B&K 4370 Accelerometer and Charge Amplifier Manual
Contents
1. 0 Weg paneo dueeJd vO of go NN eu 25 Sjuiod JO OY pue IH seDuei 9914 JO 09 pue 5910 ON SMG Bro 1715 oneueuos 269 0 2 obey t 0 vry cvd evo v d L ed Joneo 2 j 9218 9 HS amp JDue9 pu 8 8 40 640 egla POLY 9014 9019 2019 099 en 6 9 406 Ov 91 awos EZ 2002 9900 ON 9MG 3215 I99IA JDUE 9 NMVHG 151 jueuoduJo7 2 o1 ZHWIJO J L 0 50 J L 0 290 MOL 4914 4110 299 2914 471 0 LGO M00L 918 050 M00L Jdozz vro M00L 218 Jdozz 270 77401 vH 29 0 seo Mc9 2218 2 159 701 LP 7029 9218 JdOL veo MOL 9 H 2 9 JdOL 229 MOL Mc 9 PLLY Jdp 86 620 MOL Mc9 2118 341666 120 M6 vc2. Co 32 s wo Qo 22 R RoR 4 an 1 AR wo LC RoR R3 The filter poles come from the denominator of Zo with the cutoff frequency determined by wo and the damping factor by Qo These parameters are highly underconstrained in that the number of determining component values degrees of freedom is large Compo nent values were chosen to implement the desired low cutoff for the various ranges while where providing the desired Butterworth function for each range References B amp K02 Product Data Piezoelectric DeltaShear Accelerometers Uni Gain Delta Tron and Special Types http www bksv com pdf Bp0196 pdf Br el amp Kjaer October 24 2002 Chen95 Wai Kan Chen Ed The Circuits and Filters Handbook CRC Press 1995 ISBN 0 8493 8341 2 pp 2392 2400 TI97a Texas Instruments Data Book Amplifiers Comparators and Special Func tions Texas Instruments 1997 SLYDO11A vol A pp 3 767 ff TI97b Texas Instruments Data Book Amplifiers Comparators and Special Func tions Texas Instruments 1997 SLYDO12A vol B pp 6 287 ff APPENDIX Design drawings The following are the design drawings of the as built Charge Amp TR03 002 UNC Chapel Hill Department of Computer Science page 7 ON SMG ee 1715 SYANS pue sepou 40145 101 ZHWIJO
3. 1 E eN 94 2 88 ES 23 gt 9595 2 55 23 Hb 55 PH A rif ESE 5 i 92 c gross 28 e Ace 7 4 His gt 52 5 856366 55 EP 5 16 fe 25 3 j b ope gp Bg io g OQ Wr 3g 55 5 5 P pore i ok E amp 8 3 j ie sz t Tie 2 9 4 59 gt g Bes 22 5 N 1545 p H S zi z gS Hai x 5 5 2 pes 20 5 9 o A Beer BER P 3 5 BR E z EEPE Wii B ed t t o gt 8 2 Et ge eb 58 2 25 EP 135335556 Eg 58 _ 5 3 85 8893 8 o 8 58 d 5 55 6 7 45 x 053 181 Bheri 5 5 4 Bi 555 352 e 2452 lt 5 53252 g od wo vi mm sec Paper Speed Writing Speed Hz n Freq
4. 9 ALLY 9272311 SN 4101 20 YSZ LEY W29 OcviIcall vn JdOL 229 6 91 92H INL POLH 92 12911 en J G 219 MOL INOL 20184 9 12911 en 39001 209 MOL WOOL 40184 92022011 IN 401 909 YSL OL LSA 490 209 MOL 128 Week 5 SDI 6101615 SL JOY 5 1 81 1 25 V ON 375 2002 094 0 amp Jpue9 peog 2 9 0 1 o1 2 01 NOILLdIHOS3Q Jeyijdure 0 Wea AUN ISI eeu a ke en ocd 59 Led um 420 LED ced YS Ley ved GEO Bi D ol u e e IH A DI m HH 5 gt 219 peiH 1 i d 221 r8 mal qoid EI 2914 Glu 050 2 5 5 5 5078 08 ino C 1715 ON 9MG I99IA JpUE9 NMVHG 2
5. SL dO 4133 1 25 V 005 11 gzis I09IA paublsap se 47145 ZH OL o1 2 01 NOlI LdlHOS3Q 0 Wed gJILIL TSIN Kouenbai ZHAN ZH ZH ZH ZH ZH ZH ZHW ZHW 001 01 001 01 001 Ol 2 zvs w Aw OL 9 ZvS W AW 001 9 1 esuodsei jeuiduy uonejnuis 49145 ZHUJ Ot L 061 0cL 001 06 08 0 09 OS Or 06 0c Ol zS u A qp ures tas 0 37426 HM 000128 N S OLEH BE 20 S JAOUEN abieyd 01639 e gt 558 5 ges m o 8 942 N g5 5 9 wm 88 2 29 55 49 5 88 ii 29 88 5 9 uS 2 35 99 o g 9 ot i E 228 48 3 98 5 9 fore 25 85 53 925 y gt 88 v gt OS gL 5 re 28535 58055 8 85 25 dE 82882 Mi 655 8 52522
6. 0 OF 2 0 NOILdlH2S3G 0 ON 9MG I99IA A8 NMVHG ss Oe 1715 Sjeoiueuoeui OF 2 01 NOlLdlH2S3G jeyidure 0 5 Wea 82069 xog eui do y 1e 922 72 y JO 9y 12 1 5910 296 2 SION 4 0 00t 2 48208 ON 9MG I99IA JpUE9 NMVHG Im IO 1715 2 101 beyo NOILdlH2S3G jeyidure 076v Wea 202 gt 8982 gt seoel d 2 05 seoe d 2 IUD 8 1 9 6 2 IUD 8 5 08672 91152 A 49071 Y 60 F 1 9 pue 91 6 lt gt 0622 gt lt 509 gt 1715 Sjeoiueuoeui Auegeg NMvud 2 01 beyo NOlLdlH2S3G jeyidure 0 5 98 euejAdoud jod uojAu se uons Joyejnsul ejqeuiuoeuu epuq uou 000 c uupue 9ue seoe d 26 9 de
