Checkout the latest noisebudget. 500 Hz peak is gone now. This could be because of any of the changes I made in past 2 days:

1) NPMC servo gain increased. I changed R8 from 9.09k to 51 Ohms. That increased the gain by a factor of 200 which is equivalent to 46 dB. The slider for AD602 gain decreased by 23 dB to remain below oscillations wiht a 1dB gain margin. So maybe this reduced the noise a lot more (even though it is still present)

2) I did replace a normal beam dump (91,45) on the reflection path from North PMC with a triangular black glass cavity and I oriented another one (94,30) just before NPMC blocking rejection from a PBS as it was oriented horribly wrong.

3) We also changed the detector to NF1811, but that hasn't resulted in changing this feature in the past.

Note that the expected noise has now decreased roughly by 3 times in the range 200 Hz to 500 Hz. See CTN: 2423