TCCON > Sites > Caltech/Pasadena > Instrument History

Instrument History

June 2019

  • 06/24/2019
    • After many days of alignment attempts with the "Caltech" InSb detector, Jean-Francois brought down a new test cable to go directly to the voltmeter (since the Caltech one is missing) and his spare InSb detector.  We were unsuccessful at getting any signal from the "Caltech" InSb, but saw light almost immediately with JF's detector and hence concluded that the "Caltech" InSb had failed for some reason.  We installed JF's detector and the methodically optimized the voltage.  We saw (with the NIR lamp):
         1.368 V on InSb wo/filter (gain set to Ref and 1mm aperture)
         0.553 V (4105 counts on OPUS, DC mode) on InSb w/filter
         0.898 V on InGaAs (gain set to A and 1mm aperture; 4105 counts on OPUS)

April 2019

  •  04/08/2019
    • The camera in the Caltech library which captures the projection of the solar image failed. The camera is needed for the mirrors of the coelostat to follow the sun (theoretical calculation does not seem to be accurate enough for our purposes). We replaced the old 'Super Circuit PC-33C' camera by a new '2.1 Megapixel 1080P HD-TVI + 1000TVL CS-Mount WDR Box Security camera'. For that, we had to climb through we window in the Caltech library to reach the camera which is mounted behind the screen where the solar disk is projected. We were able to use the same connectors (Power: AC 24V/DC 12V, Video: Coaxial). We also mounted the old camera's telescope to the new camera. We adjusted the focus by tracking the image of the solar disk by connecting to 'sol' (computer that controls the dome and mirrors) and accessing the camera video via 'http://m7001/view/viewer_index.shtml?id=7'. The image was slighly distorted but Richard should be able to 'scale' it with his software. Weather is not optimal today. Richard will run additional tests in the next days.


March 2019

  • 03/05/2019
    • We noticed that the dome did not open in the last ~three days. We contacted Dick (he is currently on travel though). Waiting for answer. In the meanwhile we are setting the insturment status to not operational.
  • 03/01/2019
    • We ran a UPS test. We pulled the power plug of the spectrometer and the UPS went active. We waited ~5min and the UPS was still working and the battery was at about ~95% (100% initially). Seems to work!

February 2019

  • 2/25/19
    • We replaced the UPS battery following the instructions on page 50ff here:
    • We plugged the UPS into a power outlet to charge the batteries for 48 hours. We plan to run test in 2 days to test the opperation time of the UPS.

January 2019

  • 1/23/19
    • On the advise of the Tech Support at Eaton, I tested the UPS battery by unplugging the unit from the wall.  The UPS initially switched over to battery, logging that it had 93 min of opperation time, and then quickly went dead.  I have placed an order for the replacement battery.
  • 1/22/19
    • The alarm has been going off on the Eaton UPS for several days now.  Tech Support at Eaton suggested that I unplug the UPS from the wall power this evening (with the FTS in hold, of course) to see if the battery kicks in.  (I believe the laser is also on this power strip.)  They were quite surprised that this (original) battery is still working after 8 years and are sending me a quote for a replacement.
  • 1/16/18
    • John noticed a beeping noise from the solar tower in B230.  Inspecting it, I found the UPS beeping about once per second.  It stopped when I pushed a button on the front panel.

December 2018

  • 12/7/18 
    • Came in to find dome closed.  Dick said it looks like the M2 rotation has failed.  Went up and rotated the coupling between the motor and the gear.  It turned easily and smoothly.  Dick restarted and all well ... until 1pm LT, when same problem appeared.

November 2018

  • November 27
    • From Dick after visit -
      "The dome was as dirty as I have ever seen it.  It looked like smoke from the fires came in.


       I vacuumed a bit (I have to get a new bag) and washed the primary.  It had quite a bit of tree sap; how it gets there is a mystery.  Fortunately Nathan refreshed my isopropyl alcohol and it came off nicely. 

       I experimented with the Mirror 1 East West slide and lo and behold it worked! I was able to move almost all the way down and back and in fact the measured load was distinctly below max.   I have to tune the servo drive and re-enable it (this can be done remotely) and hopefully next week, the “noon gap” in the data will be gone."

August 2018

  • August 12
    • Followed the reset He:Ne usage instructions to clear the "Fail" indication in IFSloop.  Laser usage info was:
      2 years, 82 days, 20 hours, 38 mn
      since Wed, 18 May 2016 16:21:51
    • Pumped down the IFS to avoid nightly error emails.

July 2018

  • July 31
    • Zeno failure in the morning
    • The Zeno computer 'sol' failed (common DELL OLOD problem). Hard disk was removed and shipped to Richard (Dick) Treffers on August 1, waiting for replacement!
  • July 13
    • The Caltech algo is updated to collect 2 x NSS=2 low res scans between high res scans in Solar and SolarInGaAs modes and will collect 10 x NSS=2 low res cell scans on Mondays. 