7. 6 NOLLdIHOS3Gd 260 0 2 0 6 86071 1sefue 899 0 1 1 GZ j euoBeip soe 0771 6181 99 0 899 0 eet 0 TARS IIIT TITE LL LILIIIIIITII Note C42 and C44 were added after these photos were taken 40 O DESCRIPTION Circuit photos top and bottom Charge amplifier 10 to 10 kHz SIZE pwaNo DRAWN BY Leandra 4 0 5 12 15 SCALE 16 Jan 2003 DATE S JOEL 4133 divos V 5005 uef ggg 921 7 paublsap se jo 47145 o1 ZHWI O NOILLdIHOS3Q 0 1 5 IN Kouenbeu 2 2 zH zH zH z4u 2 znHuj 001 OL 001 Ol 001 OL 081 091 071 021 001 08 ures 09 Or Oz 8 001 5 JO esuodsei 30145
8. tder The resonance of the 4370 is not explicitly modeled here but the acceleration stimulus including gravity is represented by Va Changes in cause a charge transfer through the transcapacitance element Cs Parallel capacitance comprises the parasitic capacitance of the 4370 and its connecting cable The transcapacitance gain was chosen to match the manufacturer s sensitivity data so that V represents an acceleration of 1 m s Note that this is strictly a simulation model 2 4 preamp This 1s the integrating amplifier which transduces charge to voltage The transelas tance gain 1s the reciprocal of capacitance The preamp also provides one real pole to the fifth order anti alias filter The frequency of this pole is established by and the parallel combination of C and C Gain ranges are implemented with switch selectable capacitances of 1 10 and 100 nF for Co Operation at very low frequency requires an amplifier with a low noise floor as well as very low bias current U1 is a TLC2202C dual low noise precision op amp implemented LinCMOS technology with 1 pA typical bias current and 30 nV V Hz at 10 Hz in the 1 f noise regime TI97a 2 5 cauerl and cauer2 Each of these circuits provides two complex poles and two complex zeros to the anti alias filter These poles along with one real preamp pole implement a fifth order low pass Butterworth function The four complex zeros provide the desir
9. 95 seoeldz _ pue B E 00071 JUD Lii 0040 6290 05 70 05 70 006 gt ON 9MG I99IA NMVHG 91306 Eee 325 pued 2 0 NOILdlH2S3G 0 015733 991 39HVHO ZH 330119 MOT 5 30NVH 2002 2991 06 1715 ON SMG pue 2 01 o1 zHuu o 26245 0757 gt 1 IN 19 9W019 9998 suuni sui uey 5 paesu 21015 LON OG 1 41 sued 2 21018 Od 3SV3 ld 2 8 9 si ouenbay jyojno 4 5 49410 10 edoos 0 2 Jeyiduue pue 6 2154 AYVININNS e ueios SWIda SAS 21004129 90J91IN daldl1diNV dildWOH31322V 0 5 3 4 9130 mejor 18 eG 325 I99IA 7
10. B amp K 4370 Accelerometer and Charge Amplifier Manual Vices Microelectronic Systems Laboratory Department of Computer Science University of North Carolina at Chapel Hill 21Jan003 Summary The Br el amp Kjeer 4370 is a single axis piezoelectric accelerometer sensitive along its Z axis down to nearly 10 ug and conventionally use ful over a range of 0 1 Hz up to about 1 3 of its mechanical resonance frequency of 15 21 kHz depending on its mounted configuration The charge available on its output 1s proportional to acceleration sensed by the device including gravity To be useful this charge must be inte grated by a charge amplifier to provide a voltage proportional to accel eration The overall low frequency performance depends on the limiting noise floor and frequency response of the charge amplifier charge amplifier was designed specifically for extending the useful low end of the frequency range of the 4370 to 0 01 Hz and to provide anti alias filtering for adequately suppressing frequencies at the mechan ical resonance of the accelerometer This filter is fifth order Cauer filter with zeros in the stop band to cancel the mechanical resonance It provides a flat frequency response to its upper 3 dB cutoff frequency at 6 8 kHz and guarantees at least 20 dB anti alias suppression of all acceleration stimuli above 11 kHz This report comprises a user manual for the accelerometer and charge amplifier kit a theory o
