A Proposed Capacitive Position Sensor Timing Synchronization and Distribution Solution. Ben Abbott May 9, 2014 G v1

Size: px

Start display at page:

Download "A Proposed Capacitive Position Sensor Timing Synchronization and Distribution Solution. Ben Abbott May 9, 2014 G v1"

Julianna Harmon
5 years ago
Views:

1 A Proposed Capacitive Position Sensor Timing Synchronization and Distribution Solution Ben Abbott May, 0 G00-v

2 The Problem From aligo LLO Logbook report 00. Orange trace is ITMX with all chambers CPSs on, Blue trace is ITMX with the other two chambers turned off. G00

3 The Temporary Solution J 0 D Connector STS-_U+ STS-_U- STS-_V+ STS-_V- STS-_W+ STS-_W- STS-_X+ STS-_X- STS-_Y+ STS-_Y- STS-_Z+ STS-_Z- J 0 D Connector J STS-_X+ Out STS-_X- Out STS-_Y+ Out STS-_Y- Out STS-_Z+ Out STS-_Z- Out 0 J STS-_X+ Out STS-_X- Out STS-_Y+ Out STS-_Y- Out STS-_Z+ Out STS-_Z- Out 0 STS-_X+ TP C UF C UF - + UA ADARZ A TP Out+ R 0 OHM STS-_X+_Out D Connector J STS-_X+ Out STS-_X- Out STS-_Y+ Out STS-_Y- Out STS-_Z+ Out 0 STS-_Z- Out D Connector J STS-_X+ Out STS-_X- Out STS-_Y+ Out STS-_Y- Out STS-_Z+ Out 0 STS-_Z- Out TP Out- STS-_X-_Out STS-_X- TP B UB ADARZ R 0 OHM STS-_Z+ TP C UF C + UA ADARZ R A 0 OHM TP0 Out+ STS-_Z+_Out D Connector D Connector UF J STS-_X+ Out STS-_X- Out STS-_Y+ Out STS-_Y- Out STS-_Z+ Out STS-_Z- Out 0 J STS-_X+ Out STS-_X- Out STS-_Y+ Out STS-_Y- Out STS-_Z+ Out STS-_Z- Out 0 STS-_Y+ TP + C UF UA ADARZ R A 0 OHM TP Out+ STS-_Y+_Out TP Out- STS-_Z-_Out STS-_Z- TP - UB ADARZ B R 0 OHM D Connector D Connector C R R Res Res 00K 00K R Res 00K R Res 00K Repeat(STS,,) STS-_Dist_Channel.SchDoc STS-_X+ Repeat(STS-_X+_Out) STS-_X- Repeat(STS-_X-_Out) STS-_Y+ Repeat(STS-_Y+_Out) STS-_Y- Repeat(STS-_Y-_Out) STS-_Z+ Repeat(STS-_Z+_Out) STS-_Z- Repeat(STS-_Z-_Out) STS-_X+ STS-_X- STS-_Y+ STS-_Y- STS-_Z+ STS-_Z- STS-_X+ Out STS-_X- Out STS-_Y+ Out STS-_Y- Out STS-_Z+ Out STS-_Z- Out J STS-_X+ Out STS-_X- Out STS-_Y+ Out STS-_Y- Out STS-_Z+ Out 0 STS-_Z- Out D Connector J0 STS-_X+ Out STS-_X- Out STS-_Y+ Out STS-_Y- Out STS-_Z+ Out 0 STS-_Z- Out D Connector TP Out- STS-_Y- TP UF - UB ADARZ B R 0 OHM STS-_Y-_Out R R Res Res 00K 00K +Prot D N0 TP +Vin L mh, choke C uf, V Vin C 0.uF TO- D N00 U +V Power Interfaces and Regulation TP +V + + C uf, 0V +Sw -Sw K +Prot Relay-SPST D Diode P Header -Sw +Sw K -Prot Relay-SPST D Diode P Header Distribution Channels ADs Board Power +Sw P -Sw Header H TP -V U - L Vin -V mh, choke D C0 uf, V + C TO- + C TP.uF, V uf, V N0 -Vin -Prot P LED A LED K Header WM00 P LED A LED K Header WM00 R K R K +Sw -Sw TP T STS- Distribution Chassis G00

