in a selected drive to indicate that the software has acknowledged the medium This is accomplished by the setting of the Volume Valid bit in Message A T11 When Messages A and B are sent a return status is requested Drive Clear 05 Figure 7 21 This command is used to clear all error flags in the selected drive provided the error s are no longer present In addition the command resets the Status Change flip flop in the drive These actions are accomplished by the setting of the Drive Clear bit in Message A T8 When Messages A and B a
88. e the operational status of a selected drive However information obtained from the drive is not immediately available to program control until the information is validated SVAL by the setting of bit 5 which indicates that a com plete status message has been received providing a valid update Status information bits set in the RKDS register can be cleared by conventional initialization INIT CCLR SCLR However a Controller Clear CCLR operation does not affect status or error condi tion bits that are currently set in the drives In addition a Unibus Initialize INIT or Subsystem Clear SCLR operation can only reset status or error bits in a drive if the associated status or error condi tion no longer exists 11 10 09 06 03 READ ONLY 11 4205 Figure 7 7 Drive Status Register Drive Available DRA Bit 0 Drive Available is a read only bit that is conditioned message 5 to dual access operation as follows When set by a selected drive DRA indicates that the drive is available to accept commands from the controller e If an attempt is made to select a dual access drive that is currently accessed by the other controller the DRA bit will not be set This is considered an error condition by the controller resulting in the setting of controller Combined Error CERR bit in RKCSI and the generation of a program interrupt if enabled Spare Bit 1 Spare bit 1 is 0 Offset OFST Bit 2 Offset is a re
89. e the simulated transfers by unloading the Silo The 12 steps just described constitute a basic read simulation sequence However variations can be used to provide diagnostic information for error condition analysis or verification of the Error Correc tion Code ECC logic and its operation 7 4 2 Write Simulation The following sequence of operations simulates the writing of a complete sector in Diagnostic mode 1 With the exception of RKCS1 load device registers required to support a normal Write command Set the Diagnostic mode DMD bit bit 5 in 1 Load Command 1 4 bits bits 1 4 and set GO bit bit 0 in RKCSI Toggle the Maintenance Clock MCLK bit bit 8 in to produce the number of data clock cycles required to generate an NPR for each data word reflected by the word count in For example a WC value of 6410 will fill the Silo Therefore 64 NPR cycles must be generated to complete the operation 1 37 data clock cycles times 64 for 16 bit data 40 data cycles times 64 for 18 bit data if the word count is greater than 6410 Toggle 1 0 the Maintenance Sector Pulse MSP bit bit 6 in to simulate the generation of a sector pulse Toggle MCLK in conjunction with the Read Data bit bit 9 in to simu late the writing of 128 of the 255 Header Preamble Zeros 7 33 10 11 12 Check for setting of Read Gate sign
90. ed Attention Line An Attention signal is activated within a drive when its Drive Status Change flip flop is set However the Attention line is not asserted to the controller until the unit number number on the drive UNIT SELECT plug of the drive is polled via the Poll Drive 22 2 lines Thus each time a drive is polled and its Drive Status Change flip flop is set Attention will be reported The Attention signal is available to the controller approximately 400 ns after the detection of a polling address for the drive and the signal is negated within approximately 160 ns after the polling address is removed Write Clock Line This drive signal is derived from the servo surface and synchronized with the rotational speed of the disk to ensure the recording of a constant bit density on the data track Read Write Data Line This line carries encoded digital read data from the drive to the controller and encoded digital write data from the controller to the drive Write Gate Line When active for a write operation this signal enables the generation of write current in the selected drive The drive recognizes the assertion of this signal under the following conditions Drive selection achieved Volume valid set bit T6 of Status Message 0 No unsafe read write condition exists in drive No drive off track error exists No write protect mode exists Index Sector Pulse Line When a drive is selected this line transmits the once pe
91. ely by the RK611 A Seek to the cylinder in RKDC is performed When the Attention signal from the selected drive becomes true a drive status is requested When the status is obtained a Clear command is executed to clear Attention signal Transfer of data from memory to the Silo is begun and headers are read and compared with the desired disk address until the correct sector is found Preamble Data 256 words and ECC bits 32 are written on the disk If the word count in RKWC goes to zero during the sector the rest of the sector is zero filled After the sector transfer the word count in RK WC is checked and if nonzero the data transfer operation is continued into the next sector The word count in RK WC is checked at the end of each sector and when it equals zero the command is terminated by setting the Ready bit Figure 7 28 Read Header 25 The following sequence is executed A Seek to the cylinder in RKDC is performed When the Attention signal from the selected drive becomes true a drive status is requested When the status is obtained a Clear command is executed to clear the Attention signal The three words of the first header encountered are read and placed in the Silo The Ready bit is set The three header words may now be retrieved by examining RK DB three times Figure 7 28 7 28 Write Header 27 The following sequence is executed A Seek to the cylinder in is performed When the Attention signal from the
92. eneration and writing Tests partial sector write zero fill Tests 18 bit format ECC generation and data write no drive is required 5 4 Table 5 2 Diagnostic Programs Used RK611 RK06 RK611 or RK611 RK07 Disk Subsystem Field Acceptance Tests Cont Program MAINDEC 11 DZR6E RK611 Diskless Controller Diagnostic Part 5 MAINDEC 11 DZR6G Unibus RK06 RK07 Dual Port Drive Diagnostic available Q4 FY78 MAINDEC 11 DZR6H Unibus RK06 RK07 Drive Diagnostic Part 1 MAINDEC 11 DZR6I Unibus RK06 RK07 Drive Diagnostic Part 2 Function Tests multisector data transfers Tests midtransfer seeks Tests cylinder overflow checking Tests NPR transfers to memory Tests ECC error detection and correction in 16 and 18 bit modes Tests write check in both 16 and 18 bit modes and forces Tests write check errors no drive is required Ensures that port request port seize and timeout function correctly Ensures that the release function operates cor rectly Ensures that the timer inhibiting logic operates correctly Ensures that data transfers function correctly through both parts Ensures that the drive can perform all static and cycle up tests Ensures that the drive can write and read headers in both 20 and 22 sector formats Ensures that the disk can perform seek operations by doing several seek patterns Checks error detection logic by software error forcing Ensures that the disk is ca
93. er 2 RK 2 Maintenance Register 3 NOTE R W Basic Purpose Decode commands controller status Number of data words for transfer Main memory location for data word Desired track sector address Additional control information Drive status reports Additional drive status Drive status change state head offset value Cylinder address Unused Data word to from Silo Drive message select control Error position information follow ing correction sequence Error correction pattern following correction sequence Shift register A for serial Message A Shift register B for serial Message B read write selected bits in relation to Unibus R read only in relation to Unibus RK611 responds to this address with SSYN However if the bus cycle is a DATI D 00 15 is undefined 2 6 2 4 4 Silo Memory During direct memory access DMA operations in which a device bypasses the central processor and gains a direct access to main memory the storage depth of the Silo 66 words compensates for the transfer rate differences that occur between controller drive data exchanges and controller Unibus data exchanges When writing the Silo receives parallel data from Unibus and transfers the data to Data Shift Register for serialization The serial data is then encoded using the modified frequency modulation MFM technique When reading the MFM encoded data is decoded and sen
94. es 1 12 RK611 Unibus Addressable 2 6 Special Tools and Test Equipment for RK611 RK06 RK611 RKO07 Disk Subsystem Installation a erbe evel o ede 4 1 RK611 Control Switch Configurations for Desired Unibus Address and Interrupt a iai tendat tutae PH PNE DE LEAD AS 4 7 Drive Bus Caples PCT cae dee ato 4 10 Sample Table for Backplane 4 10 Unit Select Plugs for RK06 RK07 Disk Drives a 4 15 RK611 RK06 and RK611 RK07 Disk Subsystem Diagnostic Errors 5 1 Diagnostic Programs Used in RK611 RK06 or RK611 RKO7 Disk Subsystem Field Acceptance Tests 5 4 Combined Error na a n aaah 7 5 Disk Commands edes prd 7 23 1 INTRODUCTION 1 1 PURPOSE AND SCOPE This manual provides information related to capabilities installation operation programming and user maintenance requirements of the RK611 RK06 and RK611 RKO7 Subsystems This manual is intended primarily for operating and programming personnel Installation and service should be performed only by qualified DIGITAL field engineering and maintenance personnel The information is arranged as follows Chapter provides a general description of the subsystems including capabili
95. eturned from the drive when requested by the controller in a similar manner Figure 2 7b Message B Line In conjunction with the transmission of the Message A line from controller to drive a 16 bit message is simultaneously transmitted over the Message B line The Message B line transmits additional oper ational information e g cylinder address offset values etc including a status message selection code When the operational information has been assimilated by the selected drive the selected status message if requested by the controller is returned Figure 2 7b Control Clock Line As long as power is available to the controller the 465 ns Control Clock signal is continuously gener ated Strobe Line For message synchronization separate strobe signals are generated in a manner which depends on the direction of the transmission e Controller to Drive Strobe During the transmission of Message lines A and B a single strobe signal is generated as an enable for drive selection code identification Figure 2 7 Drive to Controller Strobe During Message line A and B status reporting separate strobe pulses generated by the drive clock each message bit into the controller Figure 2 7b JaesNs CONTROL CLOCK To te Ts Te Tz Te o ys ies 5 _ CTO 20 SET SE
96. f error bursts from 12 to 32 bits in length Reliable but not absolute detection of error bursts greater than 32 bits Data Error Correction Error correction involves the use of error position error burst location and error pattern error burst image information which is derived from a 32 bit Error Correction Code ECC configuration Each sector contains an ECC that is written at the end of the Data field and reflects the configuration of the recorded data The code is used to detect and correct if possible the data bits defined by an error burst pattern read from the disk The controller contains the logic required to generate the ECC configuration during a write and extract the code during a read However the actual correction of the data bits defined by the error burst pattern is a software process that occurs after the data is in main memory Data Error Recovery Data error recovery is a recommended sequence of attempted rereads that implements under program control both the head offset and error correction capabilities of the controller to improve the proba bility of a successful recovery A complete error recovery routine consists of a maximum of 28 data recovery attempts using error correction procedures with the following operations 1 Using error correction attempt to reread the sector 16 times at track centerline 2 Using error correction attempt to reread the sector twice from each of the following head offset positi
97. ft modified the RK06K RKO7K Cartridge Formatter Program DZR6L must be run using Mode 1 before any further testing can be performed At this point in the testing the RK611 is functionally checked out The only RK611 test not performed at this time is the Data Reliability Test 5 2 22 RK06 or RK07 Disk Drive 5 2 2 1 General This section of the diagnostic procedure covers the testing of the RK06 RKO7 Drive With a high degree of confidence any failures detected in these tests can be attributed to the Drive only All programs listed in Paragraph 5 2 2 2 must run through the specified number of error free passes before performing the remainder of the program Any failure detected by any of the diag nostics in this section must be corrected and all programs rerun before proceeding NOTE In the following steps repair or replace any failing component rerun the program that detected the fail ure to verify the correction then return to step 1 Load diagnostic DZR6H run two passes with no errors Load DZR6I run two passes with no errors Load DZR6J run one pass with no errors Load DZR6M run one pass with no parameters out of the specified limits printed by the Load DZR6N run two passes with no errors Load DZR6G run two passes with no errors 5 8 5 2 3 Cartridge Interchangeability Test This test requires the use of at least two drives If there is only one drive on the system being tested it will be necessary t
98. further commands BNE 4 If NE no BIT ZUNS RKER R2 Is this a Seek function BNE 4 NE yes MOV CYLADR RKDC R2 Load cylinder address CMP 113 FUNCTN Is this a recalibrate function BEQ 2 If EQ yes 117 Is this a Seek function BEQ 2 yes CMP 115 FUNCTN Is this an Offset function BNE 3 no MOV OFFSET RKOF R2 Load offset value 25 POSPRO Indicate positioning command 3 ADD 2 SP Show a good return INCB RK6ACT Set RK06 active MOV FUNCTN R2 RKCS1 4 RTS PC Return to users program sSINTERRUPT SERVICE ROUTINE INTR MOV 4RKCS1 R2 TSTB BEQ 5 BIT CERR DI R2 BNE 55 Point to CSI register Positioning in progress no Positioning complete or error NE yes Wait for positioning to complete 7 40 5 10 205 ix CLRB TST BMI TST BMI MOV TSTB BPL TST BPL CERIIE R2 Clear IE without doing CCLR POSPRO Reset positioning in progress R2 Any errors 205 Yes process directly RKDS R2 Is current status valid 20 Yes it is SELDRV R2 Select drive to get fresh status R 15 for the serial message 10 2 Any errors 40 If PL user now handles any error conditions 405 CLRB RTI RK6ACT Set RK06 inactive Return to user 7 6 2 Correction Routine The ECC Correct
