User Manual - Home Page of Alex Faveluke
Contents
1. start new test measure k factor stop managed test Status 0x80 POI CycWin Software Main Panel From this window the user may chose to enter part information stop or start a test 6 3 Create and Edit Parts Part information must be entered before starting a test To do this click on the create edit parts button This will open another screen the create edit parts Window manual sxw 2004 09 24 create edit parts window 5 Registered parts found in database name number description created K factor K offset if720 samp1 first test DUT May 08 l 2 0945m 0 6542V irf 720 samp2 samp2 a May 08 if720 jsamp3 samp3 Thu May 08 if720 jsamp4 samp4 Thu May 08 int720 jsamps samp5 Thu May 08 if720 jsamps samp Thu May 08 if720 samp samp Thu May 08 if720 samps jsamps Thu May 08 i new delete dose window Create Edit Parts Window To create a new part click the new button which will open up a screen allowing the user to enter new parts The Part Name field ought to be dalinn annniio hedas filled in with serial number the Part name o part number should be the lot part number code and part type description A short description must alsobe K factor fknown entered K offset Gf known If the K fa2. Be aware that these quantities may jump through undefined values while the machine is switching between DUTs 3 2 Back Panel Connections A labeled photograph shows the connections on the back panel 354A Back Panel Connections AC LINE IN 230VAC ONLY HEAT SUPPLY The 354A takes power from a Heat Supply on the Power terminal blocks Observe polarity positive is on the right negative on the left Power Supply Control provides an emergency stop function It carries a normally closed signal from the E stop switch to the Heat Supply Circuitry in the Heat Supply manual sxw 2004 09 24 connector end disables the supply if the E stop switch is pushed opened AUX 5V OUT is provided to run accessories requiring 5V power This output is capable of 3 amps but is not fused or protected in any way Do not short this out it will probably cause no harm but will cause the cycler to lose power RS232 Serial connects to a 9 pin female D sub connector on the back panel This provides communications from the 354A to the host computer Note that this connection ties Earth ground between the host computer and the 354A It would be advisable to run the computer and the 354A from the same AC line circuit to avoid ground loop problems AC Line is brought in on a standard IEC line entry The 354A uses auto ranging supplies internally enabling the use of any AC voltage from 110 to 240VAC Note that accompanying the Beta Unit 101 the
3. 5k 16 5k 75k or approximately Vin 0 18 Amplifiers U17 G1 and G2 control the center point voltage on H bridge outputs OUTI and OUT2 respectively Feedback is divided down by resistors R9 24 3k R8 43 2k on OUT giving dividing ratio of approximately 0 36 The op amp acts to equalize its input voltages thus it controls the voltage at OUT1 to be approximately 2 the input voltage to provide equal voltages on pins 3 and 2 after the voltage dividers Transistor Q1 acts as a high voltage amplifier to extend the range of U17 Not shown on the schematic as drawn is a feedback cap from Q1 C to pin 3 of 1000pF this is needed for stability The voltage on center point OUT2 is controlled in the same manners by op amp U17 G2 pins 5 6 7 U17 requires an isolated source of 15V to function this is provided by the DY910 028 01 isolating supply The supply takes 5V in and outputs 8 completely isolated manual sxw 2004 09 24 9 15V sources which actually range from 22V open circuit to around 16V when loaded 910 028 01 isolating supply may take 5V from an auxiliary connector on the back of the 354A unit serial 101 See the Back Panel Connection Drawing for location 4 2 2 Cable Connection to the 910 027 boards The 910 027 01 boards have already been connected to DUT interface cables The connection uses a single 2 position terminal block 4 2 3 Checking the 910 027 boards The 910 027 boards are required for cycling H bridges They may be
4. DUT channel 1 Tag 2 refers to channel 2 Read the instructions check the parts are hooked up to the correct DUT channel card and press the Next gt button The K factor taking window will then show up K factor measurement window 5 z Perform measurements at 3 different temperatures place DUTs in oven and stabilize at approximately 25 60 and 95 degrees C for each case enter the exact temperate and click measure to record K factor data The tester will reject the measurement if Yd isn t stable enough NOTE see kfm_stability in the config file TODO translate into more user friendly wording also work on better layout enter exact temp 25 6 measure approx 25C enter exact temp measure approx 60C enter exact temp measure approx 95C Cancel lt Prev It is critical that all parts selected be at the same temperature this may be accomplished by attaching them all to the same isothermal block a large block of copper that maintains the same temperature across its entire volume A thermocouple should be placed directly on the mounting tab of the device measured It is important to get the temperature of the semiconductor die as accurately and as stable as possible when making K factor measurements The 354A can be no more accurate than the K factor measured manual sxw 2004 09 24 Heat the DUTs together in an oven or a carefully controlled d hotplate Ensure than however they are