11. age from battery leakage the Charge Amp should never be stored with the 9 volt battery in stalled The external battery access feature is Intended to make it convenient to observe Figure 1 Kit consists of an accelerometer a cable a charge amplifier shown with bat tery installed and a carrying case this rule Also please note the polarity of the snap connectors and do not attempt to install the battery backwards The accelerometer should be mounted to the device under test DUT using either a 10 32 fastener or double backed adhesive tape Under no circumstances should the fastener be forcibly threaded into the accelerometer Slightly over two com plete turns is the maximum thread depth in the 4370 manufacturer s recommended mounting techniques for threaded fasteners and alternatives are excerpted from B amp K 02 and shown here as Figures 2a and 2b The charge amplifier should be set for the desired gain and low frequency cutoff Be aware that the lower frequency ranges may take a long time to settle over a minute for the 0 01 Hz range Settling can be hastened somewhat by briefly switching to a higher frequency range and back but it will still take some time Even when completely settled there is a residual DC error voltage of a few millivolts If very high accuracy is desired this should be measured and subtracted from subsequent data The output voltage range of the Charge Amp is nominally 2 5 V Depending on
12. ak of the accelerometer over the range of 15 21 kHz in which it can occur This provides at least 20 dB anti alias filtering for all mechanical excitation above 11 kHz TR03 002 UNC Chapel Hill Department of Computer Science page 3 L B amp K 4370 Accelerometer and Charge Amplifier Manual 21Jan003 Alternative Mounting Techniques When mounting techniques other than the recommended technique are used the ac celerometer s mounted resonance frequency will probably be lowered Fig 11 Alternative mounting techniques Insulating Stud Mica washer Fig 11 shows some alternative mounting techniques The section entitled Standard Accessories on page 15 lists the mounting accessories that are supplied with the individual accelerometer types These mounting techniques are described in more detail in the Br el Piezoelectric Accelerometers and Vibration Preamplifiers hand book where the effects of the different methods on the frequency response curve of an accelerometer are illustrated Figure 2b Alternative techniques excerpted from B amp K 02 irrespective of the accelerometer resonance Refer to page 13 of the design drawings for detailed response curves as simulated by SPICE The output of the charge amplifier will drive any single ended instrument input impedance of gt 50 Q through a conventional BNC cable It should be relatively insensi tive to ground loops between the accelerometer and the
13. battery voltage the output will clip at 2 9 or less If the Charge Amp is overloaded to clipping its DC offset will be upset and must be allowed to re settle before meaningful measurements can be made The remedy for this 1s either to use a less sensitive gain range so that overload does not occur or to use a low cutoff range that settles acceptably quickly after an overload The noise floor of the Charge Amp is on the order of 100 on all ranges On the most sensitive range this corresponds to an acceleration noise floor of approximately 100 ug As of this writing these