99. fy the need you think it was intended to satisfy Does it satisfy yourneeds ___ Please send me the current copy of the Technical Documentation Catalog which contains information on the remainder of DIGITAL s technical documentation Name F Street T a Se gt Company PETERE State Country Department Additional copies of this document are available from Digital Equipment Corporation 444 Whitney Street Northboro Ma 01532 Attention Communications Services NR2 M15 Customer Services Section Order No EK RK067 UG 001 FIRST CLASS PERMIT NO 33 MAYNARD MASS BUSINESS REPLY MAIL NO POSTAGE STAMP NECESSARY IF MAILED IN THE UNITED STATES Postage will be paid by Digital Equipment Corporation Technical Documentation Department Maynard Massachusetts 01754 digital equipment corporation
100. ge B with each bit position 0 15 directly related to an equivalent transmission time 15 For additional informa tion refer to Paragraph 2 5 1 PARITY HEAD SELECT SET MEDIUM DRIVE SELECT OFF LINE COMM CODE Reserved COPE SET 2 COMM RELEASE VOLUME SEC RETURN TO RECAL VALID FORMAT CENTERLINE COMM 11 4213 Figure 7 15 Maintenance Register 2 Controller to Drive Drive to Controller Message 0 For drive to controller transmissions Figure 7 16 the 16 bit Message A shift register assembles status information along with an appropriate odd parity bit from various Status and error registers within the drive As the bits are serially transmitted from the drive i in coincidence with Message B status bits they are assembled in the register with each bit position relat ing directly to an equivalent transmission time Since only one of four status messages be stored the Message shift register at a time a desired message 1 3 other than AO must be previously defined by the execution of a Select Drive com mand 7 2 15 Maintenance Register 3 RKMR3 RKMR3 is a read only register that is used for Message B drive to controller transmissions addi tional command parameters as well as Message BO B3 drive to controller transmissions status Controller to Drive Message B For controller to drive transmissions Figure 7 17 the 16 bit Mes sage B shift register assembles additional di
101. gnal is normally asserted remaining so for the entire 16 clock times Figure 2 7 When the CTD signal is negated the selected drive is allowed to transmit status information to the controller Figure 2 7b Select Acknowledge SACK Line When asserted the SACK line provides a drive to controller indication that the drive select code has been recognized and the remaining bits of the message are being accepted This occurs approximately 300 ns after the receipt of bit 5 T4 of the controller to drive message The line is negated within 100 ns of the receipt of another controller to drive message strobe 2 10 Poll Drive 2 2 Lines These three binary encoded drive address lines are used to poll all drives 0 7 available to the system to determine the attention status of each drive Regardless of whether a drive is selected or not the polling mechanism permits the controller to determine which drive if any has asserted any attention signal without serially selecting each drive The time required to update the attention status of all eight drives is 7 4 us The Attention signal in a drive can be set by any one of the following conditions The completion of a Start Spindle command The completion of a Seek command The completion of an Offset command The beginning of an Unload Heads command The completion of an RTC command return offset heads to centerline The occurrence of any fault condition Poll
102. h header information via a Write Header command The result of this operation effectively divides a track into a number of unique data storage segments that are defined as sectors Figure 2 3 For 16 bit data words a single track can accommodate 22 sectors while for 18 bit data words a track can accommodate 20 sectors In either case sectors are the basic addressable storage unit within the drives and as such are the final coordinate of track selection during data transfer operations 2 4 RK611 DISK DRIVE CONTROLLER RK611 Disk Controller presents a complete control and data interface to the PDP 11 Unibus which serves both the central processor and main memory and the RK06 or RK07 Disk Drives available to the system Figure 2 4 2 4 4 Command Control The RK611 Control interface contains all of the logic required to receive store and decode the 13 separate disk commands e g Start Spindle Seek Write etc that can be executed by the controller In addition the control interface contains all of the logic required to receive and store the status information i e modes conditions faults and errors delivered from a selected drive To accom modate the exchange of information commands status between the control interface and a selected drive information is both sent commands and received status via a pair of bidirectional serial message lines that are defined as Message Line A and Message Line B 2 4 2 Data Control The
103. h of the eight drives in the daisy chain configuration Table 4 5 Unit Select Plugs for RK06 RK07 Disk Drives Part Number 0 12 12691 0 1 12 12691 1 2 12 12691 2 2 12 12691 3 4 12 12691 4 5 12 12691 5 6 12 12691 6 7 12 12691 7 4 4 INSTALLATION CHECKS AND ADJUSTMENTS To verify performance and adjust the disk subsystem 1 Check controller power supply voltage 2 Check drive power supply voltage 3 Check head load unload 4 Check head alignment 5 Check response to abnormal conditions The required actions are detailed in Paragraphs 4 4 1 through 4 4 5 4 15 4 4 1 Controller Power Supply Voltage Checks To perform the required voltage checks on the controller power supply set controller power to ON and make measurements at the following points Test Point Voltage 04 2 5 Vdc 0 25 04 2 15 0 75 V 04 2 GND A04 D2 15 Vdc 0 75 If any voltage is out of tolerance refer to the appropriate expander box manual for adjustment and or repair procedures 4 4 2 Disk Drive Power Supply Voltage Checks To perform the required voltage checks on the disk drive power supply 1 Ensure that power cable is connected to a power outlet of proper voltage and fre quency rating Ensure that the RUN STOP switch is in the STOP position Set circuit breaker to the ON position The air system should start operating and the lid lock solenoid will energize to enable the
104. he RK06 Disk Drive 411 cylinders is 75 ms while the minimum track to track for one cylinder is 8 ms with an average access time of 38 ms The maximum seek time in the RKO7 Drive 815 cylinders is 71 ms and the minimum seek time is 6 5 ms The average access time is 36 5 ms A data transfer rate of 4 3 per 16 bit word between the controller and the Unibus provides average transfer rate of 232 500 words per second 1 2 2 Diagnostic and Error Detection Correction Capabilities Diagnostic capabilities are enhanced by an extensive status error reporting network consisting of nine Unibus addressable registers which monitor subsystem activities and are visible to the software For extended maintenance and serviceability the RK611 RK06 and RK611 RKO7 subsystems pro vide a software controlled diagnostic mode DMD of operation When diagnostic mode is initiated the drive interface lines are effectively disconnected from the Controller while the Unibus interface remains intact With this arrangement read or write commands can be initiated using simulated data and control techniques while actual transmissions between the drives and the controller are disabled In this manner closed loop data transfers to and from main storage can verify controller operation in relation to data control The RK611 Controller provides data error detection and correction logic which utilizes an error correction code ECC technique to detect and identify fo
105. he data format of Zero bits which ensures a differentiation between sectors that is required by the mechanical tolerances 2 5 4 Servo Surface Operations For the drive to affect head positioning and the reading and writing of data under the direction of the controller radial cylinder and rotational sector index coordinates are derived from factory recorded servo track signals tribits that are continuously read by the servo head When a Seek command is initiated a desired cylinder address is delivered to the drive from the con troller Within the drive the difference between the desired and the current cylinder position is calcu lated If the difference is other than zero the head positioning circuitry is energized and all of the heads R W and servo are driven across the radius of the disk in the direction defined by the differ ence information As the search for the desired cylinder position continues increasing proximity to the cylinder gradually reduces the difference to zero at which time the heads are appropriately positioned over the desired track As stated the servo tracks control this process by generating pulses that are derived from each servo track crossing and fed back to the head positioning circuits as the disk sur faces are radially traversed When the number of pulses generated equals the number expected differ ence minus track crossing equals zero the desired cylinder is located Once the heads are loaded reg
106. he multiplexers for operating the drive through the two ports 9 3 2 Timing Counters and Decoders These circuits control the timing of transmissions from the controller that has seized the drive 9 3 3 Auto Release Timer This timing circuit generates a signal that will automatically release a port if it has not been utilized within a one second period after it has gained access 9 3 4 Multiplexers These circuits act as a single pole double throw switch for each signal used by a port for operation of the remainder of the drive 9 4 DUAL PORT OPERATION 9 4 1 Seizing a Drive in the Neutral State When an RK06 or RKO7 Drive is in the neutral state it is equally available to each port Neither port has priority over the other When a controller selects a drive the drive will switch to the correct port to connect the drive with that controller If the two controllers simultaneously attempt to seize the drive a high speed clock in the dual port logic in the drive will arbitrate allowing only one of the controllers to seize the drive 9 4 2 Attempting to Seize an Unavailable Drive If a neutral drive is seized by a port and the other port attempts to seize the drive the latter port will find the drive unavailable This will assert the Combined Error bit in Control and Status Register 1 bit 15 in RKCS1 Also the Drive Available bit in the Drive Status Register bit 0 in RKDS will be cleared as a result of a port attempting to seize an
107. heads load or if the heads show an oxide buildup on the gliding surface when inspected a head crash has probably occurred The following procedure should be followed for recovery from a head crash 1 Clean the oxide buildup from the head using a head cleaning kit 2 Reinspect the head using an inspection mirror If any scratches or oxide are visible the head must be replaced Refer to the RK06 RKO7 Disk Subsystem Service Manual for the appro priate procedure 3 Ifa head shows no visible scratches or oxides install a scratch pack and load the heads onto the pack as in the head load unload checks 4 When the drive is ready cycle the drive down by setting the RUN STOP switch and reinspect the heads in question If any oxide buildup is present a catastrophic condition exists Refer to RK06 RK07 Disk Subsystem Service Manual for corrective procedures CAUTION Never use a cartridge involved in a head crash until it has been inspected and found usable i e there are no visible marks on the platter surfaces Catastro phic crashes can be propagated from drive to drive under such conditions Never use an alignment cartridge in a drive that has experienced a head crash Always use a scratch car tridge for testing 45 POWER CONVERSION The RK611 RK06 or RK611 RK07 Disk Subsystem is configured during manufacture to be either a 50 Hz or 60 Hz unit See Paragraph 1 4 Although it is possible to change the frequency character is
108. herefore all procedures requiring the application of a lateral force to the carriage must be performed with the heads unloaded Positioner parts and fixtures are precision parts and must be handled carefully The use of torque wrenches set to specified levels is mandatory to guarantee that tolerances will hold and thus prevent damaged parts When transporting cartridges by auto always keep the cartridges in the passenger com partment never in the trunk Always transport cartridges in their carrying case 4 17 4 4 41 Head Alignment Using Program Control perform head alignment checks under program control 1 2 Remove rear cover from 06 or the RKO7 Ensure that power is removed from the RK06 or RK07 and the Field Test Box Connect the head alignment cable from the Field Test Box to the RK06 RK07 Read Write board Figure 4 10 and Figure 4 11 Apply power to the RK06 or RKO7 and the Field Test Box HEAD ALIGNMENT GROUND CONNECTOR HEAD ALIGNMENT CABLE CONNECTOR lt wy Sy 2 0986 Figure 4 10 Head Assembly Area 4 18 HEAD ALIGNMENT GROUND CONNECTOR HEAD ALIGNMENT CABLE CONNECTOR WW yyy Y SBR 7 lt v MA 0985 Figure 4 11 RK07 Head Assembly Area 5 Load MAINDEC 11 DZR6N and start at address 224 Per program instructions install an RKO6K AC or RK07K AC alignment cartridge de pending on the
109. ht should be considered in relation to the weight of the existing computer system and possible future expansion 3 2 5 Installation Constraints The route the equipment will travel from the receiving area to the installation site should be studied in advance to ensure problem free delivery Among the factors to be considered are the height and loca tion of loading doors the size capacity and availability of elevators the number and size of aisles and doors en route and any restrictions such as bends or obstructions in the hallways Any constraints should be reported to Digital Equipment Corporation as soon as possible so that the requirements of the individual installation site may be considered when the unit is packed for shipment 3 2 6 Fire and Safety Precautions The RK06 and RKO7 Disk Drives present no unusual additional fire or safety hazards to an existing computer system Wiring should be carefully checked however to ensure that the capacity is adequate for the added load and for any contemplated expansion The 60 Hz version of the RK06 or is UL listed and CSA certified The 50 Hz version complies with IEC 435 standards 3 5 4 INSTALLATION This chapter includes the procedures required for unpacking and installing RK611 RK06 RK611 RKO0 7 Disk subsystem Installation will only be performed by qualified DIGITAL personnel If the equipment is installed by personnel not so qualified the equipment