4 The Temporary Solution G00

measurements from Sunday night with both ITMY and BS on the same master and ITMX still running, and in any other colors are the measurements

5 Preliminary Results You can find attached sample spectra for the CPS sensors of all BSC ISIs. In orange are the measurements from last Wednesday night when all ISIs had different masters ("comb reference"), in black are the measurements from Sunday night with both ITMY and BS on the same master and ITMX still running, and in any other colors are the measurements from last night. In all cases, those measurements were taken with St ISI isolated, St only damped. One can see that there is no more comb visible on those spectra! From aligo LLO Logbook Report 0 G00

6 Problems with the STS- Dist. Approach ADs are analog OpAmps, and are not fast enough to recreate there high-frequency square pulses. The waveforms produced are ugly, sine-ish lumps, and are easily perturbed by people touching the system in any way. There is no synchronization to the site clock. We re going to need that STS- Distribution Chassis back. G00

7 The Proposed Distribution Scheme G00

8 The CPS Synch Board Schematic +V C UF RS Driver.0 MHz Sine Wave R N/P : Turn ratio TP T PRI SEC PRI SEC J ADT-WT C C 0nF C 0nF 0.uF TP ClockIn PSoC Input Clock P[] P[] P[] P[] P[] P[] P[] P[] P[] P[] P[] P[] PSoC Interface Square Pulse Square Pulse Square Pulse +V P Header 0X Pulse Square Pulse Square Pulse Square ClockIn P[] P[] P[] P[] P[] P[0] P[] P[] P[] P[0] P[] P[0] Square Pulse Square Pulse Square Pulse Square Pulse Square Pulse Square Pulse TP TP TP TP TP TP TP0 TP VCC Y A Z Y A Z Y A Z Y A Z G ~G +V C UF VCC Y A Z Y A Z Y A Z Y A Z G ~G +V C UF U Square + Square - Pulse + Pulse - 0 Square + Square - Pulse + Pulse - SNLBCADW RS Driver U Square + Square - Pulse + Pulse - 0 Square + Square - Pulse + Pulse - SNLBCADW RS Driver Square + Square - Pulse + Pulse - Square + Square - Pulse + Pulse - Square + Square - Pulse + Pulse - Square + Square - Pulse + Pulse - Square + Square - Pulse + Pulse - J 0 D Connector J 0 D Connector J 0 D Connector J 0 D Connector J 0 D Connector V Voltage Regulator D RB0M-0 TP U +V IN OUT ADJ Power In From Breaker P D +V STPS0LG Header V Voltage Regulator D RB0M-0 TP U +V +V IN OUT ADJ TP D Diode N +V +V R Board Power.K D R +V DUAL_LED TP TP TP TP A A A A G ~G VCC U Square + Y Square - Z Pulse + Y Pulse - Z 0 Square + Y Square - Z Pulse + Y Pulse - Z SNLBCADW Square + Square - Pulse + Pulse - J 0 D Connector LMT LMT + C 0.uF, 0V Tant. R 0 C 0uF R D RB0M-0 C 0uF + C 0.uF, 0V Tant. R.K C0 0uF R D RB0M-0 C 0uF TP TP G00

9 The PSoC Program G00

10 The PSoC Frequency Divider G00 0

11 Estimated Cost at the Vertex (These numbers are raw, and contain no contingency) To fix the Vertex BSC Chambers only: Stanford DS ( ea.) $,00 CDS Electronics $,000 Cables $0 Total = $,0 per IFO To Add HAMS: CDS Electronics $,000 Cables $00 Total = $,00 additional per IFO G00

12 Estimated Cost at the Ends To fix both End BSC Chambers: Stanford DS ( ea.) $,000 CDS Electronics $,000 Cables $00 Total = $,00 per IFO There is currently no 0MHz clock at the End Station, so another timing synch solution would have to be thought up. $? As there is no crosstalk at the end stations, being that there is only one chamber, the only impetus to make this change would be to synch the CPS synch pulse to the site clock. G00

