aLIGO LHO Logbook

H1 PSL

daniel.sigg@LIGO.ORG - posted 14:00, Wednesday 14 September 2016 - last comment - 14:37, Wednesday 14 September 2016(29701)

New outer loop ISS may be ready

Keita Daniel

We were able to engage the outer loop ISS after the mode cleaner was locked and power up to 50 W in full lock. The third loop was intended to be engaged around 10 W, but due to a scaling error in the arm powers, the gain was too low. We were able to adjust the gain, when we reached 40 W. And then power up to 50 W.

How to engage the outer loop:

To turn the outer loop off, set the OUTERLOOP_ENABLE switch, the OUTERLOOP_BOOST_1 switch and and the output switch of the THIRDLOOP_SERVO to off. The input switch of the OUTERLOOP_AC_COUPLING_SERVO should be on. The SECONDLOOP_CLOSED switch in the first loop ISS may or may not be off.
Once the first loop ISS is working, the SECONDLOOP_CLOSED can be turned on.
Once the IMC is locked, the OUTERLOOP_ENABLE can be turned on.

How to engage the third loop:

When the input power reaches 10 W, set the gain of the THIRDLOOP_SERVO to 0 and the TRAMP to 3.0 sec. Now turn the output switch of the THIRDLOOP_SERVO on and set its gain to 1.0.

This is now coded on the guardian. However, the arm power scaling error needs to be fixed with an LSC model reboot.

New features in the digital ISS model:

The input switch of the OUTERLOOP_AC_COUPLING_SERVO also switches the corresponding AC coupling filter in the third loop.
When the OUTERLOOP_ENABLE is off, the integrator in the AC coupling bleeds off its offset.

We did not test turning on the boost or turning off the AC coupling of the outer loop. Changing the ISS main gain is problematic due to electronics offsets. The same goes for the boost stage.

Comments related to this report

keita.kawabe@LIGO.ORG - 14:37, Wednesday 14 September 2016 (29703)

Link

OLTF of the third loop at 40W, with the outer loop engaged with AC coupling (first attachment, red and blue).

Some notes about power scaling:

We do not change VGA during the power up process because any small DC offset from the outer loop board has a huge impact on the diffracted power.
Therefore the outer loop gain is not scaled by the power. It starts low, and goes higher as the power goes up.
- Lower UGF (made by AC coupling) starts high and goes lower as we power up. Higher UGF starts low and goes higher.
Therefore the third loop should not be normalized with power (because the loop signal is injected to the outer loop error point).
- The "power scaling error" means that the third loop is using normalized arm power as of now, which will be fixed in the next frontend model.
IF the outerloop OLTF is large, the above mentioned design makes the third loop TF independent of the power.
- However, due to AC coupling this is not true at very low frequency. Since the lower UGF of the outer loop starts high with low power, the effect on the third loop OLTF would be non-negligible at 2W.
- This is a non-issue as we only need to engage the third loop at 10W.
- Anyway, just for your curiousity, green trace is thethird loop OLTF measured at 2W. Note that the measured quantity is the negative of the OLTF, and the instability is at 0 degrees. As you can see, it's barely stable at 2W at around 0.4Hz.

Images attached to this comment