110. incrementing during data transfers This is primarily a diagnostic aid Subsystem Clear SCLR Bit 5 When the SCLR bit is set via program control the controller is cleared and the Initialize line is asserted on the drive interface to clear all of the drives available to the system Input Ready IR Bit 6 Input Ready is a read only bit that is set to indicate that the Silo input buffer is ready to accept a word Conversely the bit is reset to indicate that the Silo is full and cannot accept a word The IR bit is also set by conventional initialization INIT CCLR SCLR or by the setting of the GO bit 0 in 51 Output Ready OR Bit 7 Output Ready is a read only bit that is set to indicate that a word is in the Silo output buffer The bit is cleared by conventional initialization INIT CCLR SCLR or by the setting of the GO bit 0 in RKCSI Unit Field Error UFE Bit 8 Unit Field Error is a read only error bit that is set to indicate the following The Drive Select code bits 0 2 in the current status message 0 3 received by the controller did not compare with the Drive Select DSO DS2 code transmitted on the Message A line The setting of this error bit is disabled by diagnostic mode operation A Deselect bit T3 was transmitted to a drive on the Message A line and the Select Acknowl edge SACK signal was not negated Multiple Drive Select MDS Bit 9 Multiple Drive Select is a read only er
111. indicator will be illuminated When the alternating action pushbutton is again pressed write protection is removed and the PROTECT indicator is extinguished However if a Write Gate signal is asserted in the selected drive when the switch is initially pressed write protection will be inhibited and the PROTECT indicator will remain extinguished until the Write Gate signal is negated Once the Write Gate signal is negated the Write Protect mode is entered and the PROTECT indicator is illuminated 6 2 6 ACCESS A ENABLE Switch with SELECT Indicator The ACCESS A ENABLE switch is an alternating action pushbutton that is found on both single access drives Access A or Access B and drives containing the dual access option Port A and Port B 6 2 6 1 Single Access Drive A If Port is configured for single access i e the M7706 Interface and Timing Module is in Slot 3 of the drive card cage and the switch is initially depressed bidirec tional communication with the controller is enabled via Port A Under these conditions when the drive is selected by the controller the SELECT indicator is illuminated When the switch is again depressed released access by the controller is disabled and the SELECT A lamp cannot be illumi nated 6 3 6 2 7 ACCESS ENABLE Switch with SELECT Indicator The ACCESS B ENABLE switch is an alternating action pushbutton that is found on both single access drives Port A or Port B and drives containing the d
112. ion Routine 7 41 Maintenance Register Read 7 43 PROGRAM MING CONSIDERATIONS 7 43 CUSTOMER MAINTENANCE GUIDE INTRODUCTION 8 1 PROFESSIONAL CARTRIDGE CLEANING eene nnne nnno 8 1 USER CARTRIDGE CLEANING 8 1 SPINDLE ASSEMBLY CLEANING oso ies tiv esa tek Eee nee ENS 8 1 CARTRIDGE CARESUMMARY 8 3 vl 9 9 1 Figure No 1 1 2 1 2 2 2 3 2 4 2 5 2 6 2 7 CONTENTS CONT Page RK06 RK07 DUAL ACCESS OPTION RK611 RK06 AND RK611 RK07 DUAL ACCESS SYSTEM CONFIGURATION 9 1 DITAL ACCESS CONTROLS nini t ae a 9 2 Port AMG ees a undo peas 9 2 Port B MO sce ei eni EU tunt Lau dudes ds 9 2 Dual Access MOGE nc Naa a 9 2 DUAL ACCESS LOGIC CIRC ULES m ats teense eee 9 2 INE DICEATION aer ERN Mara En qq 9 3 Timing Counters Decoders ood ra neces 9 3 AuUto Release eue esteem 9 3 CES no PO des Ld A nid bein tetas 9 3 DUAL P
113. ion routine consists of three subroutines the first determines if an ECC correction cycle is necessary the second provides a set up procedure the third provides an ECC correction algorithm 1 Determine if ECC correction is necessary 35 BIT BNE BIT BNE BIT BNE BIT BNE TST BMI ZDLT WCE UPE PGE MDS UFE RKCS2 Controller error RETRY Branch if controller error ZJSPAR CTO RKCSI Controller time out or message parity RETRY Yes DTE HVRC RKER Drive timing or header error RETRY Yes RKER Hard error RETRY Data check error DMECC must be soft error 2 Error correction is accomplished six steps a Calculate byte offset to the start of the block transferred b Calculate offset to the first bad byte in the block c Convert correction pattern into two pattern words d Apply ECC correction using first pattern word e Apply ECC correction using second pattern word f Finish or continue previous function DMECC MOV RKWC RO negative words remaining ASL RO Convert to bytes ADD COUNT RO Calculate bytes actually transfered MOV RO SP save it BEQ DMOFST Branch to do offset if no bytes transferred BIT HVRC RKER Header VRC error BNE 40 Yes DEC RO Calculate offset to start of block 405 777 RO Clear residue BIT DTE IECH HVRC RKER Hard Error BNE DMOFST Yes do offset MOV RKECPS RI starti
114. ion to release the cartridge from the drive spindle d Lift the cartridge up and out of the shroud and place it in the protection cover e Lower the top cover handle to the horizontal position to lock the protection cover in place Figure 6 9 Engaging Lock Release 6 12 7 PROGRAMMER S GUIDE 7 1 INTRODUCTION Since the Unibus provides a unified addressing structure in which device control status and data registers are directly addressed as memory locations all operations on these registers are performed by normal memory reference instructions Under these conditions the RK06 or RKO7 subsystem does not require a special class of input output instructions to initiate the transfer of information or the manipulation of data With the exception of a Read Header command data transfers use the nonprocessor request NPR facility of the Unibus for direct access to main memory at a maximum transfer rate of 232 000 words per second 4 3 us per word in clock sizes ranging from 1 to 65 536 words To achieve a smooth flow of data from a selected disk to main memory the subsystem utilizes a first in first out 66 word data buffer known as a Silo A phase locked clock system PLO and modified frequency modulation recording technique add reliability to reading and writing operations regardless of minor changes in the rotational speed of a disk In addition program controlled head offset positioning can correct an
115. ion toward the spindle while a negative offset 1 provides head motion away from the spindle Therefore an offset value of all zero would leave the heads positioned on the centerline of a desired cylinder X Don t care RK06 RK07 Sign Magnitude Offset Offset 6 5 4 2 OFI 0 X 0 0 0 0 0 0 0 0 l X 0 0 0 0 0 0 0 0 0 X 0 0 0 0 0 1 25 12 5 1 X 0 0 0 0 0 25 12 5 0 X 0 0 1 0 0 0 200 100 1 0 0 1 0 0 0 200 100 0 X 0 1 0 0 0 0 400 200 1 X 0 1 0 0 0 0 400 200 0 X 1 0 0 0 0 0 800 400 1 X 1 0 0 0 0 0 800 400 0 X 1 0 0 0 0 1200 600 1 X 1 0 0 0 0 1200 600 Once an Offset command has been executed selected drive will remain in offset mode if read commands 1 Read Data Write Check Read Header or a Seek are subsequently initiated for the current cylinder However write commands i e Write Data Write Header to the current cylinder or any head motion initiating command to another cylinder will disable the offset condition and allow the heads to be repositioned over the centerline of the desired cylinder Attention ATNO ATN7 Bits 8 15 The Attention indicators are read only bits that reflect the unit number of each drive available to the system and the condition of an associated Drive Status Change flip flop All of the bits are continuously monitored and updated polled every 7 4 Under these conditions and during the
116. ional shakedown of the entire sub system including the Unibus interface and access to main memory Employs worst case situations involving mechani cal positioning disk addressing and data transfer Measures drive operational timing Provides numerous options that may be specified by the user Reports errors as they occur Continues functional shakedown provided by Part l Employs worst case situations involving head off setting memory addressing and data transfer Unibus cycle contention and multiple drive oper ations Provides RK06 or 07 head alignment aid Provides numerous options that may be specified by the user 5 6 Table 5 2 Diagnostic Programs Used RK611 RK06 RK611 RK07 Program MAINDEC 11 DZR6P RK611 RKO06 RK611 RK07 Performance Exerciser MAINDEC 11 DZR6Q RK06 RKO7 Drive Compatibility MAINDEC 11 DZR6R RK611 RK06 RK611 RK07 User Defined Test 5 2 1 2 Procedure Disk Subsystem Field Acceptance Tests Cont Function Exercises in a random overlapped routine from one to eight RK06 or RKO7 Drives connected to a common Unibus controller in a dedicated stand alone mode Generates read write and write write check com mands Reports system errors upon occurrence Maintains performance statistics on each drive Verifies the capability of each drive to write data that can be successfully read by all other drives and to completely overwrite data written by all other
117. ished by starting the Performance Exerciser at address 200 The exerciser will automatically write all disks on the drive bus and begin testing data transfers immediately without operator intervention While the first system is undergoing testing the second system can be entered into the performance tests by starting the exerciser at address 240 This will skip over the disk write portion of the test and allow the operator to manually begin the performance tests on the second system Both systems will then be undergoing performance testing simultaneously 4 26 CONTROLLERS MAY OPERATED FROM THE SAME UNIBUS FROM TWO SEPARATE UNIBUS RK amp 11 CONTROLLER A 1 CONTROLLER DRIVE BUSA DRIVE BUSB TERMINATORS Wie 32 3 OUT OUT LL Il jl l Ll Ll ll l a iH 2 T our our J J2 J3 J4 1 2 J3 J4 J1 1 OUT IN OUT IN OUT IN OUT IN OUT QUI IN In ouf IN our P B DRIVE UNIT 0 DRIVE UNIT 1 DRIVE UNIT 2 DRIVE priveuniT3 DRIVE UNIT 4 DRIVE UNIT 5 DRIVE UNIT 6 DRIVE UNIT 7 11 5508 Figure 4 12 RK611 RK06 or RK611 RK07 Dual Access Cable Diagram 4 27 5 TEST PROCEDURE Field acceptance testing is intended to demonstrate performance of the RK06 or RKO7 Disk Drive and or the RK611 Controller to the customer prior to his acceptance of the equipment 5 1 ERROR DEFINI
118. line indicate operational status and error conditions and provide controller access whether the drive is a single or dual access type For drive power an ON OFF circuit breaker with a POWER indicator is located at the rear of each drive 2 1 8376 2 Figure 2 1 H969 Standalone Disk Drive Cabinet 2 3 RK06K AND RK07K DISK CARTRIDGES The RKO6K and RK07K Disk Cartridge assemblies contain two platters Figure 2 2 Three of the disk surfaces 0 1 and 2 are available for data storage read write surfaces while the fourth surface S is a read only servo surface that is used for head positioning and the derivation of timing Each RKO6K disk surface is divided into 411 tracks and each RKO7K disk surface contains 815 tracks The radial location of each track is defined as a separate cylinder position that is common to all surfaces Under these conditions a cylinder can be specified cylinder address via a Seek command from the controller to cause the simultaneous positioning of all four heads to one of the 411 cylinder positions 0 410 for the RK06 or one of 815 cylinder positions 0 814 for the RK07 A single read write head 0 1 2 can then be defined track address via a Read or Write command to select one of the three read write surfaces In this manner a desired track is accessed for data transfer operations Data Track Formatting During disk formatting procedures each data track is located and recorded wit
119. ly and contaminating the disks CAUTION For cleaning purposes use only a solution of 9 per cent water with 91 percent isopropyl alcohol Water trichloroethylene or other solvents are not permit ted 8 4 SPINDLE ASSEMBLY CLEANING Using a lint free wiper dampened with the isopropyl alcohol 91 percent solution clean the spindle cone magnet and spindle rim Figure 8 1 prior to the loading of a cartridge However do not saturate the assembly remove all excess solvent with a dry wiper This is necessary to prevent solvent from entering a loaded cartridge and contaminating the disks In addition ensure that the shroud is as free of lint and dust as possible before loading a cartridge Dry lint and dust may be blown from the spindle area using filtered dry air However do not use plant air which may contain water or oil canned air is an acceptable substitute 8 1 SPINDLE MAGNET SPINDLE RIM CONE 7897 12 Figure 8 1 Spindle Assembly 8 5 CARTRIDGE CARE SUMMARY The following listing summarizes and cleaning considerations for the RKO6K or RK07K Disk Cartridge I 1 2 10 Keep cartridges clean Use cartridges at computer room temperature only Manipulate cartridges by the top cover handle only When the protection cover is removed for loading do not touch disk surfaces hub center cone or surfaces When the protection cover is removed for loading interior metal hub surfaces must be clean When
120. mmand a Start Spindle command from the controller can be used to restart the spindle and reload the heads However if the cycle down condition occurred as the result of an error a Start Spindle command is not required since the heads will immediately load when the error is cleared Finally since the RUN STOP switch has a mechanical memory a power interrupt followed by power restoration will cause the drive to automatically cycle up if the switch is currently located in the RUN position NOTE The drive cover lid is interlocked This prevents the lid from being raised when the STOP light is extin guished RUN or drive power is off REAR PANEL CIRCUIT BREAKER POWER ON OFF SWITCH FRONT PANEL AC INDICATOR CONTROL PANEL RUN 1 READY ACCESS ENABLE SW WITH ACCESS INDICATOR ON DUAL ACCESS B PORT SELECTED RUN STOP SW WITH STOP INDICATOR STOP OFF RUN UNIT SELECT SW WITH READY INDICATOR ACCESS A ENABLE SW WITH ACCESS INDICATOR ON SINGLE ACCESS DRIVE READY SELECTED OR DUAL ACCESS A PORT SELECTED FAULT INDICATOR WRITE PROTECT SW WITH ON FAULT PROTECT INDICATOR ON WRITE PROTECT 7897 11 Figure 6 1 or RK07 Disk Drive Power and Control Switches 6 2 6 2 3 UNIT SELECT Switch with READY Indicator The UNIT SELECT switch is a cam operated assembly that is actuated by the insertion of a plastic plug The face side of each ac
121. must be set and reset four times 1 0 1 0 1 0 1 0 to derive a control clock cycle Data Bit Simulation To simulate read data transfers Read Write Check Read Header it is neces sary to set and reset the Read Data MERD bit in conjunction with MCLK A data bit of one can be simulated by the setting of the MERD bit However a data bit of zero can be simulated by either the setting or resetting of the MERD bit depending on the condition of the previous data bit i e a data zero following a one requires a reset condition while a data zero following a zero requires a set condition To simulate write data transfers Write Write Header the MERD bit is set and reset in conjunction with MCLK in the same manner described for read simulation However for write data transfers it is necessary to monitor the condition of the resultant Write Data MEWD bit 7 4 1 Read Simulation The following sequence of operations simulates the reading of a complete sector in Diagnostic mode 1 With the exception of RKCSI load the device registers required to support a normal Read command 2 Set the Diagnostic mode DMD bit bit 5 in RKMRI 3 Load Command 1 4 bits bits 1 4 and GO bit bit 0 in RKCSI 4 Toggle the Maintenance Clock MCLK bit bit 8 in RKMR I a sufficient number of times to assemble Message A MR2 and Message B MR3 and simulate transmission message swap between MR2 and MR3 5 Toggle 1 0 the Maintenance
