Reports until 21:41, Tuesday 11 August 2020
H1 ISC (ISC)
georgia.mansell@LIGO.ORG - posted 21:41, Tuesday 11 August 2020 - last comment - 20:12, Thursday 27 August 2020(56461)
Fixed an issue with in-lock charge script, re-ran for ITMX and ETMX

[Sheila, Georgia]

Last week the in-lock charge scripts were revived and re-run. The scripts are found in /opt/rtcds/userapps/trunk/sus/common/scripts/quad/InLockChargeMeasurements. These scripts take 4 measurements:

There was an error in the charge_measurement_in_lock.py: when turning the bias off for the second 2 measurements it would not set the L3_LOCK_INBIAS to 0, only the L3_LOCK_BIAS_OFFSET. So for test masses where the INBIAS is used to set the bias (ITMX, ETMY, ITMY), the measurements from last week did not work. On ETMX we use the BIAS_OFFSET, so that measurement did work.

I've fixed this by adding a line to charge_measurement_in_lock.py to turn off the inbias (line 286). While I was editing the script I also changed the TRAMP on line 167 to be 60 rather than 180 (too long!).

I ran the new scripts on ITMX, then Sheila switched DARM control to ITMX and we ran them on ETMX. We saw the excitaitons in the ESD Master out's and also in DARM which is promising. I'll post the results in a comment tomorrow.

The measurements inject a line at 13 Hz, and ran from ~1281238170 until the lockloss at 1281239828 which happened while we were transitioning back to ETMX.

If possible we would like to run these measurement on ITMY and ETMY before the test mass discharge.

Comments related to this report
georgia.mansell@LIGO.ORG - 13:19, Friday 14 August 2020 (56493)
Link

The results of the in-lock charge measurements made before the test mass discharge are attached.

The usual analysis script (get_all_ESD_coefs.m) plots the alpha, gamma, beta, and beta2 coefficients as a function of time. I took the data and made bar charts instead since it has been a long time since these scripts were run.

First plot shows alpha, beta, beta2, and gamma.

  • Alpha is a measure of the dipole attraction between the test mass and ESD, and is larger for the ETMs due to the smaller gap between the ESD and the test mass.
  • Gamma is a measure of the dipole attraction between the test mass and electric fields in the chamber, eg fields terminating on the suspension cage. The measurements show it's largest for the ETMs, and the opposite sign to the ITMs?
  • Beta and Beta2 depend on the test mass charge and distribution. Beta is largest for ETMY and Beta2 is largest for ETMX.

I'm not sure why ETMY has such large error bars, I've cut them off the first plot for clarity.

Also note LLO has been running these scripts regularly and find coefficients of the same order of magniture, see latest measurement here.

 

I made a small change to the fucntion which extracts the coefficients from the data (process_single_charge_measurement.m) - it previously did not take into account the 20-bit DACs on the ETMs, resulting in a miscalibtraion of the ETM data. Now it checks if it is analysing and ETM or and ITM and uses 20- or 18- bit DAC. We'll need to update it again when everything is on a 20-bit DAC.

Images attached to this comment
georgia.mansell@LIGO.ORG - 20:12, Thursday 27 August 2020 (56613)
Link

Sheila and Aaron noticed that the gamma co-efficients presented above were negative for the ETMs. This does not make sense - gamma should be the same sign for all the test masses, and should always be positive.  
 
It turns out the script (process_single_charge_meas.m) predates an additional low-pass filter which was installed on the bias path last year (see ECR E1900071, alog 50348, and corresponding circuit diagram T1000220). So I added a 10kOhm 10uF RC low pass filter to the bias path transfer function to the charge script (line 102), and also removed a factor of 4 which I think was accidentally in the script twice (line 147). 
 
Attached are the new co-efficients. They roughly aggree with this previous measurement in alog 38656.
 
The ratio of alpha to gamma is between 0.3 and 0.4 for the ETMs, this roughly agrees with a model made by John Miller - see T1500076. Which explains that gamma is dominated by the field lines between the ESD and the cage/ring heater. It makes sense that gamma is smaller for the ITMs because the gap between the ESD and the ring heater is larger. 
 
Note that these measurements are not taking into account the broken quadrant on ETMY.
Images attached to this comment