5. Heat supply and the Host Computer are configured to take 230VAC so it would be convenient to use the same for the 354A Remember it is very advisable to have the 354A and the host computer plugged into the same AC line circuit 4 354A Accessories 4 1 281 001 01 DUT Cable Assembly The 281 001 01 DUT Cable Assembly is documented on an accompanying drawing which shows the cable flying leads and connector The 354A provides true Kelvin connections to all leads of a power FET DUT For each connection Gate Source and Drain there are two connecting wires Force and Sense Drive current flows through the Force line and voltage is sensed off of the Sense line in every case eliminating voltage drop errors from cable and switching connections This enables very accurate low voltage measurements to be made Each DUT interface module card provides a front panel connector for interfacing with a DUT This connector is a D sub type 7W2 shell with two power pins and 5 signal pins Power pins are used for Heat Supply high current Force to Source negative and Drain positive 4 signal level pins are used two for Drain and Source Sense lines one for Gate force and one for Gate sense The 281 001 01 Cable assembly is made up for conveniently connecting a MOSFET DUT to the 354A cycler It provides a male connector for one end plugging int to the module interface card and flying wire leads for the other which may be soldered or bolted onto the DUT
6. Meter for Iheat calibration shot 1 e Click the read cycle data button and read the I_heat measurement for DUT 8 Enter this in spreadsheet cell C47 f Adjust the heat current limit up to a higher level 10 A works well taking care not to exceed current ratings on the meter and test leads used Repeat steps c and d to fill in spreadsheet row 48 for calibration shot Iheat 2 g The new calibration numbers are in cells B51 and B52 and should be updated into the 354a conf file for calibration Ih_offset and calibration Ih_scale manual sxw 2004 09 24 30 8 Thermal Error Stackup 8 1 Thermal Error Several factors influence the repeatability and accuracy of the 354a What follows is a consideration of the most important errors affecting the Vd measurement which is used to calculate both absolute temperature and temperature swing Assuming the machine has been calibrated and all uncertainty comes from thermal drift properties of the differential amplifier ADC and reference voltage and that the internal operating temperature of the 354A runs 25 C to 50 C a 25 C delta Gain Errors reference error on board controller uC 100ppm 25 C 1 0 0025 ref error gain setting resistor 100ppm C on differential input amplifier nom 26 1k max 26 1k 1 0001 25 26 165k min 26 1 1 0001 25 26 034k gain setting equation is 49 4k Rset 1 therefore gain is nominal 2 89272 max 2 89746 min 2 88800 gain
7. actual cycling and thermal characterization Offset is also measured enabling absolute temperature measurements to be taken K factor and offset are recorded by DUT serial number prompted from the user The system is then able to correlate a measured diode drop Vd with a absolute junction temperature Tj to be used in further testing and data collection The graphical host software provides a routine for the user to make K factor measurements which must be done before cycling or measuring of a part starts manual sxw 2004 09 24 2 3 Thermal Characterization The 354A is capable of making thermal characterizations on a part including effective Thermal Resistance to Sink R j s efective needed for controlled cycling as well as Junction to Case Thermal Resistance Rj which needed for a good data sheet specification The units of both these characteristics is C temperature rise per Watt of power C W Effective Thermal Resistance to Sink is a measurement of junction heat rise mounted on a given heat sink at a given power level over a given amount of time in example A certain TO 247 part is mounted to a 20 C water cooled plate The junction temperature is equalized to 20 C A power pulse of 200 Watts is applied for 5 seconds and the junction temperature is measured immediately after removal of power at 118 C Rj effective 1S then 118 C 20 C 200W or 0 49 C W This value may then be used in future cycling to help deter
8. checked simply by measuring voltage from OUT1 and OUT2 to ground these should be approximately 1 2 the total voltage on the part This voltage may be adjusted by trimming resistor R6 however this is not an easy task 4 3 281 003 01 Debug Switch Assembly The 281 003 01 Debug switch assembly is used for re programming the internal software of the 354A controller unit It consists of a toggle switch and pushbutton switch wired to a small 5 pin pocket header described as followed 281 003 01 Debug cable for 354A cycler Currently this is a free hanging cable to bring out the controller RESET and PSEN signals These are used with a switch and pushbutton to reprogram the controller A shrouded header is heat shrunk to the cable which is cat 5 4 pair wire pin 1 PSEN blue of blue white pin2 DGND white of blue white pin4DGND white of orange white pin 5 RESET orange of orange white The toggle switch is wired to the PSEN signal the pushbutton is wired to the RESET See directions below in Section 5 for flashing the Controller Software 5 Flashing the 354A Controller Software 5 1 Controller Code The 354A controller uses a microcontroller to provide all control and communication This microcontroller is flash programmable over the same serial port used for normal host controller communications Wyeast Technology may send software upgrades with instructions to flash the new version into the controller The contr