122. nd the STOP indicator will be extinguished in approximately one second When the drive has completed the start spindle sequence and the heads are detented on Cylinder 0 the READY indicator on the numbered UNIT SELECT switch will be illuminated SHROUD BRUSH INDICATOR DIAL READ WRITE amp SERVO HEADS BRUSH ARM ASS Y PACK ON SWITCH SPINDLE LATCH OPERATOR SWITCHES UPPER BEZEL LOWER BEZEL 8174 3 Figure 6 2 Typical Disk Drive with Lid Released 6 5 HANDLE HANDLE SLIDE DATA CARTRIDGE PROTECTION COVER 8365 3 Figure 6 3 Typical Disk Cartridge HANDLE HANDLE SLIDE 8089 4 Figure 6 4 Engaging Lock Release 6 7 COVER 8089 3 Figure 6 5 Releasing Protection Cover 6 8 RELEASE 8365 1 Figure 6 6 Positioning for Load 6 9 Figure 6 7 Cartridge Locked in Shroud 8089 1 6 10 Figure 6 8 Installing Protection Cover 6 11 6 3 2 06 and 07 Cartridge Unloading 1 Release the RUN STOP switch STOP 2 Wait for the STOP indicator to illuminate 3 Press the lid release bar and raise the lid 4 Remove the 06 or RKO7K Cartridge as follows a Remove the protection cover and hold the cover in the left hand b Push the handle slide to the left before raising the top cover handle Figure 6 9 c Raise the top cover handle to a full upright posit
123. nerated by MCLK replace the data clock used to control the transfer of data bits to and from the Encoder Separator logic and to generate control clock used to synchronize message shifting oper ations Predicating the generation of controller timing on the system clock allows a diagnostic program to monitor response conditions following the initiation of each write operation to RKMRI For the proper simulation of data transfers in Diagnostic mode the following timing considerations must be adhered to Figure 7 29 7 30 ACTUAL DATA SIMULATED READ TIMING BIT BIT BIT CELL CELL CELL PRE COMPENSATION SIGNALS ARE VALID DURING THESE INTERVALS ACTUAL DATA SIMULATED WRITE TIMING DATA SEQUENCE DATA ONE DATA ZERO FOLLOWING A ONE AND DATA ZERO FOLLOWING A ZERO 11 4228 Figure 7 29 Maintenance Clock MCLK Timing Data Clock 1 set followed by a reset of the MCLK bit equals one data clock cycle 2 Two data clock cycles are required to provide a bit cell for a single data bit 1 or 0 there fore MCLK must be set and reset twice 1 0 1 0 3 16 bit word time is equal to 32 data clock cycles 4 18 bit word time is equal to 36 data clock cycles 5 NPR time for a 16 bit word is equal to 37 clock cycles 6 NPR time for an 18 bit word is equal to 40 clock cycles Control Clock One control clock cycle is equal to four times duration of a data clock cycle therefore the MCLK bit
124. ng bit number DEC RI Convert to relative bit number MOV R1 SP Save starting relative bit number BIC 1 lt 17 gt 1 Isolate shift count BIC R1 SP Clear shift count in relative bit number CLC Calculate offset to first bad byte ROR SP the block ASR SP ASR SP ADD SP RO Set offset to first byte in the transfer MOV RKECPT R3 correction pattern word CLR SP Clear second pattern word 415 DEC shifts to perform BLT 425 If LT ASL R3 This is in case the error burst ROL SP crossed a 16 Bit word boundary BR 418 Continue shifting 428 CALL Apply ECC correction MOV SP R3 Retrieve second pattern word ADD 2 RO Update offset to next bad Word CALL DMECOR Apply ECC correction Correction completed Finish or continue function 7 42 3 Correction Algorithm This routine calculates memory address of the data word to be corrected and using the supplied pattern word performs an XOR of the pattern and data words DMECOR CMP RO COUNT Byte Offset Within Range BHIS 505 ADRS R1 Retrieve starting buffer address ADD RO R1 Calculate address of word to be corrected MOV R1 SP Prepare to do BIC R3 R1 NOT pattern AND data word BIC SP R3 NOT data word AND pattern BIS R3 R1 Pattern OR data word 505 7 6 3 Maintenance Register Read Routine The Maintenance Register Rea
125. ntional initialization INIT CCLR SCLR initiating a new command writing into 51 selecting a new drive writing into RKCS2 or whenever an Attention signal is asserted by the selected drive for a drive status change 7 2 7 Drive Error Register The RKER register Figure 7 8 is a read only register that is used to store the error status of a selected drive However error information obtained from the drive is not immediately available to program control until the information is validated SVAL by the setting of bit 15 in the RKDS register which indicates that a complete status message frame has been received Error bits set in the RKER register can be cleared by conventional initialization INIT CCLR SCLR However a Controller Clear CCLR operation does not affect error bits that are currently set in the drive In addition a Unibus Initialize INIT or Subsystem Clear SCLR operation can only reset error bits in a drive if the associated error condition no longer exists Illegal Function ILF Bit 0 Illegal Function is a read only bit that is set to indicate that an illegal command 33 353 373 has been loaded into RKCSI Seek Incomplete SKI Bit 1 Seek Incomplete is read only bit that is set Message 10 to indicate that a seek operation has not been completed for one of the following conditions An Invalid Address Message BO T5 bit has been set in the drive indicating that the drive has
126. ntrol to initialize both the controller and all of the drives When command execution is completed the GO bit is reset and the controller is ready to accept a new command However the GO bit cannot be set if the Combined Error CERR bit is set When CERR is set the execution of a command can only occur following the initiation of a CCLR Function Code F1 F4 Bits 1 4 The configuration of the Function Code bits FI F4 with the setting of the GO bit allows the selected drive to respond to the following command control con figuration 4 2 Octal Select Drive 0 0 0 0 1 Ol Pack Acknowledge 0 0 0 1 1 03 Drive Clear 0 0 1 0 1 05 Unload 0 0 1 1 1 07 Start Spindle 0 1 0 0 1 11 Recalibrate 0 1 0 1 1 13 Offset 0 1 1 0 1 15 Seek 0 1 1 1 17 Read Data 1 0 0 0 1 2 Write Data 1 0 0 1 1 23 Read Header 1 0 1 0 1 25 Write Header 1 0 1 1 1 27 Write Check 1 1 0 0 1 31 Spare Bit Spare bit 5 can be written and read back 7 2 CONTROL ano sravus REGISTER 1 UNIBUS RKCS1 gt READ WRITE ADORESS CERR 14 11 10 07 06 05 04 o2 WORD C PUNI REGISTER RKWC R W 777442 wc WC WC WC WC wc WC WC WC we wc WC 15 14 13 12 11 10 09 08 07 06 05 04 02 01 00 BUS ADDRESS REGISTER R W 777444 DISK ADDRESS TRACK 8 SECTOR REG RKDA R W 777446 CONTROL AND STATUS REGISTER 2 RKCS2 R W 15 817 0 DRIVE STATUS REGISTER RKDS READ ONLY 15 817 0
127. ntroller with the RK06 or RKO7 Drives and PDP 11 Unibus Figure 7 1 The 15 registers are loaded and or read under program control to initiate selected disk commands and monitor subsystem status and error conditions Device register bits are generally cleared by a Unibus Initialize INIT Controller Clear CCLR or Subsystem Clear SCLR operation In the following descriptions unless otherwise specified it should be understood that the clearing of a bit by any one of these three methods is implied NOTE The RK611 does not recognize DATOB MOVB BICB etc bus cycles All registers must be written as words 7 2 1 Control Status Register 1 RKCS1 The RKCSI register Figure 7 2 be read or written via program control and as such is used to store the current disk command function code and operational status of the controller In addition the register can initiate command execution and enable a Controller Clear operation Go GO Bit 0 When the GO bit is set the disk command Function Code F1 F4 is executed With the GO bit set only two other device register bits can be set Diagnostic Mode excepted as follows Controller Clear CCLR bit 15 in RKCSI may be set via program control to initialize gen eral clear and preset certain device registers within the controller However any status and or error conditions set in the drives are not affected Subsystem Clear SCLR bit 5 in RKCS2 may be set via program co
128. o eight RK06 and or RK07 Disk Drives Each RK06 or RK07 Disk Drive is a free standing high perform ance moving head device that is connected to the RK611 Controller via a daisy chain bus arrange ment The RK611 Controller which is located in associated cabinetry provides an interface between the subsystem and a PDP 11 system via the Unibus Both RK06 and 7 drives can be used in one subsystem Dual access operation is a hardware option available to the RK06 and RK07 drives enabling either drive to be accessed through two different controllers Figure 1 1 shows the system configuration for dual access operation The eight RK06s or RKO7s shown this figure are daisy chained The two RK611 Controllers can be attached to two different CPUs CENTRAL MAIN PROCESSOR MEMORY 71 RK611 CONTROLLER B Z2 RKet CONTROLLER A rU es 6 7 pl sl RK 6 OR RK 6 OR RK 6 OR RK 6 OR RK 6 OR RKO6 OR RK 6 OR RK 6 OR RK 7 DISK DISK RK 7 DISK RKO DISK RK 7 DISK RK 7 DISK RKO7 DISK DISK DRIVE DRIVE DRIVE DRIVE DRIVE DRIVE DRIVE DRIVE RK 6 RKO7 SUBSYSTEM CONFIGURATION PDP 11 11 4326 Figure 1 1 RK611 RK06 RK611 RK07 Dual Access System Configuration 1 2 1 Media The storage medium for the RK06 Disk Drive is a top loading dual platter RKO6K Disk Cartridge which be freely interchanged between any of the available RK
129. o separate units to the RK611 The unit select switch will define the address of port A Port B address will respond to the address of A 1 Since each drive will appear as two units to RK611 a maximum of four drives can be tested by this program at one time 2 Use only even numbered UNIT SELECT SWITCHes on the drive operator panel Any other drives on the system which have an address in conflict with any of the test addresses must be deselected both ports switched off The even numbered UNIT SELECT SWITCH on any drive can be removed and inserted into the current drive under test 3 this test the RK611 is daisy chained to both ports of the RK06 or RKO7 Refer to Figure 4 13 for the cable diagram 4 Run the Dual Access diagnostic DZR6G X to test dual access seizure and release 5 After running the Dual Access diagnostics return the dual port invert switch on module M7730 to its normal position Return any UNIT SELECT SWITCHes borrowed for this test to their former drives 6 the RK06 or RKO7 Drives for normal operation as shown in Figure 4 12 4 6 2 2 RK611 RK06 and RK07 Diagnostics With the drives cabled for normal operation run all RK611 06 and 07 diagnostics through both ports 4 6 2 3 06 and RK07 Performance Exerciser To run the RK06 and RKO7 Performance Exerciser DZR6P X all disks must first be written with a predetermined data pattern through one of the systems This is accompl
130. o the actual cylinder address that corresponds to the underlined for RK06 alignment or dotted for RK07 alignment cylinder address switches on the Field Test Box Refer to the switch settings that follow Switch Position EXERCISE STATUS EXERCISE DRIVE SELECT Desired Drive COMMAND TO RK06 7 switches SEEK 22 SECTORS CLEAR ERROR amp ATTENTION SET VOLUME VALID SINGLE CYCLE CONTINUOUS SINGLE CYCLE FUNCTION SEEK ONLY 4 23 Switch Position SYNC INT CLOCK FAST HALT ON ERROR NO ADDRESSING CYLINDER SWR UNDERLINED SWITCHES 24510 for RK06 DOTTED SWITCHES 49610 for RKO7 HEAD SWR HEAD ADDRESS SWITCHES DESIRED HEAD SECTOR ALL PARITY both switches ODD Depress the INIT and START switches To change heads simply change the head address switch setting and depress the START switch Set the torque wrench to 5 in lb Tighten the forward vertical screw to move the head forward or the rear one to move the head to the rear Since these screws oppose each other the screw not in use must be fully CCW Adjust until a The red LED on the meter is ON steady state b meter on the Field Test Box reads 0 50 microinches NOTE If the needle deflects to the right of zero move the head toward spindle if to the left of zero move away from spindle Unload the heads Tighten the head to 5 in lb Load the heads and check that the conditions of steps 10a and b are still met Unload the heads remo
131. o the left of zero move away from spindle Unload the heads by typing the same head number again Tighten the head to 5 in lb Reload the heads by typing the letter R Check that the conditions of steps 10a and 106 met Unload the heads by typing the next head number Remove the head alignment fixture and mount it on the next head Repeat steps 8 through 15 for each data head When all heads have been aligned restart the program under its Exerciser mode and exercise the drive for two minutes Recheck the alignment of all heads Remove the alignment cartridge and the Field Test Box clear the WRITE LOCK condition Replace the rear cover removed in step 1 4 20 4 4 4 2 Sample Printout DZR6N D RK611 RK06 RK07 SUBSYSTEM VERIFICATION PART 2 RK06 07 HEAD ALIGNMENT AID FOR HELP TYPE H ELSE CR H INSTRUCTIONS FOR USING RK06 RK07 HEAD ALIGNMENT AID HII k k II k k k k k k k k k k k k k k k k k k k k k k k k k k k k k k k k k k k k k O MOUNT RKO6 RK07 ALIGNMENT CARTRIDGE ON THE DESIRED DRIVE AND INSURE THAT THE DRIVE IS WRITE LOCKED CONNECT THE ALIGNMENT INDICATOR TO THE DESIRED DRIVE VIA THE HEAD ALIGNMENT CABLE ONLY AND CYCLE UP THE DRIVE AFTER MOUNTING THE PACK ON THE DRIVE THE OPERATOR SHOULD WAIT 30 MINUTES FOR THE DRIVE TEMPERATURE TO STABILIZE BEFORE PROCEEDING WITH ALIGNMENT RESPOND TO ALL REQUESTS FOR PARAMETERS BY ENTERING THE DESIRED PARAMETER VALUE NEEDED
132. o use another system containing an RK06 or RKO7 to complete this test Load and run diagnostic DZR6Q If any drive fails this test make any necessary adjustments and rerun the program 5 2 4 Data Reliability Testing 5 2 41 General Data reliability testing consists of running diagnostic DZR6P to exercise the entire RK611 RK06 RK611 RKO7 Disk subsystem for at least one hour At the end of this period the statistics for each drive must be analyzed to determine whether or not the subsystem is to be accepted In analyzing the statistics for each drive no control seek or hard errors are allowed during the testing Refer to Paragraph 5 1 for error definitions The occurrence of any of these errors is sufficient to reject a drive If it is determined that an error is due to the cartridge it is sufficient to reformat the cartridge making sure that the failing sector is flagged bad The Data Reliability Test may then be rerun If the excessive errors are caused by the drive required corrective action must be performed and the acceptance restarted with the RK06 or RKO7 diagnostic Paragraph 5 2 2 If the errors are caused by the controller the acceptance must be restarted with the RK611 diagnostic testing Paragraph 5 2 1 after the corrective action has been taken 5 2 4 2 Procedure Load and start diagnostic DZR6P Write the cartridge on all drives on the sub system After one hour of running obtain the statistics for all drives If the