13 LLO alog Text (00 & 00) L SEI Link celine.ramet@ligo.org - posted :, Wednesday December 0 - last comment - :, Wednesday December 0(00) CPS Comb: ITMX measurements with ITMY+BS CPS powered off [Rich M, Marissa, Ryan, Celine] Yesterday, we confirmed that the "comb" issue was visible on all BSC-ISIs CPSs, and that it affected much more some individual CPSs than others. We were able to change which were the CPS most affected, but not to understand why or how. We did notice though that ITMY and ITMX comb seemed to start at different frequencies, while the BS one looked like a combination of the. Suspecting interaction between chambers, this morning we shut down all CPSs of ITMY and BS (from :0 am to 0: am) and compared ITMX St CPS spectra with the other chambers CPSs powered on (ref in orange on the attached spectra) and off (in blue). Shutting down the other chambers CPSs resulted on having ITMX comb disappear from ITMX St CPS spectra. We are now trying to understand the interaction between chambers in order to mitigate the issue in an effective way. Comments related to this report richard.mittleman@ligo.org - :, Wednesday December 0 (00)Link We are now fairly convinced that the problem (a frequency comb in the CPS spectra) is due to having multiple master oscillators in close proximity (what close means is still to be determined). We measured the frequencies of the sync pulses on the three BSC chambers (~.Khz) and found that the differences between the frequencies was about the same as the observed comb frequencies. We then powered down the ITMY ISI and eliminated the master oscillator on the ITMX ISI. We then added a cable to bring the sync signal from the BS chamber to the ITMX ISI. After doing this the ITMX ISI ran well and the CPS spectra showed no signs of a frequency comb. Unfortunately this loaded down the BS sync sufficiently so that it did not run well. the major problem was in the Stage V CPS sensor (which is the master). So we restored everything to the way it was before and everything is running as it was previously. It looks like the solution is going to be running neighbouring chambers off of the same master and installing a buffer amplifier so that more slaves can be driven then we currently can Rich A. has suggested that we replace all of the local master/sync signals with a gps generated signal, then all of the chambers would be running at the same frequency G00

14 LLO alog Text (0) L SEI Link celine.ramet@ligo.org - posted :, Tuesday January 0 (0) All corner station BSC-ISI CPSs running from Master [Ryan, Celine] Since the comb observed on the CPS signals was observed to be linked to the beating of the different BSC-ISI Master CPSs against each other (alog 00), we have tried to have all BSC chambers from the corner station running from one master. Using one STS distribution box (D000) installed in the LVEA in the rack located at the -X/+Y corner of the BS, we have had the BS and ITMY running from the BS St-H sensor used as Master since last Thursday (alog 0). Yesterday, we connected ITMX to the same STS distribution interface and now all BSC-ISI have only CPS Master: BS St-H. You can find attached sample spectra for the CPS sensors of all BSC ISIs. In orange are the measurements from last Wednesday night when all ISIs had different masters ("comb reference"), in black are the measurements from Sunday night with both ITMY and BS on the same master and ITMX still running, and in any other colors are the measurements from last night. In all cases, those measurements were taken with St ISI isolated, St only damped. One can see that there is no more comb visible on those spectra (all other variations are likely due to variations in ground motion when those measurements were taken). Also, it is to be noted that high frequency noise (>0 Hz) has not gotten any worse with all those changes (we were worried about our grounding configuration). Note that despite this success, this is only a temporary setup. We believe that the filtering in the STS distribution slows down the sync signal sent to the slave CPSs, and we're not sure how this might affect the ISI performances. Also, I have noticed that the system is very sensitive to any new connection made to that interface box, and that the BS sync signals need to be connected last and carefully to avoid issues. We believe that this is linked to some grounding problem. However, since it does not seem to affect ISI performances at this point, we haven't done (yet) any further investigation. All in all, this success gives us a proof of concept that we will be able to run with only Master for all BSC-ISI CPSs from the corner station, resulting in getting rid of the comb plegging our spectra. Even though we can not explain this at this time, we do not a see any signs of beating between HAM-ISI Master CPS. G00

Single Event Radiation Test Report. Microsemi LX Analog Input RAD Tolerant Telemetry Controller

Single Event Radiation Test Report Microsemi LX7730 64 Analog Input RAD Tolerant Telemetry Controller Doc. N :TDRS-0020-4 Page 1/33 TABLE OF CONTENT 1. PURPOSE AND SCOPE... 3 2. APPLICABLE DOCUMENT...