9. correct DUT tag channel Parts and position are selected by the pull down menus in the upper left of the panel Chose a heating current enough to bring the part up to good operating manual sxw 2004 09 24 22 temperature This should probably be close to the test current the part runs in cycle testing to give the part a swing of 100C Pulse duration is limited to 12 seconds at this point There is no reason to use a shorter time than the default12 seconds and it may impact accuracy to do so Set a thermocouple thermometer to measure maximum temperature on both Tc and Ts thermocouples in C The maximum temperature will occur right before pulse termination which is when the junction temperature is measured by the 354A Click the step 1 heat part button to start the heat pulse The 354A will then time and control a pulse with current slightly under the Approx heat current requested on the panel When the pulse has terminated read the maximum temperatures from the thermometer and fill them in in the boxes provided on the right side of the panel Measurements Enter case temperature celcius aa Enter sink temperature celcius AT manual sxw 2004 09 24 23 Then click the ee button to calculate Rtj c and Rtj s values These may be added to the database by clicking E RO The host software will prompt for some descriptive text to add as a comment to that specific test instance The measurem
10. error due to gain setting resistor is approximately 1 00163 gain set err gain error due to amplifier is 50ppm C or 25 00005 1 0 00125 gain err These gain errors total to a maximum of 1 00539 or 1 0 539 gain With a 100 C swing the gain induced error could be as high as 0 539 C DC accuracy value on the ADC is specified at 3 LSBs A LSB is nominally 2 5V 4096 counts so this leads to an uncertainty of 0 00183V The effect of this is reduced by the differential amplifier s nominal gain of 2 89272 for a diode voltage uncertainty contribution of 0 0006330 Given a k factor of 2 00mV C this leads to a temperature uncertainty of 0 317 C With the above gain calculation the measurement could be said to have error in the range of 0 856 C if all these factors stacked up in the worst possible combination Effects of thermally induced errors can be minimized by allowing the machine to warm up before calibration and use It is also advisiable to measure DUT K factors and offsets on the 354 rather than using other measurements as this will eliminate absolute accuracy issues in the thermal measurement path manual sxw 2004 09 24 31 9 Data Base Structure 9 1 354A System Data Storage Overview During normal cycling operation the Host Computer logs all measured heat quantities voltage and current and thermal measurements The Host Computer calculates actual quantities from the raw data sent by the 354A cyc
11. OI a Open spreadsheet calibration_worksheet sxc using OpenOffice b Open file 354a confina calibration data i calibration vd_offset 0 00872252 text editor such as vi or kwrite calibration vd scale 0 00040683 calibration Vh_offset 0 38718928 calibration Vh_scale 0 01531827 calibration Ih_offset 0 110733886 calibration Ih_scale 0 00644809 c Copy the existing calibration numbers calibration Vd_offset and calibration Vd_scale into the spreadsheet cells labeled Vd_offset and Vd_scale cells B6 and B7 Locations for user input are denoted by a manual sxw 2004 09 24 26 B E D E F G 1 34a Hnd Calibration 2 3 dD FowardDrop Calibration 4 5_ Existing Ca from 3542 001 6 lla afer eee Twas Paoa 8 a Atua tage DUTBmeas I DOLTBmeas 2 DUTB meas OLB avg fawiton ADC anple readifom meter from force meas fon farce measfonm farce meas calailatad calauiatea 027000667 644 60375751 _asesz3 12027552364 Calibration Worksheet Vd section d Attach the voltmeter leads to DUT 8 Source and Drain leads If the meter is properly calibrated it is not necessary to orient these leads to always read positive voltage just enter positive voltages only in the Calibration Worksheet e Clip the Schottky diode in parallel with DUT 8 The diode s cathode is connected to DUT8 s Drain lead the diode s anode is connected to DUT8 s Source
12. Wyeast Technology 354A FET Cycling Control User Manual manual sxw 2004 09 24 1 Introduction Power cycle testing is an important measure of long term reliability Power cycling repetitively stresses parts by thermal heating and cooling straining solder connections and other package features In many cases a certain temperature swing is given such as Tj swing of 100 C It is not an easy task to control and measure junction temperature swing with general purpose test instrumentation Even if the part is characterized correctly before testing testing itself changes thermal parameters and devices must be re characterized often during a test 2 Basic Features of the 354A The 354A is designed to outperform current in house designed cycling systems by addressing the problems of interface change and initial performance differences noted above The key feature of the 354A is the combination of a cycling controller and thermal characteristic measuring instrument After some initial measurements on the parts under test the instrument is able to accurately read and record every devices junction temperature before and after each and every power cycle This provides a running record of thermal interface performance the primary figure of merit affected by power cycle testing Even better the system is able to dynamically adjust power levels for each device to maintain a controlled temperature swing This power level adjustment is done individuall
13. ated in the main window with the edit create project button time the cycling test started time the cycling test ended not implemented yet Columns of table tests manual sxw 2004 09 24 33 project ID number ref d by fk_projects in other tables name project name The projects table may store more detailed information in the future Columns of table projects 9 3 As Shipped Database information If a user is fluent in SQL database queries they may find it useful to directly pull data from the database This can be a very fast way of determining the condition of parts currently under test as well as sorting out and selecting subsets of data Wyeast Technology does not recommend changing table structure or adding data to the database independent of the normal host control software Host system accompanying 354A serial number was shipped with host system passwords system user cycler MySQL root password MySQL user cycler password manual sxw 2004 09 24 manual sxw 34 2004 09 24