133. ollows 1 When delivered the RK06 or RKO7 is mounted on a shipping skid and covered a card board carton Figure 4 1 First remove the two plastic straps that hold the carton secure to the skid Remove the lid from the top of the cardboard carton Remove the staples that fasten the three wooden slats to the bottom flanges of the card board overlay carton Also remove the staples from that portion of the carton that extends to the bottom of the skid Visually inspect the cabinet for outward signs of shipping damage Retain the original pack ing materials and receipts in case any claims are filed for shipping damage All damage claims should be promptly filed with the transportation company involved DIGITAL should be notified immediately of any such claims Remove the rear and front covers visually inspect for shipping damage Inspect for loose cable connectors terminal points and subassemblies Inspect the logic backplane in both X and Y axes for bent pins Pull rearward on the logic assembly and power supply assembly Both assemblies will open 90 degrees for access to components Ensure that all logic modules are in proper locations refer to Figures 4 2 and 4 3 Check cable connectors and modules in logic assembly card nest to ensure tight ness Check power supply assembly for loose cables connectors and connection of air distri bution system bellows Figure 3 2 and proper installation of the mo
134. ons RK06 RK07 400 microinches 200 microinches 400 microinches 200 microinches 800 microinches 400 microinches 800 microinches 400 microinches 1200 microinches 600 microinches 1200 microinches 600 microinches 2 5 6 Bad Sector File The bad sector file is a list of the bad sectors in the cartridge and is recorded on the highest number track and cylinder on the cartridge This track is written in 22 sector 16 bit word format 2 14 2 5 6 1 Data Field Format data field in the first 10 sectors of the last track has the following format Word 0 Cartridge serial number 1 Zeros 2 Zeros 3 Zeros for data cartridge at least a single one for alignment cartridge s Bad sector 1 Bad sector 2 n 1 2 Ones filled n 3 Ones filled Last bad sector 254 Ones filled 255 Ones filled Word 0 contains the octal cartridge serial number Words 1 and 2 are filled with zeros and are reserved for future use Word 3 identifies the cartridge either as a data or alignment cartridge If word 3 contains any ones words 4 through 255 shall also contain ones The list identifying bad sectors on a data cartridge is contained in two word groups beginning with words 4 and 5 The first word of the two word group contains the cylinder address of the defect and the second word contains the track and sector address of the defect The defective sectors are listed in ascending order by cylinder track and sector addres
135. orrected using ECC Bad Sector Error BSE Bit 7 Bad Sector Error is a read only bit that is set to indicate that a data transfer has been attempted to or from a sector that has at least one of the two Good Sector Flag bits Header Word 2 bits 14 and 15 reset indicating a bad sector Header Vertical Redundancy Check Error HRVC Bit 8 Header VRC Error is a read only bit that is set to indicate that a VRC error has been detected on a sector header during a data transfer If the Operation Incomplete OPI bit bit 13 is also set the sector in which the error occurred cannot be determined However if the OPI bit is reset the VRC error was detected in the sector currently defined by the RKDA register SAO SA4 Cylinder Overflow Error COE Bit 9 Cylinder Overflow Error is a read only bit that is set to indicate that RK WC register is not equal to zero following a data transfer from cylinder 41010 for RK06 81410 for RK07 track 210 and sector 1910 or 2110 last logical sector 7 13 Invalid Disk Address Error IDA E 10 Invalid Disk Address Error is read only bit that can be set via an Invalid Address bit Message BO T5 from the drive or by the controller to indicate following 1 Invalid Address Message BO T5 bit has been set in the selected drive indicating that an invalid cylinder address 41110 through 51110 for the RK06 or 81510 through 102310 for the RK07 and or an invalid track
136. pable of performing read and write data operations in both 20 and 22 sector formats Performs worst case patterns spiral writing and reading and all offset operations Reports unloading and loading times 5 5 Table 5 2 Diagnostic Programs Used RK611 RK06 or RK611 RK07 _ Disk Subsystem Field Acceptance Tests Cont Program MAINDEC 11 DZR6J Unibus RK06 RK07 Drive Diagnostic Part 3 MAINDEC 11 DZR6K RK611 Functional Controller Diagnostic MAIN DEC 11 DZR6L RKO6K RKO7K Cartridge Formatter MAINDEC 11 DZR6M RK611 RK06 RK611 RK07 Subsystem Verification Part MAINDEC 11 DZR6N RK611 RK06 RK611 RK07 Subsystem Verification Part 2 Function Ensures that the disk is capable of performing all operator intervention functions properly Checks error detection logic by manual and by software forcing of errors Tests those areas in the controller that cannot be tested in a diskless environment Tests those areas of the drive that cannot be tested until controller operation in a diagnostic or maintenance mode has been tested Writes and verifies header and data information on an RKO6K or RKO7K disk pack at all possible disk pack addresses Uses the Bad Sector File to 1 report the serial number or the cartridge being formatted 2 deter mine whether the cartridge is an alignment car tridge and abort the program if it is and 3 identify the sectors that are to be marked Provides a funct
137. protect switch the other port is notified through the assertion of its attention The change is reported to the controller through the Write Lock Status bit 9 4 9 6 ERROR HANDLING IN DUAL ACCESS 9 6 1 Notifying the Controller of an Error Errors occurring while the RK06 or is seized by one of the controllers are reported to and should be serviced by that controller The other port is not notified of these errors The exception is a parity error in a message to the drive on the other port If the drive had been in the neutral state and an error occurred both Attention bits would be asserted The fault bit would be read by both controllers when servicing the Attention set in the drive However only one of the controllers would service the error in the drive 9 6 2 Clearing the Drive Errors and Resetting the Attention Either controller may issue a drive clear at any time The Attention in the drive will be cleared by a drive clear regardless of the drive s availability However to clear the flip flops storing the error information the drive must be seized to the port issuing the Clear command as mentioned The con troller to drive parity error is handled differently from the other errors and is discussed in greater detail as follows 9 6 3 Multiple Drive Select and CTD Parity Errors the Dual Access Configuration If a controller has seized a drive and multiple drive select is detected in the drive the drive handles this as
138. r is removed Drive Ready DRDY Bit 7 Drive Ready is a read only bit that is set Message A0 T7 to indicate that the selected drive is up to speed and the heads are properly positioned over a valid cylinder Under these conditions the drive is prepared to receive a command Disk Drive DDT Bit 8 Disk Drive is a read only bit that is internally conditioned Message 0 8 to indicate the type of drive selected For an RK06 Drive the bit remains reset and for an RKO7 Drive the bit is set However before any commands can be executed the bit must compare with the condition of Controller Drive Type bit 10 in RKCSI Spare Bits 9 and 10 Spare bits 9 and 10 are 0 Write Lock WRL Bit 11 Write Lock is a read only bit that is set Message 0 11 if the selected drive is write protected Spare Bit 12 Spare bit 12 is O Positioning in Progress PIP Bit 13 Positioning in Progress is a read only bit that is set Message A0 T13 to indicate that the heads are in motion Current Drive Attention CDA Bit 14 Current Drive Attention is a read only bit that is the logical equivalent of the Drive Status Change DSC bit in the drive defined by the Drive Select DS0 DS2 bits in RKCS2 The assertion of the Attention line indicates any one of the following conditions 1 Seek operation is completed 2 Offset operation is completed 3 Heads are unloading Message 14 4 FAULT condition
139. r revolution index pulse and each of the sector pulses derived from the servo tracks Index Sector pulses are enabled when the drive asserts a Select Acknowledge SACK and remain enabled until the drive is deselected Multiple Drive Select Line When asserted this line indicates to the controller that more than one drive has simultaneously re sponded to a given selection code To accomplish this a selected drive monitors the Index Sector pulse line for the presence of extraneous pulses If pulses other than those originating from the desired drive are detected the Multiple Drive Select line is asserted However since Index Sector pulses cannot be generated if the heads of the desired drive are not loaded the drive uses Control Clock pulses inserted on the Index Sector pulse line to monitor and detect multiple selections Finally the Multiple Drive Select signal is asserted within 300 ns of pulse detection from another drive and negated within 120 ns after the receipt of another message strobe Initialize Line When asserted this signal deselects all drives resets all error conditons and clears the Drive Status Change flip flop Controller Power on Line When asserted this line indicates that power has been applied to the controller and a physical con nection via cable has been established between the controller and the available drives When the line is negated all drives are deselected and the Strobe and Control Clock logic is
140. r software correction data error bursts up to 11 bits in length and to detect but not identify error bursts longer than 11 bits In addition and in conjunction with ECC operation each RK06 or RKO7 Disk Drive has a head offset capability which allows the recovery of data from a cartridge that may have been written by a drive with slightly misaligned heads The data recovery features operating in conjunction with the PLO clock system and the MFM record ing technique provide increased reading and recording reliability However if a disk pack sector pro ves defective in that hardware and software attempts to recover its data are unsuccessful the operating system can mark the sector bad so that future attempts to use it will be prevented 1 2 3 Maintenance Features The following standard features enhance the reliability availability and serviceability of the RK611 RK06 and RK611 RKO0 7 subsystems The RK06 and RKO7 drive interface is designed to allow a specific drive to be electrically isolated from the controller for maintenance purposes This allows an off line tester to be utilized without taking a drive off line or cycling the system down With this feature software can reconfigure the system for continued operation while maintenance is performed e Service access to 06 or RKO7 Disk Drive is from the top front or rear of the drive and not from the sides Therefore maintenance for a given drive in a multidrive system can be
141. re sent a return status is requested Unload 07 Figure 7 22 This command is used to unload the heads in a selected drive and stop the spindle This is accomplished by the setting of the Set Medium Off Line bit in Message A T10 When Messages A and B are sent a return status is requested 7 23 MESSAGE DR SEL CODE FROM RKCS2 BITS 0 2 BIT ODD RELEASE BIT X 0 1 0 1 52 BIT 3 0 1 00 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 MESSAGE B STATUS REQ PARITY FROM RKMR1 ODD BITS O 1 X 0 1 11 4217 Figure 7 19 Select Drive Command 0 02 04 05 06 08 09 m 12 18 14 15 DR SEL CODE FORMAT SET VOLUME PARITY FROM RKCS2 BIT VALID BIT ODD BITS 0 2 X 0 1 X 0 1 MESSAGE B PARITY ODD 11 4218 Figure 7 20 Pack Acknowledge Command 00 02 05 04 05 06 09 10 M 12 13 14 15 MESSAGE A DR SEL CODE DRIVE PARITY FROM RKCS2 CLEAR BIT ODD BITS 0 2 FORMAT BIT 0 1 0 1 0 02 03 04 05 06 07 08 09 oO 11 12 15 14 15 MESSAGE PARITY ODD 11 4219 Figure 7 21 Drive Clear Command 7 24 07 02 04 09 11 14 CI PLEET Bs SEL CODE PARITY P SET MEDIUM X2071 OFF LINE BIT MESSAGE B PARITY ODD 11 4220 Figure 7 22 Unload Command Start Spindle 11 Figure 7 23 This command is used to start the spindle execute a brush clean cycle and load the heads in
142. rioN n RK611 RKe11 CONTROLLER 2 UAE UJ pae t 06 OR 06 RK 6 OR RKO6 OR 0 DISK nor vd Kor DISK RKO7 DISK RK 7 DISK DRIVE VE DRIVE DRIVE p gt 7 EET UU 237 202 292 232 11 4326 Figure 9 1 1 6 1 Dual Access System Configuration 9 1 9 2 DUAL ACCESS CONTROLS A drive with the dual access option has three modes of operation with respect to its two ports Figure 9 2 shows the Port A and Port B control switches that determine which mode the drive is in When these two switches are not pressed all attempts to issue a command to the drive will generate a non existent drive error in the controller issuing the command CONTROL PANEL 1 222 WRITE ACCESS B ENABLE SW WITH ACCESS INDICATOR RUN STOP SW WITH ON DUAL ACCESS STOP INDICATOR B PORT SELECTED STOP F OFF RUN ACCESS A ENABLE SW UNIT SELECT SW WITH WITH ACCESS INDICATOR READY INDICATOR ON SINGLE ACCESS DRIVE ON READY SELECTED OR DUAL ACCESS A PORT SELECTED FAULT INDICATOR WRITE PROTECT SW WITH FAULT PROTECT INDICATOR ON WRITE PROTECT FRONT PANEL 11 5329 Figure 9 2 RK06 or RKO7 Disk Drive Power and Control Switches 9 2 1 Port A Mode With only Port enabling switch depressed drive will respond to Port A Any attempt by B to issue a command to the drive will generate a
143. rive When bit 12 is set a 20 sector format 18 bit data words is defined for the sector counter located in the drive and when the bit is reset a 22 sector format 16 bit data words is defined Drive to Controller Parity Error DTC PAR Bit 13 The DTC Parity Error is read only bit that is internally set to indicate that a parity error has been detected in a serial message received from the selected drive Drive Interrupt DI Bit 14 In relation to program control Drive Interrupt DI is a read only bit When set the bit differentiates between a drive initiated interrupt and a controller initiated interrupt The DI bit is set when any drive sets an associated Attention ATNO ATN7 bit 8 15 in RKAS OFP Thus if the Interrupt Enable IE bit is set the setting of the DI bit with the set condition of Controller Ready RDY bit 7 in RKCSI indicates a drive initiated interrupt The DI bit is reset by Unibus Initialize INIT Subsystem Clear SCLR or by the execution of Drive Clear commands to all drives asserting Attention 7 4 Combined Error Controller Clear CERR CCLR 15 As a Combined Error CERR indicator bit 15 can be set by the controller or any one of the drives to indicate that a subsystem error has occurred Table 7 1 However when the bit is set via program control a controller initialize CCLR operation is enabled which clears the controller and results in the clearing of bit 15 itself Thus if the