measurements are very approximate and need to be TR03 002 UNC Chapel Hill Department of Computer Science page 2 L Mounting Fig 9 Recommended mounting technique Fig 10 Recommended tolerances for the mounting surfaces Dimensions and symbols in accordance with ISO 1101 B amp K 4370 Accelerometer and Charge Amplifier Manual 21Jan003 Br el accelerometers can be mounted with their main sensitivity axis aligned in any direction Recommended Mounting Technique Fig 9 shows the recommended method for mounting most of the accelerometer types The accelerometers are screwed using a threaded steel stud onto a clean metal surface meeting the requirements specified in Fig 10 Under normal circumstances the absolute minimum depth of 4 mm will not be sufficient to accommodate the mounting stud but is the minimum dept
14. dra Vicci SIZE DATE 30 Dec 2002 1 15 REV 40 SHEET 1 OF 15 05 li TW T zm B amp K 4370 Charge amplifier co 5 L DESCRIPTION g p pwrgnd Figure 3 Charge Amp schematic showing SPICE subcircuits 2 1 pwrgnd The Charge Amp uses a single 9 volt battery and needs to have a local ground refer ence preferably midway between the positive and negative power rails This 1s generated by U5 a TLE2426 Virtual Ground Generator This local virtual ground is connected to the instrument s case and through the connecting cable to the body of the 4370 It 1s therefore important not to allow a ground loop to be established between the 4370 and the Charge Amp The Charge Amp has insulating rubber feet to isolate it from its supporting surface TR03 002 UNC Chapel Hill Department of Computer Science page 5 L B amp K 4370 Accelerometer and Charge Amplifier Manual 21Jan003 2 2 diffout Ground loop noise between the Charge and the instrument it drives 18 however suppressed U2b is embedded in a differential amplifier configuration which isolates the local ground from the output common This prevents a ground loop from occurring between the Charge Amp and the driven instrument The amount of ground noise suppression depends on the CMRR of the differential configuration of diffout which should approach 40 dB 2 3
15. driven instrument However care should be taken to prevent the Charge Amp box from electrically contacting anything as ground loops between the accelerometer and the Charge Amp itself can cause noise problems The Charge Amp is furnished with rubber feet to help provide the necessary electrical isolation TR03 002 UNC Chapel Hill Department of Computer Science page 4 L B amp K 4370 Accelerometer and Charge Amplifier Manual 21Jan003 2 Charge Amp Theory of Operation The Charge Amp is logically composed of five functional blocks and a power manage ment block Referring to Figure 3 these are the charge integrating preamplifier preamp two stages of anti alias filter cauer1 and cauer2 an integrating feedback circuit dcfb an output driver diffout and power management pwrgnd Also shown is an electrical model tdcr of the accelerometer in which the acceleration stimulus is represented by voltage Va Notes 1 R6 and CO are gang switchable for three gain ranges 2 R1 R2 and RO gang switchable for four low cutoff points PREMIERS Eak cte eI E 2 54 R1 R2 s in 1 1 R4 R3 jet 93 id T 040 R2 cauer2 Charge amplifier 10 mHz to 10 kHz with SPICE nodes and subckts shown DRAWN BY Lean
