Date: Mon, 1 May 2023 12:12:11 -0600
Subject: Re: [sloan-25mlog 6701] Re: 2.5m obslog 20230430 (SJD 60065)
From: "Daniel Oravetz"
>> ===================================================
>> =====>>>>> 04:38:20 Config Loaded: 102378-121680-9758
>> ===================================================
>>
>> 04:41:34 FVC exposure /data/fcam/60065/proc-fimg-fvc1n-0014.fits
>> 04:41:16  fvc_perc_90= 57.9 fvc_percent_reached= 2.0 fvc_rms= 46.82
>> 04:42:05 FVC exposure /data/fcam/60065/proc-fimg-fvc1n-0015.fits
>> 04:41:54  fvc_perc_90= 8.5 fvc_percent_reached= 58.1 fvc_rms= 6.82
>>
>> It failed to clear the Rot axis afetr the FVC images.
>> I used this stop to correct very large Rot fiducials.
>> Then tcc track moved me to new place (I used for the correction),
>> so I re-took the FVC images, just in a case.
>> 8 minutes were lost to it.
>>
>> 04:50:18 FVC exposure /data/fcam/60065/proc-fimg-fvc1n-0016.fits
>> 04:49:59  fvc_perc_90= 61.9 fvc_percent_reached= 1.4 fvc_rms= 44.97
>> 04:50:47 FVC exposure /data/fcam/60065/proc-fimg-fvc1n-0017.fits
>> 04:50:37  fvc_perc_90= 9.2 fvc_percent_reached= 56.4 fvc_rms= 6.55
>
> This sounds strange to me. Will a small rotator motion cause so big
> differences in FVC fitting.
> If robots didn’t move between fvc image 15 and 16, I thought the fitting
> result from them should be
> Very similar, the differences are BIG: rms 6.82 vs 44.97?
>
> Dmitry: Can you please clarify whether the configuration was re-executed,
> or robot moved in any other way before fvc image 16 was taken?
>
> Conor: If there are no robot motions before fvc image 16 was taken, is it
> possible for such large difference in fvc image 15 and 16?
>
> I thought we might be able to continue if fvc loop indicates fibers are in
> position (like fvc image 15 did) without re-do fvc loop even rotator is
> slightly off the position (85K in fiducial means it is off by ~0.3 deg).
> We can just correct the rotator axis, and take cal, setup the field.
There was a 'configuration reverse' between the two sets of FVC loops
which should explain why the second set of rms values was so similar to
the first since the positioners would've been starting "from scratch"
given the 'reverse'.