14. ctor information is Cancel OK known it may be entered If K New Part Definition Window factor data is not known the software provides a routine for the user to determine it using the 354A K factor data must be known before testing of a part This is the only way the machine can determine the before and after temperatures If an error is made in part data entry highlight the corresponding row in the Create Edit Parts window by clicking over it Then use the delete button to remove the row 6 4 Determining K Factor After creating the parts in the previous screen a routine may be run to determine K factor This is selected by the measure k factor button from the main startup window manual sxw 2004 09 24 A K factor instruction window will appear K factor measurement window 5 K factor measurement happens as follows the parts you hawe selected will be hooked up to the tester and placed in an owen stabilize the parts at one temperature and the tester will measure Vd then repeat for a couple more temperatures This text will be rewritten for clarity RJM Hook the following parts up to the tester exactly as specified Tag 1 attach part irf720 samp2 samp2 Cancel lt Prev Next gt K Factor Instruction Window It is critical to hook up the parts exactly as described When the instructions say Tag here they mean DUT channel Thus Tag 1 refers to
15. e process to get a sample of ordinary room temperature DUT diode drop Insert these values into row 11 in the spreadsheet h The new calibration values are in cells B14 and B15 Edit the 354a conf file values calibration Vd_offset and calibration Vd_scale to update these 7 4 Vheat Calibration Procedure Vheat will be calibrated in the same way by measuring at two separate points To get the two points an adjustable Heat Supply must be connected to the 354a This procedure will use the pulse device function to apply voltage but not current to the DUT It is convenient to remain with DUT 8 Leave the voltmeter attached but be sure to check the Schottky diode is removed a Open file 354a conf ina text editor such as vi or kwrite B Existing Ca from Atua tage BOLTS V heat raw for ADC feaedtom meter Fon cvoe chiz cafauiziec 507 00312242 318500153590 c b Copy the existing calibration numbers calibration Vh_offset and calibration Vh_scale into the spreadsheet cells labeled Vh_offset and Vh_scale cells B24 and B25 Locations for user en are denoted by a Calibration Worksheet Vh Section c Inthe Admin Panel right hand side chose select D U T 8 Enter the value 0 zero into the white DAC setting box and click the button write DAC The shell window should respond wrote 0x0 to D U T 8 response 0x0 This will set the current command to 0 where it will not cause the a
16. end These are bundled together in expandable polyester braid Wire colors are Signal Wire Drain Force AWG 12 large Black manual sxw 2004 09 24 Signal Wire Source Force AWG 12 large White Drain Sense AWG 22 small Red Source Sense AWG2 small Blue Gate Force AWG2 small Yellow Gate Sense AWG2 small Violet Note that in operation Force and Sense wires are connected at the DUT Ensure that this wiring is correct before starting testing as misconnections may cause damage to a interface module card or DUT 4 2 910 027 01 H Bridge Balance Adapter Board These boards are documented in schematic 910 027 01 This description will be more clear if the schematic is on hand The 354A Cycle Controller has been designed to power cycle single MOS transistor devices An H bridge is four MOSFETS arranged of two in series to form one leg and two in series to form the other The purpose of the 910 027 adapter boards is to make an H bridge appear like a single transistor to the 354A This is done by equalizing the current through both legs and making the center point of each leg approximately one half the total voltage With 1 2 current and 1 2 voltage on each MOSFET each one takes the full heat power Voltage balancing is done using a single reference generator and two OP Amps R6 and R7 form the reference generator splitting the input voltage by Vout Vin R7 RT R6 16
17. ent will be added to store Rt measurement oO x About to store measurement data into database Please provide a description and press OK description Type a description here Cancel OK the table rt_measurements when OK is clicked manual sxw 2004 09 24 24 7 Calibration Procedure The 354A is calibrated at the factory and user calibration is ordinarily not recommended Mistakes in calibration may lead to incorrect data and improper calibration procedure could possibly damage the Beta unit 354A 7 1 Calibration Concepts The 354A makes three critical measurements e Diode Forward Voltage Vd which corresponds to DUT temperature e Heat Voltage Vheat e Heat Current heat These measurements are made by the central controller using the same analog hardware and analog to digital conversion for each DUT This results in only one calibration being valid for all DUTs it is convenient to do this with DUT 8 especially in calibrating Diode Forward Voltage The hardware is assumed to be linear Two calibration numbers are supplied for each measurement offset and scale The host records data as measurement raw scale offset where raw is the data directly from an analog to digital conversion scale and offset are numbers to convert this into a engineering unit quantity These calibration numbers are determined by taking data under two different test conditions Calibration number
