Engineering 2012MayJun

Computer OSes

• May 27: Debian Lenny source packages were moved to archive.debian.org. FB updated the packages to reflect this.

• May 27: aptitude on wolverine found more than 65535 (16bit limit) packages to update, causing a bug. FB updated the apt cache size, then updated the distribution source by source.

• May 27: current status = we have the final versions of all Debian Lenny packages (gcc 4.3.2, python 2.5.2, pygtk 2.12.1, gtk 2.12.12)

• May 28: also upgraded lothlorien to ease work on CHAMP

CHAMP -- hardware

• May 26: Larry started pumping CHAMP

• May 27: vacuum ~ 2e-3

• May 28: vacuum reached 1.0e-3, so not quite ready for pumping.

• May 29: CHAMP cool, pressure ~7e-7

• May 29: Despite attempts to reproduce hardware crashes/freezes from last run, none have happened so far.

CHAMP -- software

• May 28: fixed CVS that had been erased from ctrscrut due to a HDD failure in Jan/Feb. Fixed all memory leaks on spooler and astropci reported by cppcheck, especially realloc issues. Reintroduced star info and uv info to spooler, hopefully compatible with CHARA libraries (tests TBD).

• May 29: worked on spooler code, now fully functional, i.e. star + uv info (though this uses partially the old method). Tried unsuccessfully to reproduce the "negative ftnum bug".

• May 30: cleaned up a lot of the python code and debug of automatic calibration -- would need consecutive spots to fully try.

• May 31: continued Pico automatic calibration and automatic spiral search. Spot detection routine not robust yet but spiral search works on single spot. Made backup of code + added to CVS the new spooler code.

• Jun 1-2 : been working on the aligment GUI, fixed many bugs, see

https://www.astro.lsa.umich.edu/intra/research/monnier/internal/index.php/CHAMP_bugs_to_fix

• Jun 2-12: code for new pzts, cophasing with MIRC (see below).

• Jun 12: W1(B1)-W2(B2), ref= W1, delay line fringes W1: 2.200000m, W2: 2.007073, gain is +0.25 on B1B2 to track fringes.

Thresholding tests:

• lowering SLtoS threshold at 2-3 instead of 4 helps to not loose weak fringes.

Final final cophasing of MIRC+CHAMP with CHARA

Work on June 10, 2012 (JDM/XC)

REFS: MIRC IR6 87750, CHAMP IR6 is not adjustable

BEAMS RETRO MIRC CHAMP

5&6 5.720 mm IR5: 55250 IR5: 136685

4&5 5.841 mm IR4: 84450 IR4: 146754

3&4 5.375 mm IR3: 58262 IR3: 155554

2&3 5.940 mm IR2: 77030 IR2: 171589

1&2 5.024 mm IR1: 68580 IR1: 179630

6&1 DL below ---- BC: 44300

DL: W1(B1)-W2(B6), W1:1.00000m, W2:1.222532m

Starting 5:07pm

Prep: Homing RETRO

---

BEAM 5-6

MIRC Quality Control checks

Beam 5: 66 cts peak / 732 cts, FWHM 7.6 mu (Fluxes here are xchan / fringe quad)

Beam 6: 51 cts peak / 642 cts, FWHM 7.1 mu

FRINGES (VIS~90%, very little dispersion):

ir5 67750, ir 6 87750, retro 4.723.

Goal: FIX IR6 87750, RETRO 5.720mm, adjust IR5: 55250 (CALIBRATION: 12.5 microns IR% per 10 microns of RETRO)

CHAMP alignment:

Beam5 image quality is great.

Beam5 R6 comment says "image clean" but actually two spots. I will choose bottom one (top one seems to lead to failure)

Beam6 L6 image looks very elongated... vignetting? ugh.

Looking for fringes... FOUND: champ IR5 136685, RETRO 5.720

---

BEAM 4-5

MIRC Quality Control checck

Beam 4: 85 cts peak / 950 cts, FWHM 6.7 mu (Fluxes here are xchan / fringe quad)

Beam 5: 75 cts peak / 860 cts, FWHM 6.6 mu

Fringes (vis ~ 70%, detectable but small dispersion)!: MIRC-IR4: 84450, RETRO 5.841

CHAMP Alignment:

Problems getting all spots to overlap. It hink the soluiont was to choose different spot for B5 R5... champ ir4: 146754, RETRO 5.841

finished 6:45pm

---

BEAM 3-4

MIRC Quality Control checck

Beam 3: 130 cts peak / 1240 cts, FWHM 7.9 mu (Fluxes here are xchan / fringe quad)

