Displaying report 1-1 of 1.
Reports until 14:46, Tuesday 04 May 2021
H1 PSL
daniel.sigg@LIGO.ORG - posted 14:46, Tuesday 04 May 2021 - last comment - 15:45, Tuesday 04 May 2021(58845)
ISS Second Loop

Marc Keita Daniel

We swapped the modified ISS second loop servo with the spare. However, this resulted in an unstable AC coupling servo. The gain had to be lowered significantly to make it a well behaved servo. We then modified the original ISS to see, if this has something to do with the spare. Indeed, the original ISS was table with the original gain, but had like less than 2° phase margin! We set the gian to -0.08 from -0.3, but put it into a filter module, FM10, so the gain in the filter module is now 1. Also, added a ramp of 10 seconds. This now allows us to turn the ISS on DC-coupled. Here are the steps:

  1. Set the ISS_SECONDLOOP_ENABLE to off, SECONDLOOP_OUTPUT_SWITCH to on, SECONDLOOP_CLOSED to on (ISS second loop is off).
  2. Turn on the AC-coupling by setting SECONDLOOP_AC_COUPLING_SERVO_GAIN to 1, and disable the HOLD on the SECONDLOOP_AC_COUPLING_DRIVE.
  3. Make sure that the ISS diffraction power is around 2.5%. If not, adjust ISS_REFSIGNAL.
  1. Set SECONDLOOP_AC_COUPLING_SERVO_GAIN to 0.
  2. Wait unitl the 10 second ramp has expired, then enable the HOLD on the SECONDLOOP_AC_COUPLING_DRIVE.
  3. Check that the ISS_SECONDLOOP_ERR1_MON value is fluctuating around zero and doesn't have a significant offset. If so, undo and repeat steps 1 and 2 above.
  4. Set the ISS_SECONDLOOP_ENABLE to on.
  5. Turn on boosts 2 and 1, ISS_SECONDLOOP_BOOST_2 and ISS_SECONDLOOP_BOOST_1.
  1. Set SECONDLOOP_OUTPUT_SWITCH to off. (Don't use ISS_SECONDLOOP_ENABLE to turn the servo off!)
  2. Turn the boosts off and redo the preparation steps.

To reduce the time for the output offset to converge, we increased the gain in ISS_SECONDLOOP_REFERENCE_IN_MTRX_1_2 and ISS_SECONDLOOP_REFERENCE_IN_MTRX_1_4 by 10 and 20, respectively.

We also measured the ugf using an SR785. It was about 15kHz and had 40° phase margin with the boosts on. (There is a gain difference of 10 between the two BNCs of ERR2 and ERR1.)

Comments related to this report
keita.kawabe@LIGO.ORG - 15:45, Tuesday 04 May 2021 (58850)

OLTF w/o boost (1st attachment) and with 2 stages of boost (2nd attachment, note the different vertical scale from the 1st plot). UGF ~20kHz for both, phase margin is ~50+deg w/o and ~40deg with boosts.

Little structure close to the UGF didn't look like a real TF structure as it changed significantly from measurement to measurement, but neither integrating longer nor making the injection bigger didn't help much. A line or a noise bump around here?

Images attached to this comment
Displaying report 1-1 of 1.