Thermal cycling primarys with LBC FPIA
Observer(s): A. Rakich, K. Boutsia and O. Kuhn at start of the night (LBTO)
Telescope Operator: S. Allanson (LBTO)
Telescope Support: M. Midkiff (LBTO)
Instrument Support: none
Attempted tuning of RED FPIA parameters for Z11: phRedcorrection and RedfinalZ11zero
, to restore Red FPIA to same stability as Blue. This is completing work that had to be left incomplete in March. However seeing and clouds precluded this, switched to getting hot mirror data to complete the set from last night. Reset astigmatism gain from 200 to 100 and noticed a dramatic improvement in astigmatism stability. Pending a few more tests will be recommending this new gain factor for science, should remove one major source of FPIA slow or failed convergence.
All times are UTC
unless otherwise noted.
Clouds clear, humidity low, open. Doing FPIA then rangebal.
Have converged, ran FPIA FIRST as a sanity check twice to check the backout. Seeing is ~ 1.5". Now changing two parameters to a) increas the size of the central hole in RED that FPIA converges to, and b) back out the extra Z11 that results from (a).
Changing sphRedcorrection to -325 and RedfinalZ11zero
to -375 nm which backs out the +300 nm Z11 we get from the first change.
Taking a pair of extra focal red and blue images for a baseline before making adjustments: 85613b EF and 85703b IF and 85609r EF and 85658r IF.
Seeing is 2" , not good enough for this sort of test, switching to thermal cycling M1s to get hot mirror data for LBC that we didn't get last night.
Have removed above mentioned changes to RED Z11 and have changed astigmatism gain from 200 to 100 so we can compare astigmatism stability to data we took last night. Not at the circumpolar field but not changing elevation much through the couple of hours we have remaining.
clouds gone, continue thermal driving data collection.
Using the 50% damped gain for astigmatism has made a marked improvement in astigmatism stability, have not had a single case of an astigmatism value out of convergence range for 6-7 iterations.
Notice that on RED we get coma oscillations of the order of 1000 nm because the central hole is sometimes fit to the actual central hole and sometimes fit to a dark permanent feature near the top of the Red spider. Issuetrak 3504.
Seeing has improved to ~ 1" but given limited time we will leave the RED Z11 tweaking to a later technical night and continue thermally cycling primary mirrors.
The astigmatism improvement is so great we are proposing to leave this gain factor of 100 in for science, contingent on some tests where we converge at a high elevation then preset to a low elevation and do FPIA to make sure we don't converge too slowly. Another "feature" that we may decide to build into FPIA, contingent on more testing, is to use a higher astigmatism gain on the start-of-the-night "first" type of FPIA (which also preloads more spherical).
So far in 36 iterations we have only had a single occurance of a Z5 or Z6 value on red or blue out of the convergence criteria, and that was assoiciated with a big coma/focus jump (see issutrak3504). Typically in the past we would have an out of range value every two or three iterations.
Notice pupils looking like they have spherical as per last night. C00 and do FPIA again.
Example of blue bad pupil fit 105738b. Pupil fitting algorithm needs to handle pupils like this better. Also 110035b.
closing for dawn.
- 05 Jul 2011