2011 September 18: LFBG and LBC checkout, plus MODS restart

S. Allanson, T. Shih, M. Midkiff, O. Kuhn
J. Hill in Tucson
F. Pedichini, M. Centrone, R. Pogge, R. Stoll

Power outage all day from about 8am to 4:30pm. At 19:00, still restarting network and systems.

From about 4:00am, Fernando & Mauro worked on the LBCs.
  • Swapped in spare controller for Blue Tech chips, but this did not have an effect. The noise problem arise outside the controllers, shielding, grounding of the cables, e.g.
  • New version of LBC software which should make a difference in the darks.

02:03 Started LBC and started a sequence of Darks. 020406 B and 020404 R. The Blue Darks (180 s and 360 s) show the pattern still. Fernando notes that red is skipping the flush (from the windows computer he sees this). On the blue side, the UI says "hard flushing..." but I did not see this for Red.

03:00 Fernando modified the program. Both did the hard flush and flush (you have to watch the windows PCs not the UI which only reports snapshots of what is really happening. Blue shows only slight pattern now, Red shows strange stripe from cols 2000-2099. 25921 25913. First two red darks show this dark column band on chip1, but last one is good. Last Blue dark good too.

03:09 25 biases. Look good, now trying again the set of 3 180s darks. Lost first Blue image of sequence. Blue RPC error. Red SPC error
2011/09/18 03:14:00.505418 E B CAMERA            RPC Client/Host communication failed retval:5 [src/camera/camera.c:874]
2011/09/18 03:25:13.743000 E R CAMERA   CCDCTRL  "timeout reading SPC status" (code:68) controller error [ccdctrl.c:2432]

03:30 OB and User Interface was hung. We had no Blue and 2 Red images from the previous Dark OB. lbckill/lbcstart and restart OB now: 033252 B and 033248 R. These are looking OK now. 10 biases immediately following.

04:00 Wait for 10 minutes - so safety off done at 04:13 After this, hard flushing should be triggered. checked - it did do hard flush and flush. On the first blue dark, both patterns are seen. Pattern may be linked to safety off. Fernando says Blue and Red may do the readout differently - he'll set blue to be like red.

05:30 Fernando changed the software to make the Blue side do what the Red side does. About 1 month ago, he removed the pulse during integration on Red, but not on Blue. Pulsing the amplifier during the integration prevented a gradient on readout. The amplifier was pulsed also before readout - that remains. (This is all not so clear to me). After this change, we did not see the pattern in the Blue science.

05:30 Going to test the 10-min safety off again - so as of 05:28 LBC is idle. Wait 10-15 minutes and run the 3-dark OB.

05:39 safetyoff was triggered at 05:35. Running the "fast" OB which was already loaded and will take a 180s dark on both sides. It did a hard flush. 054159 B shows same pattern BIases ending in 060619 for Red show the vertical stripes. I didn't see earlier - whether we had such high counts on some like yesterday.

06:49 safety off triggered at 06:16. Crazy red biases again with vertical bands... 065251. But subsequent darks are better - 25 biases, followed by 3 darks. These look fine (073944 B).

08:00 Start another series of 3 darks after a "safetyoff". First pair, both blue and red are bad. See 080123 B and 080111 R. Patterns diminish as the 3 dark set completes, and as a set of 25 biases is taken, however Red, between 082940 and 083006 re-gained this pattern of vertical banding, which continued to the next and last one. A new set of 25 biases was started after this. This new set started with the pattern (left over) but then it diminished pretty quickly. Both look flat.

08:45 Letting current series of biases finish and will trigger "safetyoff" again. Taking new set of biases 090025 R and 090036 B. These look good. Stopped and ran set of darks - hard flushing again: 090358 R and 090409 B. Pattern is seen in on 090409 and diminishes to second one. So --- biases look ok after the safetyoff, but the first dark is still messed up. And we see that hardflushing and flushing occur before the bias set and the dark set.

09:15 LBC swing arms in place. Start of LBC checkout.

09:19 Slew to WT10_353 to collimate/focus and then correct pointing. The new collimation models had been in place. I reverted to the old ones (which were dated 20110903), and copied the new ones to ?XPMLBCCollimation.20110918.dat. Steve re-initialized.

09:28 Run dofpia - syntax error in lbcfpia.pro print statement. It occured where I tried to write the channel name to the screen.

09:40 Initial dofpia crashed - pupils too small on blue and red and on red strong coma. T=5.8 C. Mirrors in equilibrium. Steve gives Z_go = -1mm to both sides. Helps, but DX has huge coma and FPIA fails - does not even get to Blue. Steve gave DX: Z7=+2000, Z8=+1000 and SX: Z7 = +1500. Now FPIA is analyzing pupils.

09:54 WT10_353 observation: 095637 R and 095641 B. lbcrangebal. Redo WT10_353 observation. 100005 R and 100008 B. Done.

10:03 Slew a bit to offset from star and run dofpia. Will run superfoc to confirm no tilt. Superfoc for rV PA=0: 100937/100934... B/R... focV0.list and focr0.list. Data obtained. Need to analyze.

10:20 dofpia - converged quickly. Run IB superfoc: 102202/102159
Run rV superfoc: 102612/102613
Run ig superfoc:
Run rV superfoc again:
103611 is crazy red image with vertical bars (see attachment)

10:43 Running dofpia but Red pattern persists and only two pupils, out of the pattern, are found. Will stop dofpia and take biases. Pattern did not disappear in biases.

10:51 Slew to Triangulum. Will try biases here again. The pattern is worse - see 105515 and previous image.

11:16 Red science image is not useable - vertical banding as fine as individual columns - has to do with response to light?, see 111426. But there is also horizontal trailing/bleeding for bright stars.

11:50 dofpia and then rV superfoc. 115117/115113 elev=55 deg.
YS superfoc
rV superfoc

12:05 Slew to new field, auriga open cluster at 05:52 32:33. Problems with LBC-Red seem much less in initial 1sec exposure here. elev=75 deg. collimation/focus. Neither blue nor red is far from collimation. Pattern on red is still visible in 16-sec pupil image fpia convergence looked good initially, but then wavered around - we are in twilight. Eventually FPIA crashed, and the red pupils show -ve z22.

12:25 Twilight and too light to continue. Closing.

-- OlgaKuhn - 18 Sep 2011
  • Vertical bars (2x background) on focus sequence image:
    LBCR20110918VerticalBars.jpg
I Attachment Action Size Date Who Comment
LBCR20110918VerticalBars.jpgjpg LBCR20110918VerticalBars.jpg manage 177 K 18 Sep 2011 - 10:40 OlgaKuhn Vertical bars (2x background) on focus sequence image
Topic revision: r2 - 18 Sep 2011, OlgaKuhn
This site is powered by FoswikiCopyright © by the contributing authors. All material on this collaboration platform is the property of the contributing authors.
Ideas, requests, problems regarding Foswiki? Send feedback