Because the Staging Building quad test stand (X1) has very old electronics there has been some doubt as to whether the open light counts for OSEMs were comparable between it and the new hardware in the LVEA. The most conspicuous difference is that X1 has -v1 versions of the AA boards (D070081) which have a sign flip relative to the latest -v5. But they also have inputs with an impedance of only about 1K, which could conceivably have loaded the satellite amps and reduced the signal. And something in the X1 chain seems to saturate before the ADC because before the satellite amp gains were reduced, no OL counts below about -31500 were observed, even for BOSEMs that gave 32K on H2.

To test this we (Jeff B and I) remeasured on H2 six AOSEMs that had recently had their OL counts recorded as part of electronic noise testing on X1. Five were chosen at random but we made a point of including SN367 which had the largest OL observed so far at -30819.

We adapted the procedure for AOSEM noise testing in E1100860 to use the M2 channels on FMy because the cables for those positions were readily accessible. However we encountered various computing issues on cdsws10 and had to improvise at several points. MEDM wasn't working (whited-out fields) so we couldn't confirm the settings of offsets and filters, so we abandoned trying to take noise spectra. Partway through we started to see digital noise on M2-LR, so we ended up using only M2-LL. And shortly after that, tdsavg stopped working partway through, so we resorted to reading OL values from DataViewer. However we ending up with 6 usable readings. 

It turns out that the H2 values are almost identical to those on X1: about 0.8% +/- 0.3% high. And the OL for the "hottest" AOSEM is not significantly depressed on X1 relative to the others, so the safe range apparently extends as high as -30819. See attached results spreadsheet and printout.

This measurement could usefully be redone after the issues with cdsws10 are fixed to see if there is any effect on the noise spectra. (AOSEM SN367 had curiously low noise when measure on X1, which might be an artefact of running it close to the ceiling of the electronics.)