Takano, Ushiba, Izumi, Aso
We have been struggling to stably lock the ALS DARM because of the noise increase when the arms are near the IR resonances. We spent the whole day today investigating this problem.
Conclusion
In a nutshell, the culprit was REFL WFS QPDs.
For the moment, we blocked the beam going to the QPDs. We also turned off the RF QPD interface. Therefore, you cannot use these QPDs now.
With the QPDs disabled, now the ALS DARM is much more quiet.
What we did
Single arm lock test
First, we locked the X-arm with ALS plus CARM IR feedback. Then adjusted the PLL frequency to keep the IR near the resonance. The spectrum of the MCL feedback signal gets noisier when IR is near the resonance. This is a similar behavior we observed before the REFL PDA1 cable was fixed (11014). However, this time, the amount of noise increase was smaller (a factor of 2 or so).
Just like before, by turning down the REFL power with the remotely controlled half-wave plate, we could eliminate this problem.
We also tried the same thing with Y-arm. In this case, we did not observe any noise increase.
ALS DARM check
Then we moved on to the ALS DARM/CARM configuration. By putting X-arm on the IR resonance, the noise increase can be observed. This noise increase kicks the 7.5Hz ETMY resonance in pitch, making the hand over of the control to IR signals impossible.
Again, the noise increase can be eliminated by turning the wave plate.
In the tunnel
We went into the tunnel and tested many things, such as blocking beams before PDs/QPDs, disconnecting cables, turning off PD interface boxes and so on.
In the end, we found that the noise increase is eliminated only when we block the beam goes into the WFS QPDs, or turning off the QPD interface.
At first, QPD1 seemed to be the source of the problem (blocking only QPD1 reduced the noise). However, after we did a swap of the D-Sub cables connecting the QPDs and the interface box, both QPD1 and QPD2 started to cause the problem, i.e. we need to block both of them to eliminate the noise increase.
We also observed that by turning on and off the QPD interface box, the ALS got a sudden offset change. There may be a strong coupling to ALS electronics through the power line.
We run out of the time at this point. So we blocked the beam going to the QPDs. We also turned off the RF QPD interface. These measures eliminated the ALS DARM noise increase. So we are now back to the interferometer locking.
To Do
We still need to investigate what is causing this problem. We identified that QPDs are doing something. But we still don't know what is the coupling mechanism.
One possibility is that the DC power supply for the rack is operating near the current limit. We should check this tomorrow.
Other findings
We found that there is about 200Ohm resistance between the bodies of the PD/QPDs and the REFL optical table.
We disconnected the PZT cables on the REFL table, then this resistance became infinite. So there is a ground connection between the optical table and the racks through the PZT cables. However, the removal of this ground connection did not solve the ALS DARM noise problem. After this test, we re-connected the PZT cables.