Engineering 2011AprMay
Daily Log:
2011 April 24:
Xiao filled the dewar before FB arrived, and pressure seems high (3e-6). After refilling both dewars, the pressure went down to 6e-7.
Did the alignment on beams 4, 5, 6 so as not to impede on CLIMB run (+ we need to measure the fringe position again now that the IR5 repeatability problem has been cleared).
All computers had to be rebooted due to a long power outage this night, including lothlorien, wolverine, zoot. Had to restart all servers (shutters, etc.). Green laser is completely disaligned, need to restart crtscrut again to launch server to use pico to realign it...
2011 April 25:
Full alignment with 6 beams done, may look for missing offset B1B6 this week. Started to work on shutter control with RMG, and on loading/saving config files from IRCam. Also starting to review all possible issues when trying 6 beams, FT not ready, need to double check the bootstraping code too.
2011 April 26
Reinstalled CHARA libraries properly. The code now compiles cleanly, everything is on CVS including CHAMP code. Finished the addition of shutter code, working and compliant with CHARA code. RMG modified the alignment sequence for 6 beams. Added the bootstrap for 6 beams (see CHAMP:Documents for the crazy equations...). Started to implement the wraparound carefully.
2011 April 27
RMG finished adding the 6 beam sequence and tested it.
FB worked on trying to integrate the pico controls to the align gui. This took a while but I managed to integrate the controls into the gui, and they do work. However there are cosmetic issues to solve... Also the idea was to be able to use the PicoGui.py code directly, using the equivalent of a C #define depending on whether we use the standalone version or not (in particular we don't want to launch the zaber code in the align gui!). Once this is done, though, the first phase of the alignment gui will be complete (for the second phase we would need to calibrate the picos automatically, and have a close loop positioning of the spots, which should take only ~2 days to implement if we think ahead).
FB also modified the FT gui in prevision for the next run, fixing display issues when using 6 beams. Haven't had the time to continue on the wraparound as we needed the server to run most of the time.
Plan for tomorrow: finish the wraparound and finish integration of the picos to the align gui.
Possible plan for the next days: OPD maps ?, attempt to find the offset for BC61 again
2011 April 28
Fixed the following cosmetic issues with the align gui : invisible icons on devel@wolverine or ubuntu ssh, picos menu size, optimized viewport sizes, optimized icon sizes. Added greying out buttons during alignment sequence for 4 and 6 beams. With automatic shutter/quadrant/pico selections implemented, now the user just has to click on the arrows.
The spooler code should be able to add most of the CHARA infos soon (uv coords, shutters, ). For the moment I implemented Shutter info only, but the original code from MIRC has a bunch of code which should be easily added back. Need more work on this obviously, but I've prefered continuing the wraparound coding... the main parts are in place, but still a bug or two to find as fake fringes do not get estimated perfectly.
Note: what about installing a more recent version of gcc in the far future, as vectorization in gcc4.4 should really improve performance.
2011 April 29
Not feeling too good at all today, so it took me a while to finish the wraparound, but this works. Cleaned code in the Align gui that was useless (lots of junk from the template that was used for MIRC gui). Tried to find the offset for BC61 but no luck (barely enough time to make a real attempt)....
2011 April 30
The wraparound introduced a curious behaviour for beams 5 and 6, checking this. Murderer on loose on the mountain, so we had to evacuate the mountain... Wasted most of the day as I now need the server for most of the remaining work. Crashed at Theo's place (this happened to be his birthday).
2011 May 1
Back on the mountain. CHAMP pressure increased from 6e-7 on the 29th to 5e-5. Both dewars only half-full, so now need to wait for pressure/temperature to settle down. Worked on the GUIs a bit to clean up things. More tests on wraparound with 6 beams and fake fringes, phase estimators ok, gdt seems ok too. Started pumping MIRC from pressure = 2e-1 at 11:30 down to 1.3e-3 at 18:30.
2011 May 2
No luck on B6B1 again.
Tested coherent integration code on various fake fringes, out of array indexing due to recent changes from wraparound and other bugs fixed.
Committed changes to CVS for champ server, uploaded GUI code to wiki.
Code priorities for next time:
Server:
- Better debias, need to discuss with JDM.
- Kalman, already all the ingredients for this now.
These two require live DL fringes to test.
- I should examine all these calls to champ_bs, there seems to be too many of them.
GUI:
- Automatic "coarse" alignment if time.
2011 May 15
Back to CHARA, realigned for 6 beams. Found fringes after a bit of mirror tweaking.
Started testing the wraparound on real fringes, I am tracking some issues.
2011 May 16
Fixed issues with the wraparound (bad realloc of ChannelLastFlux + offset problem).
There was an issue with beam sampler server refusing to launch, fixed after theo investigated (computer refused to boot with serial plugged in).
I could not find the fringes B1B2/W1W2 at the usual place but I suspect where they should be thanks to fluctuations of flux. Hopefully this is not bad piezo again :-(
champdev and devel now use secure rsa passwordless ssh and sudo whenever possible, making debugging less tedious.
2011 May 17
Data taken in directory 2011May18/
chamo_00001.fits and champ_0002.fits : non-moving fringes at standard mode rate
chamo_00003.fits and champ_0004.fits : moving fringes at standard mode rate, vel = 0.001 mm/sec
chamo_00005.fits : dummy
champ_00006.fits and champ_00007.fits : non-moving at 6.61465478 Hz
Nreads 50
Config: 16/ 16/ 7 / 22 / 50 / 42/ 21
FT; 6.61465478 Hz / 10.5 / 13.2 / 0.05 / 0.0
champ_00009.fits and champ_00010.fits : moving fringes at same 6.61465478 Hz, vel = 0.001 mm/sec
champ_000011.fits to champ_00019.fits : junk
champ_00020.fits and champ_00021.fits: non-moving, nreads = 25, FT 12.9178 Hz
champ_00022.fits and champ_00023.fits: moving, nreads = 25, FT 12.9178 Hz, vel = 0.001 mm/sec
Need to retake data as spooler had a writing bug
