Summary
I remeasured the RFAM-to-DARM TFs for the 9 MHz and 45 MHz sidebands.
The 45 MHz measurement agrees with the previous result of ~0.1 mA/RAN. However, the 9 MHz measurement is also ~0.1 mA/RAN, which is a factor of 10 higher than what was measured previously. Note that the previous "9 MHz" RFAM measurement was really a simultaneous measurement of 9 MHz and 45 MHz RFAM, since we had no 45 MHz RFAM stabilization in place.
Details
For the 45 MHz measurement, I injected into the error point of the 45 MHz RFAM stabilization servo and measured the TF from the OOL RFAM stabilization detector (which is already calibrated into RAN) to the DCPD sum.
For the 9 MHz measurement, I temporarily replaced the OXCO with an IFR running at 9.1 MHz and +10 dBm. Then I used the spare DAC channel to inject into the IFR modulation port, which was set to 10 % deviation, dc-coupled (which means a RAN of 0.071 for 1 V of input, though I did not measure this directly). The signal from the spare DAC is buffered by an SR560 and sent back into one of the spare ADC channels. Then I measured the TF from the spare ADC channel to the DCPD sum. This measurement relies on the 45 MHz RFAM servo suppressing the resulting fluctuations in the 45 MHz sidebands before they are applied to the EOM; looking at the OOL readback, this seems to be satisfied below 1 kHz. Above 1 kHz, there is a RAN increase of <2 compared to no 9 MHz injection.
Templates live in my folder under Public/Templates/Osc/(45|9)_RFAM_2016-02-08.xml.
In addition, I took noise measurements of the 9 and 45 MHz RFAM spectra.
The 45 MHz measurement is straightforward, since we already have a calibrated, dequeued RFAM monitoring channel. (Actually I used the faster, undequeued IOP channel, calibrated it, and undid the AA filter.) the noise between 50 Hz and 1 kHz is a few parts in 109 / Hz1/2.
We don't have a similar readback channel for the 9 MHz RFAM close to the EOM, so I made a mixer-based measurement by taking an output from the ISC 9 MHz distribution amp, splitting it, and driving both sides of a level-7 mixer. I had 9 dBm into the LO and −3 dBm into the RF, so the LO was being driven hard and the RF was below the mixer's compression point. The mixer IF was terminated and then bandpassed with a 1.9 MHz filter. The IF dc was −135 mV or so.
To read out the noise, I took one of Rai's low-noise preamps (measured to have <2 nV/Hz1/2 input-referred noise) and ac-coupled the input with a 20 µF capacitor (giving a high-pass pole at <0.1 Hz). Then I read out the noise with an SR785. I have not yet verified that the signal is above the noise floor of the mixer measurement.
Finally, I also include the RFAM-to-DARM coupling TFs with the DARM loop undone.
