【Sheila, Evan】
Robert previously reported (67075) that the coupling of grounding noise at EX into DARM could be minimized if the ESD bias is set to a particular voltage (+128 V). We wanted to understand if this also corresponds to the voltage that minimizes the ESD actuation strength. To do this, we actuated on DARM using EY and monitored a 17.6 Hz EX ESD drive line while also varying the EX bias. We found that an EX ESD bias of +33 V minimizes the actuation strength, and therefore does not correspond exactly to the bias that minimizes the ground noise coupling.
EX ESD bias values tested (as witnessed by H1:SUS-ETMX_L3_ESDAMON_DC_OUTMON): [-453, -298, -159, -90, -21, 49, 118, 261, 413, 261, 118, 48, -20, -90] V
Observed ratio of DCPD SUM / H1:SUS-ETMX_L3_CAL_LINE_OUT_DQ at 17.6 Hz: [7.8, 5.3, 3.1, 2.0, 0.9, -0.3, -1.4, -3.6, -6.1, -3.6, -1.4, -0.4, .9, 2] × 10−4 mA/ct (neglecting an overall 90° phase shift in the transfer function)
The test began around 2023-02-09 23:51:00 Z.
Tagging CAL.
FYI: This may be the beginnings of interest in reducing the bias voltage permanently, which will impact the EX ESD actuation strength.
@Evan & Sheila (or others who wish to follow in their footsteps) -- a friendly reminder that if you go for this permanently, as a first step to retain decently low systematic error in the overall calibration, you can fudge the gain in the CAL-CS calibration pipeline
H1:CAL-CS_DARM_ANALOG_ETMX_L3_GAIN
to match the drop in actuation strength that you've either (a) measured independently elsewhere, or (b) found the impact of in the broadband DELTAL/PCAL transfer function and want to minimize it.
We'll catch up with you and "fix" the pipeline... eventually (a reminder that the down side is that when we fudge, the primary downside is that we lose the accuracy in all front-end computed TDCFs).