Beam 4: 65 cts peak / 770 cts, FWHM 7.7 mu

FRINGES!: (VIS~60%, no obvious dispersion) MIRC IR3: 58262 RETRO 5.375

CHAMP FRINGS: CHAMP IR3: 155554 RETRO 5.375

---

BEAM 2-3

MIRC Quality Control checck

Beam 2: 110 cts peak / 1200 cts, FWHM 6.5 mu (Fluxes here are xchan / fringe quad)

Beam 3: 107 cts peak / 1180 cts, FWHM 7.5 mu

FRINGES!: (VIS~90%, most dispersion so far...) MIRC IR2: 77030 RETRO 5.940

CHAMP FRINGE: CHAMP IR2: 171589 RETRO 5.94

---

BEAM 1-2

MIRC Quality Control checck

Beam 1: 98 cts peak / 1250 cts, FWHM 6.9 mu (Fluxes here are xchan / fringe quad)

Beam 2: 113 cts peak / 1100 cts, FWHM 7.7 mu

FRINGES!: (VIS~85%, little dispersion....) MIRC IR1: 68580 RETRO 5.024

CHAMP FRINGE: CHAMP IR1: 179630 RETRO 5.94

---

BEAM 6-1

Delay line fringes on W1(B1)-W2(B6) with W1 at 1.000000 m and W2 at 1.222532 m.

MIRC IR1 68580, IR6 87750

CHAMP BC: 44300 (-1000 on BC is equivalent to +50 um on W2)

Data Analysis

• UT2012Jun14

Looked at BEAM Ratios for MWC 361 on sky performance.

Beam order; W1 S2 S1 E1 E2 W2

Note that B2/S2 has the worst quality. B5/E2 is the best.

with a camera rate of 250Hz:

Based on shutters, approximately champ files 00081 to 0096

	B1 W1	B2 S2	B3 S1	B4 E1	B5 E2	B6 W2
Total	149.2	54.1	87.6	69.3	112.7	91.3
Li	0.37	0.28	0.28	0.39	0.35	0.36
Ri	0.23	0.27	0.23	0.24	0.33	0.25
Li+1	0.14	0.12	0.06	0.12	0.05	0.17
Ri+1	0.26	0.33	0.44	0.25	0.27	0.22

(JDM)

based on these funny results on B3 especially we are looking carefully at engineering beam. we arechecking flux ratios for beam 3 and image quality.

Nightly Log

• UT June 8, 2012

got lab fringes on beams 1,2

got sky fringes on beams 1,2 (w1-s2).

mirc found fringes with S2 at 5.8mm, we had to move s2 to 7.1mm, so a change of 1.3mm (hmmmm?)

but we tooks some data 2012Jun08/champ0001-0002 and later background files.

---

• UT June 14, 2012

goal track fringes with CHAMP.

problems: software accumulator for envelope was corrupted by a recent change. spent time tracking that down.

W1 telescope has constant tracking problems between 30-42 degrees or so. frustrating.

MIRC found fringes on one CAL but not on MWC 631 (except one cross fringe).

Now trying to get fringes on a new cal.

MWC361 had about 50 counts at 250Hz. but coudln't find fringes

problems with tracking W1 and S1.

12:15am

Now going to a cal that is closer to mwc 361: hd 205776

MIRC is tracking fringes fine. CHAMP found fringes for champ IR4: 147084 IR5: 136845. BC: 44660 IR1: 179550

GAINS: BC1 +0.25, BC2: -0.25 BC3: +0.25, BC4: -0.25, BC5: +0.25, BC6: -0.25

MIRC: 00291 champ 0004-005.. SHUTTERS: mirc 309->408

champ is taking data during shutters.. will have to sort this out! at 0006-> 0040 or something.

note: W1 was not tweaked at all so the fluxes will not be representative of performance except E2W2 were optimized.

hd 200775

W1 tracking fine above 41 m.

mirc is optimizing.

MIRC: S2 5.89 E1 2.38 E2 1.92 W1 1.25 W2 ref S1 XX

FT offsets: BC1 -.325 BC2 BC3 XX BC4 XX BC5 .209 BC6 .060

We found fringes between W2-W1, e1-e2, e2-w2.. but not for w1-s2. We

We could never find S2-W1 fringes. not sure why. we found on calibrator but could not find on mwc 361. probably too resolved !!

taking shutter sequence.. mirc data + champ (champ0081=>).

plan is to next re-optimize fluxes and search for fringes on mirc.

note -- signal on L2 was VERY WEAK.. maybe alignment? beam 6 L1 also weak

maybe mwc 361 is resolved W1-S2 baseline... try another option?