144. ror bit that is set by the assertion of the MDS signal line via the selected drive when a determination is made that at least one other drive has been simultaneously selected This bit can only be cleared by a Unibus Initialize INIT or a Subsystem Clear SCLR 7 8 Programming Error PGE 10 Programming Error is a read only error bit that is set if any controller register is written bits for CCLR and SCLR excepted while the GO bit in RKCS1 is set The error is disabled in diagnostic mode Nonexistent Memory NEM Bit 11 Nonexistent Memory is a read only error bit that is set when the controller attempts to execute a bus cycle and SSYN is not returned within 10 us of the assertion of MSYN Nonexistent Drive NED Bit 12 Nonexistent Drive is a read only error bit that is set to indicate the following e A Select Acknowledge SACK signal from a selected drive has not been asserted on the inter face in response to a message other than a Release sent to the drive A complete status message 3 B0 B3 was not received from a selected drive within 20 us of a message request from the controller Unibus Parity Error UPE Bit 13 Unibus Parity Error is a read only error bit that is set to indicate that a parity error has been detected in a 16 bit data word format from main memory during the performance of a Write or Write Check command During 18 bit data word transfers a Unibus parity error cannot be detected
145. rror Type Control Control Control Control Cartridge Hard Soft Seek Control Hard Soft Control Control Control Soft Control Hard Control Seek 5 1 22 Hard Errors any failure to read data correctly after a complete recovery sequence with ECC enabled constitutes an irrecoverable i e hard error Errors that are not ECC correctable include 1 bursts greater than bits in length and 2 isolated dropped bits that are separated by more than 11 bits within a sector The allowable error rate for hard errors is one error per 10 bits read 5 1 3 Soft Errors A soft error is defined as any read error that is correctable by ECC and or a complete recovery sequence The allowable error rate for soft errors is one error per 10 bits read 5 1 4 Cartridge Errors A cartridge error caused by imperfections in the recording surface is regarded as cartridge attributable or media dependent If the imperfection is less than 11 bits long it is ECC correctable and will appear as a soft error If longer than 11 bits it will appear as a hard error On a given cartridge attributable errors will always appear at the same cylinder sector and track addresses The definitions of hard and soft errors see 5 1 2 and 5 1 3 apply only to randomly distributed errors and do not take into account errors that are cartridge attributable Imperfections in the cartridge surface may be found by mapping the cartridge by mean
146. rtridge may not be used on an RK06 Drive 6 3 1 RKO06K and RK07K Cartridge Loading 1 Depress the drive lid release bar and raise the lid Figure 6 2 2 Prepare an RKO6K or an RKO7K Cartridge Figure 6 3 for loading as follows a Lift the cartridge by grasping the top cover handle with the right hand Figure 6 4 b Support the cartridge with the left hand by holding the protection cover handle Figure c Lower the top cover handle and push the handle slide to the left with the thumb of the right hand Again raise the handle to its full upright position to release the protection cover Figures 6 4 and 6 5 3 Lift the cartridge from the protection cover and place it in the drive shroud with the top cover handle recess facing the rear of the machine Figure 6 6 4 Rotate the top cover handle a few degrees clockwise and counterclockwise to ensure that the shroud locating studs are properly seated within the cartridge housing detent slots 5 Gently lower the top cover handle to a horizontal position to engage the drive spindle and place the protection cover over the top cover Figures 6 7 and 6 8 6 Close the lid 7 Depress both of ACCESS ENABLE switches depending capability 8 If write protection is required press the WRITE PROT switch 9 Press the RUN STOP switch RUN If the lid is properly closed and no drive errors exist FAULT extinguished the spindle will turn a
147. s After the pair of words identifying the location of the last bad sector the remainder of the data field is ones filled 2 5 6 2 Track Format The bad sector file format consists of the same preambles gaps header ECC and postamble as those on any other 22 sector track The data field of sectors 0 2 4 6 and 8 will contain the addresses of all sectors identified as defective in 22 sector 16 bit format and the data field of sectors 1 3 5 7 and 9 will contain the addresses of all sectors identified as defective in 20 sector 18 bit format Sectors 10 through 21 will have words 0 through 3 of the data field record as specified in 2 5 6 1 Words 4 through 255 will be ones filled At least two even sectors and two odd sectors out of the first 10 sectors must be error free and two even and two odd sectors out of the last 12 sectors must be error free 2 15 3 SITE CONSIDERATIONS AND PREPARATIONS 3 1 ENVIRONMENTAL CONSIDERATIONS 3 1 1 General The RK611 RK06 or RK611 RK07 Disk subsystem is designed to operate in a light industry or a business environment 3 1 2 Temperature The operating temperature range of either Disk Subsystem is from 10 to 40 C 50 to 104 maximum gradient is 30 F per hour The nonoperating temperature range is from 40 C to 66 C 40 F to 151 F The disk cartridge operating temperature range is from 15 6 C to 48 9 C 60 F to 120 F 3 1 3 Relative Humidity
148. s each of which has a diameter of approximately 355 mm 14 in and a height of approx imately 101 mm 4 in No more than five cartridges should be placed in a single stack 3 2 2 Cabling Each RK06 and RK07 Disk Drive is equipped with a molded line power cord with a length of 2 7 m 9 ft 3 2 3 Power Requirements Both 60 0 5 Hz and 50 0 5 Hz drives operate from 90 128 Vac 180 264 Vac power sources at a power factor greater than 0 80 While operating at 60 Hz the RK06 input power is 450 W nominal and 500 W maximum While operating at 50 Hz the input power is 500 W nominal and 550 W maximum While operating at 60 Hz the input power of the RK07 is 500 W nominal and 550 W maximum or 550 W nominal and 600 W maximum while operating at 50 Hz Receptacles that will accept the 90 128 Vac plugs are designated variously as DEC 12 05351 NEMA 5 15 Hubbell 5000 M9 and General Electric 4050 IT A Receptacles that will accept the 180 264 Vac plugs are designated variously as DEC 12 11204 NEMA 6 15R Hubbell 566251 and General Electric 4092 ITA Appropriate circuit breakers are also necessary Digital Equipment Corporation must be notified of available input power well in advance of shipment so that the proper drive may be shipped 3 2 4 Floor Loading The weight of the RK06 or RK07 148 kg 325 Ib alone is not sufficient to place unusual stress on most office building or industrial plant floors However the added weig
149. s imposed by the Unibus to controller disk commands concluding with data formatting requirements and data error control oper ations 2 5 1 Controller Drive Interface The following information describes the operations performed by the controller interface lines Figure 2 6 that are common to all of the drives available to the system READ WRITE DATA MESSAGE LINE MESSAGE LINE STROBE WRITE GATE POLL DRIVE 20 SELECT ACKNOWLEDGE SACK INDEX SECTOR PULSES WRITE CLOCK POLLED ATTENTION RK611 DISK CONTROLLER UNIBUS INTERFACE CONTROLLER DRIVE INTERFACE LINES 11 4197 Figure 2 6 Interface Lines Message A Line This line transmits from controller to drive a drive selection code as the initial 3 bit portion of a 16 bit serial message the remainder of which includes additional operational information e g Seek Start Spindle Head Select Coding etc for the desired drive Figure 2 7a Notice in the Figure that the control clock timing 15 reflects the availability of each one of the 16 bits of the message relation to time With these timing considerations desired drive is selected by the initial coding TO T2 of the serial format When drive selection recognition is achieved the remaining drives ignore the remainder of the message T3 T 14 When the operational information has been assimilated by the selected drive one of four possible 16 bit status messages can be r
150. s of the formatter program 5 1 5 Seek Errors Any positioning operation that terminates with seek incomplete or with the positioner in an incorrect location constitutes a seek error The allowable error rate for seek errors is one error per 106 seek operations 5 22 DIAGNOSTIC TESTING When all installation procedures have been completed the tests detailed in the following subsections should be performed Table 5 2 briefly describes the diagnostic programs used in field acceptance testing of the RK611 RK06 or RK611 RKO07 Disk subsystem The total time required to run these tests except DZR6Q in the absence of failures is 2 5 hours per drive 5 2 1 RK611 Controller 5 2 1 1 General In diagnostic testing the RK611 is tested first All five of the diskless controller diagnostics Table 5 2 DZR6A DZR6E and the RK611 functional controller diagnostic DZR6K in the Table must run for two error free passes before the remainder of the acceptance program can be carried out Any failure detected by any of the diagnostics must be corrected and all programs must be rerun before proceeding complete recovery sequence consists of 28 retries 16 of which at the nominal head position and 2 each at offsets of 400 800 and 1200 microinches for the RK06 200 400 and 600 microinches for the 7 respectively 2 run time of this diagnostic depends on the number of drives being tested 5 3 Table 5 2 Diagnosti
151. sk command information including the status message request numeric of 0 1 2 or 3 along with an appropriate odd parity bit from various device registers and command logic within the controller As the bits are assembled in the register they are serially transmitted to the drives in coincidence with the transmission of Message A with each bit position 0 15 directly related to an equivalent transmission time TO T 15 For additional information refer to Paragraph 2 5 1 Drive to Controller Message B0 B3 For drive to controller transmissions Figure 7 18 the 16 bit Message B shift register assembles status information along with an appropriate odd parity bit from various status and error registers within the drive As the bits are serially transmitted from the drive in coincidence with the Message A status bits they are assembled in the register with each bit position relating directly to an equivalent transmission time 7 20 Since only of four status messages be stored Message shift register at a time desired message B1 B3 other than BO must be previously defined by the execution of a Select Drive com mand NOTE Maintenance registers 2 RKMR2 and 3 RKMR3 only contain valid status messages following the exe cution of a Select command 73 RK611 DISK COMMANDS Disk commands are divided into two groups One group nondata handling is concerned with the various Operational requirements of
152. ss Register RKDA 7 6 Control Status Register 2 RKCS2 7 8 Drive Status Register RK DS Yee 7 10 Drive Error Register RKER a a 7 12 Attention Summary Offset Register RKAS OFPF 7 15 Desired Cylinder Register RK DC 7 16 Data Buffer Register RKDB 7 17 Maintenance Register 7 17 ECC Position 5 2 2202 7 19 ECC Pattern Register RKECPT 7 19 Maintenance Register 2 2 7 20 Maintenance Register 3 RKMR2J3 7 20 RK6LEDISK COMMANDS u u Nusu a 7 21 DIAGNOSTICMODE DMDE tede d one ad uu tu I 7 29 R ad TT 7 32 WHte SIBIIIAD OB s uu a aysan a uta 7 33 ERROR DETECTION AND 7 34 Correctable Error 7 36 Uncorrectable Error 22 2 7 37 PROGRAMMING EXAMPLES nana Ea a eren 7 38 RK06 Device Driver Routine tee eec e e andes Eure b d 7 38 ECC Correct
153. stem unit in an expansion box Power requirements 5 5 15A 15 590 175 mA 15 Vdc 5 400 mA Number of drives controllers 8 max No Unibus addressable registers 16 Device base address 777440 octal switch selectable Device interrupt vector 000210 octal switch selectable NOTE The RK611 Device base address overlaps the Look Ahead Register in the RC11 Disk subsystem and the RK611 Device Interrupt Vector overlaps the same vector in the RC11 subsystem Interrupt priority Small peripheral controller options Data transfer rates Unibus to Controller average word rate Drive to Controller burst rate Time to update drive status Time to update drive attention Error detection correction Time for correction Maintainability Number of sectors track Level 5 plug selectable Up to three slots one of which must be quad height Other two can be quad or hex height 22 Sector 20 Sector 16 bit 18 bit data words data words 4 3 us word 4 6 us word 3 7 us word 4 2 us word 15 us maximum 7 4 us 32 bit ECC word sector detection 11 bit ECC word sector correction Less than one revolution Diagnostic mode places all controller timing and data paths under software control Software programmable for 20 or 22 sector format 1 5 Table 1 2 RK06 Disk Drive Performance Specifications Characteristics Storage type Medium Disk diameter Magnetic heads Recording
154. t to the Data Shift Register for conversion to parallel data The parallel data is then sent to the Silo for eventual transfer to the Unibus Figure 2 5 PARALLEL DATA OUTPUT HEADER BUFFER COMPARE 1 DATA LOGIC SERIAL DATA DATA SHIFT REGISTER PARALLEL TO SERIAL AND SERIAL TO PARALLEL SILO IC 64 DATA WORDS ENCODER SERIAL DATA 8 SEPARATOR LOGIC BUFFER 1 DATA WD PARALLEL DATA 11 4196 Figure 2 5 RK611 Data Interface 2 7 2 5 RK611 RK06 AND RK611 RK07 SUBSYSTEMS The basic RK06 RKO7 Disk Drive subsystem consists of an RK 611 Disk Controller and up to eight RK06 or RKO7 Drives that can be connected to the subsystem via a daisy chain bus arrangement The controller interfaces to the system via a PDP 11 Unibus However if the optional dual access RK06 or Drives are installed an expanded subsystem can be configured With the expanded subsystem dual controllers A and B are installed which allow each drive to be accessed by either controller via appropriate dual port selection A or B When two controllers are used in this manner they may be connected to the same Unibus in a single processor system or each controller may be connected to a separate Unibus in a dual processor system Figure 1 1 The following paragraphs describe 06 or RKO7 subsystem operations in relation to the functions performed by the controller drive interface lines 17 under directive
