We did a (weak) health check of the OMC DC PD B and saw no signs of excess noise in the dark state or in the single bounce configuration.
In klog 36277, excess noise was seen on DC PD B following a lock loss on February 3, 2026. In klog 36569, later in March 2026, DC PD B showed no excess dark noise. So did DC PD B have some intermittent noise that went away? Or could it have some damage that only makes excess noise when there is a DC power? We tried to test this by looking again at the dark noise and comparing with the noise in a single bounce configuration (IMC locked, ITMX misaligned).
Dark noise:
9/6/2026 05:38:30 UTC dark noise: shutter closed, 1 stage whitening transimpedence 400 ohm (to match settings in klog 36569)
The DC PD dark noise agrees with previous measurements made in klog36569. No excess noise in DC PD B.
Single bounce:
OMC locked with single bounce configuration, OMC DCPD settings the same as above.
9/6/2026 05:47:40 UTC Single bounce with 7.8mW on DCPD B.
The noise here of 1.4e-7 mW/rt(Hz) is higher than shot noise (7.8mW should give a shot noise of 5.4e-8 mW/rt(Hz)).
To double check that this is not shot noise, we increased power by a factor of 2 and look at how the noise scales.
9/6/2026 05:56:20 UTC Single bounce with 16.2mW of power on DCPD B.
The noise scales close to linearly. Turning the power up a factor of (both in mW): 16.2/7.8 = 2.08 the high frequency noise went up a factor (both in e-7 mW/rt(Hz)) 2.8/1.4 = 2. So there is some noise on the light hitting the DC PDs that’s scaling close to linearly with power. We’re not sure what the source of that noise is.
So we try to reduce the linear noise by reducing the power on the DC PDs. Locked OMC with IMC output of 0.9W.
10/6/2026 00:21:45 UTC Single bounce with 1.6mW on DC PD B.
Using the photodiode powers for DC PD B, we can make some estimates for the various noise components, comparing the 7.8mW and the 1.65mW single bounce cases.
The power (mW) reduced by factor 7.8/1.65 = 4.7
The unknown linear noise should now be at the level 1.4e-7/4.7=2.98e-8 mW/rt(Hz)
The shot noise for 1.65mW should be 2.48e-8 mW/rt(Hz)
The dark noise was measured to be 1.4e-8 mW/rt(Hz)
So for 1.65mW on DC PD B, the quadrature sum of dark+shot+linear = sqrt(2.48^2+2.98^2+1.4^2) = 4.122 e-8 mW/rt(Hz)
And we measure just a little higher, about 5e-8 mW/rt(Hz). Pretty close to shot+dark+linear noise.
From these tests, there is not any evidence of excess noise in OMC PD B. But the unknown linearly-scaling noise on the light in the single bounce configuration isn’t allowing us to make a shot-noise-limited test.
Some notes on the measurement setup:
1. For the single-bounce configuration, the following suspension states were requested:
PRM: MISALIGNED_BF
ITMX: MISALIGNED_BF
ETMY: MISALIGNED
SRM: MISALIGNED
All others: LOCK_ACQUISITION
2. In the case of locking with 7.8 mW and 16.2 mW on DC PD B, OMMT2 and OSTM were actively controlled to optimize their alignment using the in-vacuum OMC QPDs, whereas no alignment control was applied during locking with 1.6 mW.
This is because the alignment control using the in-vacuum OMC QPDs did not work well due to the very low power on the OMC QPDs; however, this is expected to have little impact on the results.