Evan, Sheila, Elenna

We have been plagued recently by 2 Hz instability in the yaw ASC loops, and have so far been unsuccessful at finding the origin of this instability. Today, Evan and I took a CHARD Y OLG at 10W, and I compared it with an OLG measurement I took back in December after I updated the control loop design. The measurement showed 6 dB less gain in the loop now.

3 dB of that gain loss can be explained by a choice we made recently regarding the 9 MHz reduction compensation. We going to high power, we reduce the modulation depth of the 9 MHz by 3 dB. We then compensate for that reduction digitally, except for some time we have failed to compensate that reduction in the 9 MHz ASC WFS. This meant that we essentially reduced the gain of any loop using the 9 MHz WFS by 3 dB before going to high power. CHARD Y is sensed on both A and B REFL RF9. I corrected the guardian to properly compensate for the 9 MHz reduction, and then reduced the 9 MHz input matrix values by 3 dB so we do not change the ASC loops that have bee running stably for a long time.

We have no explanantion yet for the other 3 dB of lost gain. However, we increased the CHARD Y gain by 3 dB (125 to 175) and that has seemed to improve stability. While at NLN, we kept the CHARD Y gain hight at 175 and it still seemed to keep everything stable. I tried to measure the CHARD Y loop to see if we actually needed that much gain with the new soft loop changes, but I caused a lockloss by exciting too hard. For now, the guardian will keep CHARD Y at a high gain until we can determine what is best for stability. This may be one reason why the low frequency noise is still high, as CHARD Y contributes significantly to DARM.

CHARD P is sensed with both 9 and 45 MHz WFS. We measured the loop and there was no change in gain. INP1 loops are also sensed on 9 MHz but they have very low bandwidth (0.01 Hz), so we determined they are likely fine.