155. ted for a drive in which the heads have already been loaded a controller interrupt can be returned on completion of the command but an attention interrupt cannot be returned Unload Command If an Unload command is initiated for a drive in which the heads have already been unloaded a controller interrupt can be returned on completion of the command but an attention interrupt cannot be returned Drive Error Unload Reload When a drive is in the RUN state and the heads unload due to a drive error refer to Paragraph 7 2 6 Current Drive Attention Bit 14 a successful error clear will result in a reload of the heads and the generation of two interrupts the first interrupt will occur with the drive error and the second interrupt will occur up to 6 seconds later when the heads reload Enabling an Interrupt The Interrupt Enable IE bit bit 6 in RKCS1 must be set prior to the initiation of an operation or an interrupt will not be generated Simultaneous Interrupts If interrupt conditions simultaneously occur in the controller due to the overlapping of an operational and data transfer command e g Seek done and Read Data done the data transfer interrupt will take precedence and the operational interrupt will not be processed until another operation is completed Therefore to ensure that no interrupts are lost a program should scan all interrupt indicators in conjunction with the recognition of an interrupt Drive Ready Requirement
156. the drive while the second group data or header handling is concerned with the transfer of data or header information to or from a selected device Refer to Table 7 2 The controller recognizes a command by the configuration of the 4 bit command code F4 F1 that is loaded into RKCS1 However the command will not be decoded for execution until bit position zero GO of the register is set In addition two other 51 bit positions are significant to command flow The Ready RDY bit bit 7 is set when the execution of a command is completed and the combination Error Clear CERR CCLR bit bit 15 will be set if a device or controller error occurs during execution With these considerations and the simultaneous serialization of Message A and Message B Figure 2 8a using input derived from additional device registers the disk commands can initiate the operations shown in Table 7 2 15 13 12 11 10 09 08 07 06 05 04 02 O 00 DRIVE SPINDLE OFFSET DRIVE VOLUME SELECTED PARITYSTATUSposiT write formar TYPE prive VALID orive DRIVE CHANGE IN LOCK 1 20 SECT READY AVAIL ADDRESS PROGR 0 22 SECT 15 4 13 07 06 05 02 0 00 A1 UNLDG HEADS SERVO PARITY HEADS LOADING HOME HEADS SIGNAL SELECTED 2 DOOR HOME PRESENT DRIVE OK LATCHED ADDRESS 15 14 07 00 A2 Tuc p CYLINDER DIFFERENCE OFFSET POSITION SELECTED PARITY DRIVE ADDRESS 15 14 13 12 n 10 09 08 06 05 04 03
157. the drive is an 06 or the bottom if the drive 15 RKO7 If the M9016 module does not have these resistors the serial number must be wired to the back plane To do this a Setup a table similar to this Table 4 4 Sample Table for Backplane Wiring Backplane Pin Assignments b Fora serial number ABC enter the BCD equivalent of digit A in Row 1 Column enter the BCD equivalent of digit B in Row 1 Column B enter the BCD equivalent of digit C in Row 1 Column C c Enter the complement of Row in Row 2 d For every 1 in Row 1 wire the appropriate backplane pin below Row 1 and Row 2 to D6H1 3 V For every 1 in Row 2 wire the appropriate pin to A6T1 ground the drive is an RK06 wire 01V1 to ground If the drive is an RK07 wire 01V1 to 3 V If the installation is multidrive proceed to step 8 otherwise proceed to step 10 Connect drive bus cable PN 70 12292 xx from J2 A OUT of the previous drive to J1 A IN of the next drive Repeat for each additional drive The standard length between drives is 2 44 m 8 ft however optional length cables are available Total drive bus length must not exceed 30 48 m 100 ft Daisy chain each drive using a ground wire PN 7212827 8 between the ground studs in the drive cabinets 000000 SERIAL UNIT SELECT PLUG RKO06 XA OR 7 115 VAC 60 HZ RKO6 XB OR RKO7 XB 230 VAC 60 HZ RKO6 XC OR RKO7
158. the protection cover is removed for loading ensure that the disks are not moved or rotated since improper disk motion may generate plastic particles which can result in disk damage When loading or unloading an RK06 or RK07 Drive insert and remove cartridges gently In addition do not use excessive force when manipulating the top cover handle If during operation a cartridge makes rumbling or continuous tinging sounds or errors occur discontinue use of the cartridge Use of a damaged cartridge on other drives may damage the drives resulting in additional damage to all other cartridges used thereafter Each cartridge should be professionally cleaned every 6 months and or whenever a specific cartridge is not operating properly Cartridges are factory repairable only Disassembly in the field is not permitted and such action will void the warranty on a cartridge as well as any drive on which the cartridge may be operated 8 3 9 RK06 RK07 DUAL ACCESS OPTION 9 1 RK611 RK06 and RK611 RK07 DUAL ACCESS SYSTEM CONFIGURATION Dual access operation is an option available to the RK06 Drive enabling it to be accessed through two different controllers Figure 9 1 shows the system configuration for dual port operation The eight 06 RK07s shown in Figure 9 1 are daisy chained The two RK611 Controllers can be attached to the same or to two different CPUs CENTRAL MAIN PROCESSOR MEMORY UNIBUS oP
159. tics of the drive in the field this requires such parts substitution as the power supply transformer and spindle motor pulley plus a belt change Changing of voltage range 120 Vac to 240 Vac or vice versa merely requires a circuit breaker change at the rear of the drive 4 25 4 6 DUAL ACCESS DRIVE INSTALLATION AND ACCEPTANCE 4 6 1 Dual Access Installation Procedure Site preparation and environmental considerations for the RK06 or RKO7 dual access disk drive are the same as those for the single access drives and are discussed in Chapter 3 of this manual The adjustment and installation procedures for the RK06 or RKO7 dual access drive are also identical with those of the single access drive except for the added cables for dual access Refer to the single access installation procedures described earlier in this chapter Figure 4 12 is a dual access cable diagram 4 6 2 Dual Access Acceptance Procedure 4 6 2 1 Dual Access Diagnostic The Dual Access diagnostic DZR6GA X must be run first to establish that a port can seize and release the drive To accomplish this perform the following pro cedures 1 On each drive to be tested the dual access test switch located on module M7730 must be placed in the invert position for the duration of this test Refer to Figure 4 2 or 4 3 for the location of this switch When in the invert position this switch grounds bit 0 of the port B unit select lines and causes each drive to appear as tw
160. ties standard and optional features and specifications Chapter 2 provides separate drive cartridge and controller descriptions concluding with a subsystem overview to familiarize the user with certain hardware and technical characteristics Chapter 3 contains the details to be considered when selecting a location for installation of the subsystems Chapter 4 describes the installation process for the subsystems and the dual port option In stallation should only be performed by qualified DIGITAL field personnel Chapter 5 describes the procedures necessary to demonstrate that the drive and controller are suitable for user acceptance Chapter 6 contains step by step procedures for the operation and control of the subsystems Chapter 7 provides the programmer with general reference material subsystem software re quirements and recommendations to facilitate the design of user programs Chapter 8 provides the user with basic maintenance requirements and procedures Chapter 9 contains a description of the subsystem dual port option including control and operation procedures for systems with this option This manual will have a much higher value to the reader if he is already familiar with PDP 11 proces sors and peripherals 1 2 GENERAL INFORMATION RK611 RK06 Disk Drive subsystem and RK611 RK07 Disk Drive Subsystem manufac tured by Digital Equipment Corporation both consist of an RK611 Disk Controller and up t
161. tions it is possible that the remainder of the data will be recovered in the same offset position It should be understood that for every required reread cycle the Disk Address DA register must be reloaded and GO bit 0 in RKCS1 reasserted Thus a disk revolution is lost when uncorrec table error is initially detected and an additional revolution is lost every time a reread cycle is exe cuted 7 6 PROGRAMMING EXAMPLES The following material provides several examples of RK06 subsystem programming including Device Driver ECC Correction and Maintenance Register Read routines RK07 subsystem programming is done in a similar manner 7 6 1 06 Device Driver Routine The 06 Device Driver routine allows a user to establish communications with a device and deter mine subsystem status and availability TITLE RK06 DEVICE DRIVER CALLING SEQUENCE JSR PC RK06 DRIVE UNAVAILABLE RETURN NORMAL RETURN INPUTS UNIT DESIRED UNIT NUMBER IN BITS 0 2 DSKADR TRACK SECTOR ADDRESS BUSADR LOW ORDER 16 BITS OF UNIBUS ADDRESS WCNT TWO S COMPLEMENT WORD COUNT CYLADR DESIRED CYLINDER ADDRESS FUNCTN DESIRED FUNCTION IE 16 17 OFFSET OFFSET VALUE FOR OFFSET FUNCTION OUTPUTS RK6ACT SET IF RK06 IS ACTIVE 7 38 81 RKCS2 RKDS RKER RKDC RKOF PACACK SELDRV SCLR SVAL DRDY VV DRA UNS CERR DI IE 177440 RKCS1 10 12 1
162. tor release pin Inspect the absolute filter area for proper attachment of hoses Clean the foam prefilter if necessary Check the head cam area to ensure that the head arms are engaged properly on the head cam If any problems are encountered during execution of the above procedures refer to the RK06 RKO7 Disk Subsystem Service Manual for corrective procedures Close the hinged power supply and card nest assemblies and replace all access covers Remove both bolts securing the cabinet to the shipping skid Remove the skid blocks and position the cabinet leveling feet so that the cabinet rests on its casters Remove the cabinet from the shipping skid using a ramp or other suitable device At this point there are no other shipping restraints for the drive WARNING Use care in moving the drive cabinet 4 2 FULL TELESCOPE 9905446 5 PANEL FOLDER 9905976 CRATING SLAT 7606868 CUSHIONED SHIPPING SKID 11 4979 Figure 4 1 06 07 Shipping Package Configuration 4 3 TRACK POSITION SENSOR M7708 SERVO ANALOG PCB M7729 LD REV FWD 20 21 HDS SET S SPEED DUAL PORT DRIVE CONTROL LOGIC 7730 SEE NOTE 2 PORT B ADDR INV SW SEIZED REQUESTED INTERFACE amp TIMING LOGIC M7706 SEE NOTE 1 DRIVE SELECT A INTERFACE amp TIMING LOGIC M7706 SEE NOTE 1 DRIVE SELECT B CABLE BOARD NOTES 1 TWO
163. tuator plug contains a single number 0 7 while the other side consists of raised plastic cam actuators that are separately configured to encode the plug with the number With this arrangement the insertion of a plug into any UNIT SELECT switch assembly will automatically encode the associated plug number when power is applied to the drive When the heads of a drive are settled not in motion the numbered indicator on the UNIT SELECT switch assembly associated with the drive is illuminated READY 6 2 4 FAULT Indicator The FAULT indicator is illuminated whenever one of the following error conditions is detected in the drive More than one drive selected Positioner when detented has moved too far e g the drive has been jarred A parity error has occurred in a message transmitted from the controller to the drive e A read write unsafe condition in the drive e g serve track error signals write gate signal errors etc A write lock error condition 1 the receipt of a Write Gate signal when the drive is in Write Protect mode Low ac voltage in the drive An incomplete Seek operation The receipt of Write Gate or Seek signals whil Volume Valid is reset 6 2 5 WRITE PROT Switch with PROTECT Indicator If a Write Gate signal from the controller is not currently asserted in a selected drive initial depression of the WRITE PROT pushbutton will immediately set the drive in Write Protect mode and the PROTECT
164. ual access option Port A and 6 2 7 1 Single Access Drive B If B port is configured for single access 1 the M7706 Interface and Timing module is in Slot 2 of the drive card cage and the switch is initially pressed bidirectional communication with the controller is enabled via Port B Under these conditions when the drive is selected by the controller the SELECT B indicator is illuminated When the switch is again pressed released access by the controller is disabled and the SELECT B lamp cannot be illuminated 6 2 7 2 Dual Access Drive A and B If the ACCESS B ENABLE switch is independently pressed on a dual access drive the operations performed are similar to those previously described However if both the ACCESS B and ACCESS A switches are initially pressed arbitration logic will determine on a priority basis which port B or A will be accessed by its associated controller B or A 6 3 OPERATING PROCEDURES This material describes RKO6K and RKO7K Cartridge loading and cycle up procedures that are re quired to place an RK06 or an RK07 Disk Drive on line The procedure assumes that ac power is available red indicator lamp on the rear panel circuit breaker is on the ac circuit breaker is on blower motor is energized and the STOP indicator lamp is illuminated on the control panel If these conditions are met initiate the following NOTE RK06K Cartridge may not be used an RK07 Drive and an RK07K Ca
165. uals zero However if the word count does not equal zero after the transfer of the last sector 19 or 2010 on the last track 2 of the last cylinder 41010 for an RK06 or 814 0 for an RK07 Cylinder Overflow Error COE will occur Spares Bits 11 15 Spares bits 11 15 are 0 7 7 7 2 5 Control Status Register 2 RKCS2 The RKCS2 register Figure 7 6 can be read or written via program control and is used to store the current drive select code subsystem operational status and Silo control information In addition the register can initiate a Subsystem Clear SCLR operation 07 06 15 14 13 12 11 10 09 08 05 04 03 01 MM Y READ ONLY RO 11 4204 Figure 7 6 Control Status Register 2 Drive Select 050 052 Bits 0 2 The Drive Select bits are configured 0 7 with the unit number of the drive to be currently selected Release RLS Bit 3 When the RLS bit is set and a command is initiated the drive specified by 080 2 will be released The bit is used in dual access configurations to effect the release of a drive from one controller and make it available to the other controller This is necessary because once a drive is accessed by a specific controller it is not available to the other unless one second has elapsed since the first controller accessed the drive Bus Address Increment Inhibit BAI Bit 4 When the BAI bit is set via program control the RKBA register is prevented from