18. erator to select parts to test Selected parts are highlighted by clicking with the mouse and using ctrl click or shift click to add to the selection Ensure the correct parts are selected and hooked up exactly as instructed in the window appearing next All data logging is dependent on the fact that the correct devices are hooked up to the correct DUT channels Start a New Test Highlight parts with mouse to select Select up to 8 parts to test 88 99 anotherpard irf3711_1 3711_1 0220 properl irfp3415_11 freds transis SOM SLFLLX3 3 test part TO SOMBSJVRP_1 1 GOT227 pard Select Parts for Test Window manual sxw 2004 09 24 The software will allow the user to chose which part gets connected to which DUT interface Just click in the Position Number field to change this number You must select a different position for each DUT Start a New Test specified press next Position ay Position l2 gt Position Ed Position ay For the parts you have selected to test choose a Tag number to which you will now hook up the part When all of the parts are hooked up to the tester exactly as attach part itf3711_1 3711_1 o220 properl attach part irfp3415_1 1 dreds transis attach part SOM 5LFLLX3 3 est part TO attach part SOM85JVRP_1 1 GOT227 part Select Position for Test Window Be sure to hook up the parts exactly as listed in this screen Next fill in the inf
19. heated you are able to take a stable measurement at close to the requested temperatures When the part is stabilized at approximately 25 C enter the exact temperature measured in the fill in box on the left and click the measure at approximately 25 C button This will probably be room temperature just ensure that that the parts are well stabilized not warming or cooling and that the measurement typed into the window is the actual temperature when the button is pressed Heat the part and stabilize at approximately 60 C Again enter the measured temperature and click the measure at approximately 60 C button It is not critical what the exact temperature is so long as it is measured accurately and entered correctly Repeat for the 95 C measurement The software will then run a linear regression to determine K factor and offset for the part The information from the k factor determination procedure is saved in the system database Devices of identical size from the same wafer lot generally have very close to the same K factor You may decide to take K factor on a sample of parts only and copy in data part to part using the K factor if known option in the New Part Definition Window manual sxw 2004 09 24 6 5 Starting a Test Once part numbers have been entered and K factors have either been measured or entered a test may be started Select the R from the main window A panel will appear asking the op
20. host control window then chose the desired project and test name A firsttest v 500 Extract of rows from dataset optional est firsttest has 0 cycles message will then appear with the location of the file created You may select the number of rows to go into the comma delimited file For example if there are 5000 cycles taken and the user asks for 500 rows the software will select every 10 row to export 4G Report stored in CSV format in file home tester reports default projecuwatercooled1 csv This report file contains comma delimited data tabulated for each DUT organized in separate columns for each part making it convenient for comparisons between devices under test manual sxw 2004 09 24 6 7 Rtj c Rtj s Testing The 354A may be used for Rtj c and Rtj s measurements on characterized FETs Rtj c and Rtj s measurements require measurements of junction temperature device case temperature and sink temperature Thermal resistance measurements are important for datasheet characterizations Designers use Rtj c and Rtj s numbers to design cooling systems that provide adaquate cooling to a semiconductor in an application 6 7 1 Definition of thermal resistance Thermal resistance can be loosely defined as the opposition of a region to heat flow An analogy to electrical circuits can be drawn with temperature differences substituted for voltage and heat flow power substituted f
21. le values calibration Vh_offset and calibration Vh_scale to update these manual sxw 2004 09 24 29 7 5 Theat Calibration Procedure The Theat calibration is done almost exactly like the Vheat calibration It uses the current limit of the power supply to provide two separate test currents instead of test voltages a Copy the existing calibration numbers calibration Ih_offset and calibration Ih_scale into the spreadsheet cells labeled Ih_offset and Ih_scale cells B43 and B44 b Turn the heat voltage supply down to about 4 volts and set the current limit for approximately 3A It is very important to have the voltage backed down to a low level or the output capacitance of the supply may deliver a damaging spike into the interface module cards c Remove DUT 8 and attach a current meter Current should flow FROM Drain Force into the of the meter out the and back into the Source Force A spare DUT cable is handy for this simply connect the meter to the large white and black heat current force leads In the picture to the right ameter is attached to an open male connector plugged into the interface module It is not necessary to connect any other leads to the interface module card urrent Meter Attached to Drain Force and Source Force C d Observe the meter and click the pulse device button Read the current meter when the heat current has stabilized and record it in spreadsheet cell B47 Actual Current read from