16. ed signal suppression over the frequency range of 15 5 kHz through 21 4 kHz to cancel out the 4370 resonance Refer to pages 13 15 of the design drawings for a quantitative comparison 15 5 kHz is the resonance of the 4370 mounted on a very large DUT mass while 21 4 kHz is the resonance of the unmounted device The circuits themselves are nullor realizations of generalized impedance converter GIC derived biquadratic functions Chen95 2 6 This circuit sets the low frequency cutoff of the Charge In combination with the accelerometer transfer function and the preamp transelastance it implements a crit ically damped second order high pass filter function that is second order Butterworth This function has two DC zeros and a pair of complex conjugate poles which set the cut off frequency This provides 40 rolloff below the switch selectable cutoff frequencies of 10 1 0 1 and 0 01 Hz see page 13 of the design drawings TR03 002 UNC Chapel Hill Department of Computer Science page 6 L B amp K 4370 Accelerometer and Charge Amplifier Manual 21Jan003 The circuit itself synthesizes the equivalent of a parallel RL circuit which is connected across in the preamp Accounting for the combined DC gain A of the Cauer filter sections we have R Ro R4 L Ro R4 X ARR The equivalent feedback impedance of the preamp circuit is then 1 1 8 1 7
17. f operation section and an appendix containing the design drawings Casual users need not delve further than the user manual although the theory of operation provides more detailed technical insights which will be useful to the more serious user The design drawings along with the theory of operation are intended to be sufficiently detailed for anyone skilled in the art to adapt and construct the charge amplifier design for other accelerometers TR03 002 UNC Chapel Hill Department of Computer Science page 1 L B amp K 4370 Accelerometer and Charge Amplifier Manual 21Jan003 1 User Manual The accelerometer and charge amplifier Charge Amp are provided as a complete kit and should always be kept together special coaxial cable provided to connect the accelerometer to the charge amplifier is a calibrated component with microdot connectors which will not mate with any of the standard connector families To encourage keeping the kit together a convenient carrying case 1s provided usage synopsis is posted on the inside of the case Use of the kit 1s straightforward that s necessary 1s to power the Charge Amp mount the accelerometer on the device under test DUT connect them together and to a data acquiring instrument such as a DAQ card or oscilloscope and set the gain and low frequency cutoff switches To power the Charge Amp 9 volt bat tery should be snapped into the front panel receptacle To prevent dam
18. h required to hold a stud securely The optimum torque for tight ening 10 32 UNF steel studs is 1 8Nm 151b in for steel studs it is 0 6Nm 51 and for M8 steel studs it is 4 6Nm 381b in 10 32 UNF M3 M4 M5 or M8 1 6 This mounting method is used in obtaining the specifications of all the accelerometers with the following exceptions Type 4374 due to its small size cannot be mounted using a stud The recommended mounting technique used to obtain the specifications utilises a quick setting methyl cyanoacrylate cement Br el amp Kjaer QS 0007 The tolerances on the clean metal mounting surface shown in Fig 10 are required gt Type 8309 has 5 metric screw stud as an integral part of its base The tolerances shown in Fig 10 apply and the optimum torque is 1 8Nm 15 1b When using the recommended technique it should also be noted that if the mounting surface is not perfectly smooth the application of a thin layer of grease to the base of the accelerometer before screwing it down on the mounting surface will improve the mounting stiffness Figure 2 Mounting techniques excerpted from B amp K02 quantified more carefully if the kit is to be used at sensitivities where this is an issue The upper cutoff frequency or 3 dB rolloff point of the Charge Amp s passband is nominally 6 8 kHz In the stopband a special rejection band is provided to suppress the resonance pe
Download Pdf Manuals
Related Search