Switching S1 <-> S2

back on sky. we changed pops too .

3:43am. peaking up on the

found fringes on cal. found fringes on mwc361 but NOT S2-E1 .. no fringes again . BC1: -0.756 BC2 XX BC3 XX BC4 (cal) -1.071 BC5 -0.023 BC6 0.227

but no S2-E1 fringes... mwc 361 fringes must be too weak for long baselines!

we took 250Hz data champ 0165-182 as a test to see if its better to go slower. so we are taking more data at 125Hz. 70mu scans weres saturateda bout 3/4 of the way through on mwc361 (k=4.6) -- so not too practical.

(sorry these logs are bad -- but wiki logs suck and I am tired...!)

==

• UT June 15, 2012

cal: HD 156928. got fringes on mirc and champ -- cophased.

now trying mwc 275.

we got fringes easily on S1 S2. saw weak fringes on W2W1 and E2W2. (E1 was out of delay).

at 250Hz we had about 60-100 counts on each channel.

our algorithm did not 'see' or lock on the weak fringes. but maybe we can tweak the algorithm.

we tried 500Hz to see if the SNR improved. but seems ot have gotten worse. we no longer saw the weaker fringes.

tomorrow we will improve our fringe detedtion algorithm for intermiittant weak fringes in a way more like how classic/climb works.

we took some 500Hz data for later study on mwc 275 but .. it caused some odd display problems on ft gui so not sure if it saved properly.

• UT Jun 16 2012

using H pass dicroic for beams 1-2. using HK pass for beam 3-6 to allow test of J band mode.

HD 156928 (JHK 4.1 4.0 4.2): all 6 fringes !! CHAMP 0005->0007.

in file 8 we will do a by-hand shutter... BG.. B1 B2 B3 B4 B5 B6 ALL

NEXT we are putting j band bfilter and the 1.3mu SWP on beams 1-2 only. the others are close dbeause they are swamped with laser light.

MWC 275 (JHK 6.2 5.5 4.8). CRAPPY J band fringes ! in B1 B2 (BC2) !!! Can't keep the fringes .. but we are going to take a little data for posterity.

we are using ~40micron opd amp..

champ 00009-> 10.

Trying a longer stroke mode and also at 500Hz to see if the fringes will stay put for tracking before trying MWC 275.

fringes look cleaner.. but still FAINT!!!

WE CANT SAVE DATA! it seems for this longer stroke the spooler is hanging the system.

more debugging is needed !!

now on MWC 275: we have very few counts .. at 500Hz on BC2 (w1-s2) , 6 counts ! (yes.. 6 counts). on the calibrator we had about 50 counts per channel. This does not look good.

Next up. v1295 Aql Calibrator.

1:19am

HD 187923 using 50 micron, 250Hz scanning.

Found 3 fringes but the others eluded us... That is IT for the 2011 CHAMP RUN.

• good things:

new PZTS work really well

new 1.3micron filter works

first J band fringes

tracked 4T for MIRC on mwc 361

software more robust

good cophasing with mirc and chara system -- well documented

• bad things:

couldn't track mwc 275 firnges -- saw 3 out of 5..

seeing did not cooperate at the end

tracking algorithms not optimal for super faint fringes (come and go)

stubborn software bugs (accum, saving large scans,

j band mwc 275 looked hopelessly weak

• to do:

check fluxes using the saved files. how far away are we from performance of climb? this is crucial to know what to focus on next run !

MIRC high-res Engineering

2012Jun25

Camera setting: rows: 72, cols: 252, row off: 69, col off: 2, Nreads: 5, # of frames: 350., frames per reset: 350. We deinterlace routine in real time code so that it will read the top two quadrants as the mirc and PCs quadrants. On the python gui side, nothing requires changing and everything looks exactly the same as before.

Note: under high-res mode, the # of rows should be 72 rows (in camera setting, it should be 72*2=144), but the integration time is 48ms which is too long, so we reduced the # of rows to half, and the integration time is now 24ms.

Method: we move fringe to one of the edges as the beginning point, and then take 10 files. Then move 30 micron each step, and take 10 files until the fringe reaches the other end.

Experiment 1: W1 (beam 1), ealing: 68580, delay line offsets on MIRC gui: 1000 mm W2 (beam 6), ealing: 87750, delay line offsets on MIRC gui: 1220.4mm file number: 83 - 142

Experiment 2: W1 (beam 1), ealing: 68580, delay line offsets on MIRC gui: 1200 mm W2 (beam 2), ealing: 77030, delay line offsets on MIRC gui: 1005mm filen number: 168 - 242