22. lead In this way the diode is effectively in parallel with the internal body diode of DUTS8 f From the Admin Panel click the force measurement button Schottky Diode in Parallel with DUT with voltmeter Measurements will appear in the leads attached shell window Immediately read the attached voltmeter for the actual diode voltage A hold function is usefull here Copy the voltmeter reading into cell B10 It should be much lower than an ordinary diode drop Shell 354A Test Session Edit View Settings Help Copy the last DUT8 measurements from DUT 2 measurement 0 5763 0 5763 0 57631 2 A 7 DUT 3 measurement 0 5730 0 5734 0 5730 the shell window into the spreadsheet cells put 4 measurement 0 5763 0 5758 0 5758 P DUT 5 measurement 0 5730 0 5730 0 5730 C D and E10 In the picture here these DUT 6 measurement 0 5795 0 5803 0 5795 DUT 7 measurement 0 5893 0 5889 0 5889 are 0 2744 0 2744 0 2744 DUT 8 measurement 0 2748 0 2752 0 2744 DUT 1 measurement 1 6747 1 6747 1 6747 DUT 2 measurement 0 5763 0 5763 0 5763 DUT 3 measurement 0 5734 0 5730 0 5726 DUT 4 measurement 0 5763 0 5763 0 5763 DUT 5 measurement 0 5730 0 5730 0 5730 DUT 6 measurement 0 5803 0 5807 0 5803 DUT 7 measurement 0 5893 0 5893 0 5889 8 measurement 0 2744 0 2744 0 2744 Shell output from force data manual sxw 2004 09 24 27 g Remove the paralleled Schottky and repeat th
23. ler using calibration values in the configureation file All data is written to a database currently using MySQL residing on the host machine The host software is also capable of writing to a network accessible Oracle database if the tables are correctly set up 9 2 Description of Tables As shipped 2003 08 28 unit 780102 the database name is dynalink_354a Tables are all logged data for each cycle each part kfactor_measurements K factor measurement storage parts description of part including calculated or entered K factor data projects project name and ID tests description of tests including which DUT channel is connected to which part Tables of Database dynalink_354a T_before temperature before heat pulse in C calculated by Host from diode forward drop using K factor and offset from parts data Columns of table cycledata manual sxw 2004 09 24 32 the part ID number Keys to parts table measured forward diode drop average of 3 measurements sent by 354A temperature that measurement was taken at the date and time the measurement was taken Columns of table kfactor_measurements Columns of table parts fk_parts_dut1 8 columns Each dut column contains the part ID number which references the part ID in the parts table This connects the part names to fk_parts_dut2 the DUT channels fk_parts_dut8 fk_projects which project this test is a member of New projects are cre
24. mine the required power level for a given temperature rise The 354A measures Vd on every device in testing before and after every power cycle The host computer uses recorded K factor and offset to obtain junction temperatures before and after the power cycle occurred If the temperature difference temperature swing is too low power level is increased in subsequent cycles If the swing is too high power level is decreased The host computer records details of every cycle including power level and temperature swing actually recording R j s effective in testing for every DUT for every cycle Internally the 354A contains a precision current source presently fixed at 10mA This current source is used for every Vd measurement Ibe fixed to 10mA All measurements are taken with the same test current including K factor offset and in test temperature measurements 2 4 Power Cycling The 354A automatically controls power cycling Before cycling the graphical host software host machine will allow the user to enter certain parameters including 1 Cycle quantity The number of cycles to run 2 Target swing C Desired temp swing Usually 100 C 3 Max Current Amps The maximum current allowed for heating 4 Thmax C Maximum final junction temperature 5 Tlmax C Maximum starting junction temperature manual sxw 2004 09 24 The host computer will then control the 354A to produce cycling of the given temperature excursion i
25. n the given cycle time If the 354A is not able to heat the part through the requested swing within the maximum current limit testing will terminate with an error Thmax and Tlmax are protection limits to terminate cycling if a DUT becomes overheated say in the case of a fan or cooling system failure During power cycling the host computer will record the cycle count the before and after temperature for each DUT current and voltage applied to the part and atime stamp This information is stored in a mysql database for easy and fast access by the host software manual sxw 2004 09 24 3 354A Connections 3 1 Front Panel Connections 3 1 1 DUT interface cables The main connections on the front panel of the 354A are for the DUT interface cables All connection to the DUTs are made by over this single connector The connection is described in detail in the below with the 281 001 01 DUT cable assembly in Section 4 354A Accessories 3 1 2 Current and Voltage auxiliary output connections Two BNC style Connections are provided on the controller unit on the lower right hand side of the display panel These provide analog signals corresponding to the instantaneous heat current and voltage at the DUT They are referenced to the digital ground of the system or HTP The current signal is calibrated to 0 1V A measured heat current and the voltage output is 0 05V V measured heat current These may be monitored with an oscilloscope or voltmeter
