Looked at the spectum of th 50W CO2 lasers on the fast VIGO PVM 10.6 ISS detectors when the CO2 is locked (using the laser's PZT) and unlocked/ free running: time series attached. Small differences <6Hz, see spectrum attached.
Gabriele, Camilla.
We are not sure if this measurement makes sense.
Attached is the same spectrum with the CO2X laser turned off to see the dark noise. It appears that measurement is limited to dark noise of the diode above 40Hz. The ITMX_CO2_ISS_IN_AC channel dark noise is actually above the level when the laser is on, this doesn't make sense to me.
- Is there anything strange happening in models or chassis? No.
- The filters in the DC _OUT channels include a CountstoVolts gain(0.000610352) and UndoGain gain(0.00196078) which is 1/510 from the PD electronics D1201111 as expected
- The AC _OUT channel includes a CountstoVolts gain(0.000610352) and de-whiten zpk([20], [0.01], 0.011322, "n").
- There is nothing else in the h1hsccs model.
- Looking at the TCS block diagram E1100892, the ISS PD outputs go though the TCS ISS and Interface Chassis D1300649 (D1300015 shows there nothing happening to the signals inside this chassis), before going to the ADC. So the model should see the signals as if they were coming straight from the PD DB9 output + counts to volts gain.
- Does DC value of 7e-3 V make sense? Yes.
- In CIT#389 saw saw -1V (DCMON so gain of 255) for 60mW of laser power (can jump ~by a factor of 2 dependent on wavelength/polarization/how the laser is feeling). From this, expect 1mV from diode (before electronics gain) is 15mW of laser power. So that the DC value measured of 7e-3 V would be around 100mW of power on the PD (within a factor of 2 is 50mW to 200mW). This makes sense as we would expect around 250mW on the ISS PD (50W out of laser x 99/1 BS x 50/50 BS).
- Does Dark noise value make sense: Unsure.
- In CIT#541, Gabriele shows DC dark noise values of 4e-7 /rHz, this is straight out of the diode (DCMON so x gain of 255) so the comparable value to our _OUT channels would be 1.5e-9 /rHz. This is a factor of 3 different from the 5e-9/rHz dark value we measure at LHO on the DC OUT.
- Debatable math: The power noise of the diode is 1e-8 W/rHz (from Keita's computations attached to CIT#549). So the RIN should be noise-limited to 1e-8 W/rHz / 0.250W = 4e-8 /rHz. We are measuring down to 5e-9/rHz on the DC channel which is a factor of 10 lower than expected possible
Gabriele and I checked that the H1:TCS-ITM{X,Y}_CO2_ISS_{IN/OUT}_AC filter: de-whiten zpk([20], [0.01], 0.011322, "n"), is as expected from the PD electronics D1201111, undoing the gain of 105dB with the turning point around 20Hz, foton bode plot attached.
This means that both the AC and DC outputs should be the voltage out of the photodetector before the electronics, where the PD shunt resistance was measured to be 66 Ohms.