166. uld be ON Note that this vector overlaps the interrupt vector in the Disk Subsystem 4 Set the switch configuration for the desired Unibus address and interrupt vector into M7900 module Use Table 4 2 as a guide when making these settings 5 If the controller is to be connected to a PDP 11 40 system remove the M7234 from the processor Remove W5 the M7234 and replace the module in the backplane This jumper must be removed to successfully run the RK06 or RKO7 diagnostics 6 Insert the modules into the RK611 backplane using Figure 4 4 as a guide 7 Attach the transition connector assembly PN 70 12415 0 0 to the cabinet containing the RK611 using the nuts and bolts provided 8 Connect the flat cable PN BC06R 06 from the M7904 module to the Berg connector side of the transition connector Refer to Figure 4 5 for placement of the striped edge of the cable 9 If the RK06 RKO7 subsystem is to be dual ported repeat this procedure for the second controller NOTE Check for a different address and vector on the sec ond controller 4 7 D E F UNIBUS IN G727 6727 7903 7902 7901 UNIBUS OUT 11 4981 Figure 4 4 RK611 Controller Backplane from Module Side 4 8 RK06 OR RK07 DRIVE LEFT CAB UPRIGHT REF ALTERNATE TRANSITION BRACKET MOUNTING 7 POSITION SEE NOTE 3 A ST
167. um altitude specified for the RK06 or is 2440 m 8000 ft However maximum allowable operating temperatures are reduced by a factor of 1 8 C 1000 m 1 F 1000 ft for operation at high altitude sites Thus the maximum allowable operating temperature at 2440 m 8000 ft for example would be reduced to 28 C 82 F 3 1 3 1 7 Radiated Emissions The effects of radiated emissions can be reduced by 1 Grounding window screens and other large metal surfaces 2 Shielding interconnection cables with grounded shields 3 Providing additional grounding to the system cabinets 4 extreme radiation environments providing a grounded cage for the system 3 1 8 Vibration Mechanical Shock Attitude The 06 07 is designed to operate within specifications in a moderately controlled environ ment such as an office or computer facility during the following sine vibration test applied once in each of three orthogonal axes sweep rate 1 octave min 50 60 Hz 0 002 inch double amplitude DA displacement 50 5000 Hz 0 25 G pk 500 50Hz 0 25 G pk 50 5 Hz 0 002 inch DA It is also designed to operate while a half sine shock pulse of 3 G pk and 10 3 ms duration is applied once in either direction of three orthogonal axes 3 pulses total Performance of RK06 or RKO7 will not be affected by an attitude where the maximum pitch and maximum roll do not exceed 5 percent 3 1 9 Cleanliness Although cleanliness is import
168. ume Valid bit set If a port issues the Unload Heads command the one second auto deselect timer is disabled This has been instituted to allow a change of cartridge and not have the drive return to the programmable state If the drive were allowed to return to the programmable state after an unload command the other could seize the drive and issue a conflicting command Once the drive is cycled up the completion of the head loading sequence asserts the attention bit of the port having access to the drive For one additional second the drive will be seized by that port The port having access must issue a command to the drive within this time Failing to issue a command within this time will cause the dual port logic to examine the other port s Port Request flip flop If the flip flop is not set the drive will return to the neutral state If the flip flop is set the drive will become seized by the other port and the other port s attention will be asserted If the drive is unavailable to a port and the RUN STOP switch is changed the port is notified by the assertion of its attention The change is reported to the controller though the spindle on status bit When Write Protect is asserted neither port will be able to do a write operation There is no provision to Write Protect the drive through one port and not have it Write Protect through the other If the drive is seized to a port and the operator changes the status of the write
169. unavailable drive In addition to clearing bit 0 in RKDS a port attempting to seize an unavailable drive will set the port request flip flop in that drive The dual port logic checks the status of this flip flop when the controller currently operating on the drive releases the drive If the port request flip flop is not set this indicates that the port other than the one just in use did not attempt to seize the drive The drive then returns to the neutral state and it may then be seized by either port If however the port request flip flop is set this means the other port attempted to seize the drive when it was unavailable In this case the other port seizes the drive For example assume Port A controls the drive Port B tries to seize the drive and finds the drive unavailable Thus the Port Request flip flop is set When Port A is finished with the drive it releases the drive with a release command Since the Port Request flip flop is set Port B then seizes the drive 9 3 9 4 3 Caution in Issuing a Release Command to an Unavailable Drive The programmer should be careful when designing software for use with the dual port option to avoid the following problem Assume Port A has seized the drive and Port B attempts to seize it and finds it unavailable Suppose Port B then attempts to cancel its port request If the release command from Port B is completed before Port A actually releases the drive there is no problem If however Port
170. unction is a read only bit that is set Message 0 8 to indicate that a Seek command a Write Gate signal has been received by the selected drive while Volume Valid was reset The occurrence of this fault allows the CDA bit bit 14 to be set in RKDS the Attention signal to be asserted for RK AS OF the DSC bit to be set in RK MR2 and the Fault bit to be set in RKMR3 Control to Drive Parity Error DRPAR Bit 3 Control to Drive Parity Error is a read only bit that is set Message 9 to indicate that the drive detected a parity error on controller to drive mes sage transmission The occurrence of this fault allows the CDA bit bit 14 to be set in RKDS the Attention signal to be asserted for RKAS OF the DSC bit to be set in RK MR2 and the Fault bit to be set in RKMR3 Format Error FMTE Bit 4 Format Error is a read only bit that is set when the format status bit Message 0 9 returned from the selected drive does not compare with the CFMT bit 12 in RKCSI Drive Type Error DTYE Bit 5 Drive Type Error is a read only bit that is set when the drive type status bit Message 0 8 returned from the selected drive does not compare with bit bit 10 in RKCSI reset for 06 or set for RK07 Error Correction Hard Bit 6 Error Correction Hard is a read only bit that is set to indicate that a data error detected by the Error Correction Code ECC logic in the controller cannot be c
171. until the word count reaches zero or until a failure to compare occurs If the data fails to compare the command is termi nated immediately Figure 7 28 7 4 DIAGNOSTIC MODE DMD When Diagnostic mode is entered under program control the drive interface lines are effec tively disconnected from the controller while the Unibus interface remains unchanged Under these conditions read write commands i e Read Write Read Header Write Header Write Check can be initiated using simulated data transfer techniques while actual transmissions to and from the drives are disabled In DMD read write commands are decoded normally and messages are assem bled in the shift registers for serial transmission However the output of each shift register is redirected to provide an alternate input to the opposite register resulting in a message swap at the conclusion of a simultaneous transmission Thus message transmission to a drive is simulated by a message swap between shift register A RKMR2 and shift register B RKMR3 Diagnostic mode is entered by the setting of the DMD bit in Maintenance register 1 MRI and subsequent operations are controlled and monitored by additional bits within the register Data trans fers are simulated by the continuous setting and resetting of the control bits 1 MCLK MIND MSP MERD PAT in a predetermined sequence while the program monitors the read only bits i e MEWD WRTGT RDGT ECCW
172. ure 4 6 and if a discrepancy exists correct it before continuing with the installation Connect a ground wire PN 7212827 25 from the ground stud on the H969 cabinet to the ground stud on the cabinet containing the RK611 Controller Figures 4 8 and 4 9 Lower the card nest assembly to its horizontal service position and install a drive bus cable PN 70 12292 xx between the transition connector PN 70 12415 0 0 of the controller cab inet and I O connector J1 A IN of the RK06 or RKO7 Figure 4 7 and Table 4 3 The standard drive cable length from controller to drive is 7 62 m 25 ft However optional cable lengths can be used if the total drive bus does not exceed 30 48 m 100 ft See Table 4 3 for part numbers Table 4 3 Drive Bus Cables Part No 70 12292 8 2 44 m 8 ft 70 12292 12 3 66 m 12 ft 70 12292 25 7 62 m 25 ft 70 12292 40 12 19 m 40 ft 6 On some drives there are two rows of 12 zero ohm resistors on the M9016 module They represent the last three digits of the drive serial number and must be clipped to give the BCD representation of these three digits If they have not already been clipped remove the top resistor of a pair to represent zero or remove the bottom resistor for a one for each of the 12 bit locations Also there is a pair of zero ohm resistors representing the drive type placed to the left of the serial number resistors If neither has been removed clip the top one if
173. ve the head alignment fixture and mount it on the next head Load the heads and repeat the procedures of steps 8 through 15 until all data heads are aligned 4 4 5 Response to Abnormal Conditions The following procedures are for handling abnormal conditions encountered during a system installa 4 4 5 1 Clearing an Unsafe Condition Clearing of an unsafe condition can be accomplished by releasing the RUN STOP switch on the front panel Press the RUN STOP switch again If the fault no longer exists the FAULT light will go off When the READY light is on the drive is ready for further operations 4 24 4 4 5 2 Removing a Cartridge When Drive Power is Lost If power to the driver is lost and it is necessary to remove the disk cartridge use the following procedure 1 Insert a small flat blade screwdriver into the opening above the door latch 2 While pushing on the screwdriver lift the pack loading door 3 Ensure that the brush drive knob is turned fully clockwise 4 Ensure that the head loading arms are fully retracted 5 Remove the disk cartridge in the usual manner 4 4 5 3 Detecting Head to Disk Interference The read write heads actually fly over the disk surface at an altitude of approximately 50 millionths of an inch When for any reason the read write heads touch the disk surface a condition referred to as head crash damage to either heads or disk surface is possible If there is an audible ticking sound after the
174. will lose its CSA rating The drive is designed for installation as a free standing device 4 1 UNPACKING AND INSPECTION When packaged for shipment the RK06 or 7 in its H969 cabinet weighs approximately 148 kg 325 Ib Although excessively heavy and bulky for single person handling the package does not re quire the use of a forklift or similar equipment for moving or lifting Table 4 1 lists the special tools and test equipment required for the RK611 RK06 or RK611 RKO0 7 Disk subsystem installation CAUTION When moving or lifting the RK06 or RK07 always grasp the frame structure Do NOT hold any part of the top or side covers Table 4 1 Special Tools and Test Equipment for RK611 RK06 or RK611 RK07 Disk Subsystem Installation Part No Part RK6 7 TA Field Test Box 115 Vac 50 or 60 Hz RK6 7 TB Field Test Box 230 Vac 50 or 60 Hz RK06K AC Alignment cartridge RK06 RKO7K AC Alignment cartridge RK07 93 06058 Wrench 93 06240 Glass alignment tool 93 06076 Head installation removal tool 93 06087 Head alignment tool 93 06122 01 Shim 0 003 in 93 06122 02 Shim 0 00015 in 93 06123 01 Shim 0 020 in 93 06123 02 Shim 0 015 in 93 06171 Head separation tool 29 19633 Mirror 29 20065 Magnifier 29 20906 Torque wrench 29 21487 Head torque wrench 29 21832 Hex ball driver set 29 22526 Screwdriver 14 5 in Phillips 93 06056 Carriage roll alignment tool 4 1 The procedure for unpacking disk drive is as f
175. with an image of the 11 bit burst that includes a One bit for every error bit regardless of whether the error bit is an erroneous one or zero Consequently the program increments through the Data field image area in main memory a number of times that is equal to the value in RKECPS and locates the start location of the error burst The error burst is then extracted from memory and exclu sively ORed with the value contained in RKECPT The result including a complement of each of the eight error bits is then returned to the 11 burst locations and the correction is complete One disk revolution is lost as a result of the execution of an ECC correction cycle However the Disk Address RKDA register has been incremented to the next sector Therefore only the GO bit 0 in RKCS1 must be reasserted to resume reading DISK DATA WORD 2 WORD 1 gt 55 2 5 5555 5 4 32 31 50 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 09 11 BIT ERROR BURST ON READ 8 ERRONEOUS BITS PATTERN ECPT REGISTER CONTENT RESULT OF EXCLUSIVE OR CORRECTED CORRECTS ERROR BITS DATA IN MEMORY 15 00 POSITION 0 5 REGISTER 11 4230 Figure 7 31 Example of 8 Bit Error Correction Process 7 36 7 5 2 Uncorrectable Error DCK When uncorrectable data error is indicated
176. y slight mechanical misalignment that may occur between read write heads and a disk cartridge by allowing a head to be positioned about the track centerline in incremental steps Also error detection and correction hard ware ECC provides adequate information for correcting any error burst of up to 11 consecutive bits within a sector For increased reliability parity checking is performed on the control information that is transferred between the drives and the controller In addition the controller also detects and flags data errors from the Unibus Upon the indication of an error condition or the completion of a command the controller can interrupt the processor Thus extensive error indicators exist for on line diagnosis while numerous status indicators provide complete program control On multidrive systems positioning operations can be overlapped 1 Seek operations may be per formed on one or more drives before starting a transfer on another drive Thus while one drive is reading or writing others can be positioning to a new cyclinder for the next transfer The following paragraphs provide detailed descriptions of controller register content and usage fol lowed by disk command descriptions and programming considerations related to the use of an RK06 or RK07 subsystem with a PDP 11 processor 7 1 7 2 RK611 DEVICE REGISTERS There are 15 usable 16 bit device registers contained in the RK611 that are used to interface the co

RK06/RK07 Disk Drive User`s Manual

Contents

Download Pdf Manuals

Related Search

Related Contents