26. ndow will appear asking the user to confirm the limits and part hook up order then to start the test If anything is wrong correct by using the lt Prev button or Cancel The user should carefully double check wiring and test limits before starting the test manual sxw 2004 09 24 6 6 Test Action Parts will be cycled in order As heating current is applied to a part the green Channel indicator LED is lit up on the corresponding module card These should stay lit for 4 seconds before the next unit is energized The test will presently start at zero heat current commanded and slowly ramp up until the target swing has been reached The user should monitor the test until it has reasonably stabilized and ensure cooling is properly functioning The host software queries the 354a on every cycle calculating temperature swing If the swing is to high it reduces heat current command If the swing is too low the heat current command is increased It will take several tens of cycles before the temperature swing is truly stabilized this may be improved for faster convergence in the future 354A host controller oO x System Help Data is logged to an SQL database residing on the host machine One report function is currently implemented in the host control software To access this report generator select the Reports tab test reports default projet Y Select a project Select atest from the top of the 354A
27. nductive media in Rtj c testing Water pulls itself between the part and the block by capilary action greatly improving thermal contact Thermally conductive grease would work but be very messy to clean up and deal with It is important to correctly position the DUT over the case reading thermocouple The thermocouple must touch the case directly under the semiconductor die mounted in the package manual sxw 2004 09 24 21 The DUT should be securely clamped down to the heat block during Rtj c testing for repeatable results Good physical mounting is very important to get accurate values in this test The heat block should be secured to a cooling plate with water running 6 7 3 Host Software Operation for Rtj c testing Host software provided with the 354A supplies an Rtj c measuring routine Select measure Rtj c from the main System Control panel This will open the Rtj c measuring screen select tag for the as Measurements part to measure Enter case temperature celcius select a part test3 3 dut3 X p EE Enter sink temperature celcius Approx heat current amps oO set pulse duration seo 12 heat power 0 0 W Step 1 heat part Tj 0 0 C calculated Rtj c ____ W calculated Rtj s ____ W cancel amp close window Just as in setting up cycle testing it is critical that the correct well charcaterized part is selected and that it is on the
28. oller should manual sxw 2004 09 24 not be flashed w o specific instructions and an upgrade file 5 2 Cautionary statement The 354A must not be connected to an energized heat supply while flashing the code Connection to heat power could damage the 354A as it is not in operational mode when being programmed The supply should be disabled or ideally off before proceeding with flash upgrades 5 3 Upgrade file name The name of the upgrade file to be programmed into the controller is always cycle hex Wyeast Technology will place this file into the proper directory The customer should check to be sure the date of this file is as expected 5 4 Flashing procedure a stop any running tests and ensure that the heat supply is off b exit the host control software c attach the 281 003 01 debug cable assembly to the debug connector just under the AUX 5V out connector on the back panel and free hanging This connector may be left inside the 354A unit if so unplug the AC line power before removing the chassis cover to extract it Never trust a single pole switch to truly disconnect AC line power d double check the heat supply is off Flick the toggle switch to P then reset the controller with the pushbutton It is now in programming mode e send the upgrade code to the controller by running the program aprog from a command line shell in home tester flick the toggle switch back away from P and reset
29. or current TI T2 a Rt in units of C W Power T1 T2 is the temperature differnce across the region in interest In our case for junction to case thermal resistance Tj Te as C W Power Rtjc and for junction to sink thermal resistance Tj Tsink Power Rtjc C W Junction temperature of a part is measured in the same manner as the 354A uses to track temperature swing during cycle testing by voltage drop at a given test current over the DUT s integral body diode manual sxw 2004 09 24 20 6 7 2 Use of Heat Block Case and sink temperature may be measured using a special copper heat block which provides thermocouples measuring the block s own sink temperature and the temperature of the mounted DUT directly below the semiconductor die Case sink and junction temperature should be measured simultaneously In practice measuring maximum temperature on the case and sink and taking a thermal measurement immediately after a heat pulse gives good results Area Directly Under FET Die Tc Case Thermocouple Heat Block and FET In the picture above the Tc case thermocouple can be seen protruding a small amount over the surface of the copper heat block This thermocouple is isolated from the copper by a small insulator fit into the hole The case thermocouple must protrude enough to put gentle pressure on the metal of the DUT case Water is used as a thermally co
30. ormation requested by the Test Limits Entry window Full temperature refers to the temperature swing the test will shoot for Default temperature swing is 100C default number of cycles is 10 000 but this may be changed by the user Start a New Test Please complete the following select project for test test name description of cycles full temperature swing maximum start temperature for a part Cancel lt Prev Next gt a a r default project new_test test of parts 10000 100 0 50 0 Test Limits Entry Window Saved reports will be grouped into subdirectories by project To change the project use the pull down selection in the select project for test box Projects must have already been created using the create edit projects button on the main panel manual sxw 2004 09 24 17 Can Start a New Test e nt eS Confirm and hit START project default project test name new_test description test of parts test will run for 10000 cycles before terminating parts will be cycled in a 100 0 degree swing start temp will not be allowed to exceed 50 0 degrees these parts will be used tag 3 SOM S5LFLLX3 3 test part TO tag 2 irfp3415_11 freds transis tag 4 SOMS5 JVRP_1 1 SOT22 part tag 1 irf3711_13711_1 0220 proper Start Test Confirmation Window Finally a wi
31. pplied voltage to drift while the calibration shot is in process b Adjust the heat supply for an output voltage of approximately 8V c Observe the voltmeter and click the pulse device button Read the value on the voltmeter when the voltage is stabilized Ideally the meter should be read two seconds after the pulse goes into effect green channel light on DUT module but the results will be fine as long as the heat supply is stable Record the voltmeter reading manual sxw 2004 09 24 28 actual heat voltage into spreadsheet cell B28 d Click the read cycle data button on the left bottom of the Admin Panel Cycle data will scroll into the shell window Read the reported V_heat measurement for DUT 8 5502 0 5502 Vd_after 0 3537 0 3533 0 3533 DUT 7 cycle data cycle 221 V_heat 7 2804d I_heat 8 7125 Vd_before 0 5685 0 5685 0 5685 Vd_after 0 3708 0 3712 0 3712 DUT 8 cycle data cycle 222 V_heat 8 0923 I_heat 0 1107 Yod_before 0 5860 0 5860 0 5856 Vd_after 0 5960 0 5856 0 5860 m Cycle Data Scrolled Into Shell Window For example the value read here is V_heat 8 0923 8 0923 would be entered into spreadsheet cell C28 e Repeat this process at a higher heat voltage ideally a little over the actual DUT heating voltage used for testing to get information for spreadsheet row 29 the vheat2 calibration sample f The new calibration values are in cells B32 and B33 Edit the 354a conf fi
32. s are stored in the file 354a conf If needed this file is edited to insert new calibration numbers A spreadsheet is available to do the calculations for calibration Without this the these calculations would be tedious and very error prone The raw ADC unscaled data is not accessible to the user and must be back calibrated from reported values and current calibration numbers in use 7 2 Calibration Equipment Suggested calibration equipment should be on hand before starting calibration 1 text editor on host computer 2 Spreadsheet calibration_worksheet sxc 3 Calibrated voltage meter with clip leads 34401A is ideal 4 Calibrated current meter capable of 10 amps with heavy clip leads manual sxw 2004 09 24 25 5 Extra DUT cable or male connector 6 Schottky Diode and small clip leads 7 Ordinary MOSFET DUT 8 Current limiting heat supply 7 3 Vd Calibration Procedure Vd will be calibrated by measuring at two separate points It is convenient to have a Schottky diode to simulate a DUT at high temperature This will eliminate the need to stabilized the DUT at two different temperatures 354A host controller 5 a Open the 354a Host Control software and switch to the Admin Panel using the top tabs System Help Test Control Reports Admin Pane can tweak layout later anal select D UT 8 read DACs force measurement PLS CEE read cycle data pulse and measure device status 0x80 P
33. the controller again using the 88 y pushbutton The 354A is now back in normal run mode g remove the debug cable assembly and store in a safe place h re start the host GUI software in its startup text window it should report the new software version This should be verified against the information sent by Wyeast Technology along with the upgrade file manual sxw 2004 09 24 6 CycWin 354A Host Software 6 1 Software Introduction Software running on an attached Host Computer handles all control and data logging features of the 354A This software runs and monitors the 354A FET Cycle Controller and logs all data into a flexible database for reporting and analysis It uses a Graphical User Interface GUI for ease of use and clarity 6 2 Logging in and Starting the Software Login to the host computer checker with the user name tester The password for user fester as delivered is test354 Start the Host Software GUI either with the desktop shortcut CycWin 354A Host Software or from terminal shell by clicking on the Terminal Shell and running the script CycWin from the shell prompt If run from a Shell useful debugging information will appear in the text terminal along with the GUI The CycWin Host Software starts with a main window 354A host controller e System Help Test Control Reports Admin Panel Tests etc Parts etc create edit project create edit parts
34. y for every device and with no manual interaction required so as 2a Be fe 3 F OAA MN w 3s i cr es g a noe te A i 354A Front Panel View With Heat Supply manual sxw 2004 09 24 2 1 Overall Function of the 354A The 354A attaches to all leads of FET devices under test DUTs It takes power from an external DC supply Heat Supply and routes it to the correct device according to timing and test parameters Voltage is controlled by the Heat Supply The 354A controls current by active control of DUT gates A host computer connects to the 354A to provide user interface data collection and dynamic adjustment of cycling conditions All physical timing parameters are controlled by the 354A This eliminates fine timing dependency on the host computer s operating system and allows very low jitter cycle timing and measurement triggering HOST COMPUTER DATA STORAGE ANALYSIS AND PRESENTATION l iil ii T e amp zi 5 6 354A FET CYCLING CONTROL f exe a HEAT CURRENT SUPPLY 354A and Surrounding Components STEADY STATE CONTROL 2 2 K Factor Measurement With the host computer and an appropriate thermometer system the 354A may be used to measure device K factors The K factor is a number that relates body Diode Forward Voltage Drop Vd to temperature at a given current K factor is needed by the system to enable measurement of DUT junction temperature in
Download Pdf Manuals
Related Search