June 2018

May 2018

April 2018

  • April 17-23
    • LN2 Pump did not go active in InSb_Fill
    • Dick here (April 17-18) to work on coelostat
    • Finally cleared water droplet blockage by blowing into pump head through silicon tubing which normally attaches to black fill line.
  • April 16
    • Realigned the FTS with JEan-Francois present.
  • April 13
    • Realigned the FTS, but not confident with results.
  • April 9
    • Zeno failure. Indicated by Zeno stale counter.
    • Hundreds of "Unexpected InSb Not Ready Error. Reverting to InGaAs " messages started coming.
    • Router toured off; IFSloop indicating a laser failure.
    • Restarted the FTS; we still get an HeNe fail (which might be real), but the instrument went back into SolarInGaAs.
    • Just kidding, IFSloop keeps sending emails... IFSloop quit again. I can imagine that the laser error caused the scanner stop and that this somehow causes the emails. Also, I disconnected the LN2 pump from the Herc (Com2) to make sure it's not the pump/temperature sensor that causes the emails.
  • April 5
    • ifsreduce: Invalid analdir argument
    • IFSloop restarted, "sw status reinit" command issued; it seems to be working now.
  • April 4
    • Vacuum failure. IFS_P = 9.51 Pump_P = 0.25
  • April 3
    • Zeno failure. Indicated by Zeno stale counter.
    • Filled the InSb detector by hand, but software does not switch over to Solar (instead of SolarInGaAs)
  • April 1
    • LN2 Pump did not go active in InSb_Fill

March 2018

  • Mar 31
    • LN2 Pump did not go active in InSb_Fill
  • Mar 30
    • LN2 Pump did not go active in InSb_Fill
  • Mar 29
    • Zeno failure.
  • Mar 28
    • LN2 Pump did not go active in InSb_Fill
  • Mar 27
    • Zeno failure.
  • Mar 16
    • Zeno failure.
  • Mar 8
    • Zeno failure.
  • Mar 1
    • Zeno failure.

February 2018

January 2018

  • Jan. 21/22, 2018
    • Dick Treffer cleaned and tweeked the mirrors of the coelostat; must check the alignment.

December 2017

  • Dec 11, 2017
    • Noted that last night's data did not transfer properly and that the system did not initiate scanning this morning.
    • Attempted "sw InSb fill", but LN2P dropped rather than increased and eventually errored out.  The large LN2 tank in lab was empty, so got another one ordered and manually filled the InSb detector.  Auto fill line is removed and presumably is blocked.
    • Network connection was re-estabished after a power cycle of the router.

July 2017

  • Jul 25, 2017
    • The LN2 pumpwas put back into the 50 L dewar this morning, but it would not pump the LN2 over again today.  The pump was removed from the dewar and laid at an angle.  FTS is in Hold.  Problems continued with the blockage until the end of the month.
  • Ju 24, 2017
    • The LN2 fill line still seems to have some sort of blockage, so the pump was again removed from the 50 L dewar.  Also had Zeno failure/Coelostat or driver communication problems indicated by suntracker stale counter.
  • Ju 23, 2017
    • The "LN2 Fill not finished after 15 minutes" error appeared this morning, so removed the pump from the 50 L dewar to dry out overnight.
  • Jul 22, 2017
    • LN2 Fill not finished after 15 minutes
  • Jul 20, 2017
    • Zeno failure, followed by Coelostat or driver communication problems indicated by suntracker stale counter.
  • Jul 7, 2017
    • Zeno failure, followed by Coelostat or driver communication problems indicated by suntracker stale counter.

June 2017

  • Jun 26, 2017
    • LN2 Pump problems, followed by Coelostat or driver communication problems indicated by suntracker stale counter.
  • Jun 21, 2017
    • Started having LN2 Pump problems
  • Jun 15, 2017
    • Zeno failure, followed by Coelostat or driver communication problems indicated by suntracker stale counter.
  • Jun 12, 2017
    • Vacuum failure. IFS_P = 7.30 Pump_P = 5.11
  • Jun 11, 2017
    • Vacuum failure. IFS_P = 6.72 Pump_P = 3.66
  • Jun 10, 2017
    • Vacuum failure. IFS_P = 6.29 Pump_P = 2.26
  • Jun 9, 2017
    • Vacuum failure. IFS_P = 5.69 Pump_P = 0.15
  • Jun 2, 2017
    • Zeno failure

April 2017

  • Apr 25, 2017
    • Problems with the data transfer again last nigt.  Power cycled router and Herc again. 
  • Apr 23, 2017
    • Continuing to have trouble with the communications with almus. It transferred data to ccycle fine Thu. night, ran all day Fri., did not transfer ok Fri. night, and apparently never started up on Sat.  
    • Power cycled router and Herc again. 
  • Apr 20, 2017
    • The system seemed to run fine until the dome closed at 7:30 pm last night. The ifsreduce.log was produced, but the data did not transfer (again) to ccycle properly. Also, could not ssh into almus from ccycle. Power cycling the router did not fix the problem and was forced to power cycle the Herc (and the FTS just for completeness!) After several sw reinit's, I got IFSloop to start.

      However, that the values on the detector seem very low:
      20:54:49 MXY1: 0.071 MNY1: 0.007 MXY2: 0.682 MNY2: 0.147
      20:54:49 MXY1: 0.037 MNY1: 0.037 MXY2: 0.434 MNY2: 0.433
      20:54:49 MXY1: 0.037 MNY1: 0.037 MXY2: 0.434 MNY2: 0.432
      20:54:49 MXY1: 0.037 MNY1: 0.037 MXY2: 0.433 MNY2: 0.432
    • J-F: The logic for setingt the InSb gain is a bit complicated in the code because the IFS125 will refuse to set the InSb gain if the detector is not cold. Re-staring the system when the detector is already cold bypasses the fill procedure, and it seems also the gain setting. Simply using re-init only sets the InGaAs and Si gains. So go to "Hold" and issue the these commands:
      IFS Direct DTC=0x4041
      IFS Direct PGN=3
  • Apr 19, 2017
    • Installed a SIOS laser (S/N 410481110-L/E on 4/17/17).  With the scope, saw that the sine waves disappeared as the scanner traveled from ZPD to MOPD.  So, sent scanner to Back Short Adjust, adjusting the MOPD Laser Adjustment Screws as it moved.  Then, optimized signals at MOPD, and recorded values below:
      Location Laser Vmax Vmin Vmax-Vmin Vmax+Vmin Zero ME
      Front A -250 mV -8.275 V  8.025 V  -8.525 V -8.975 V  0.85
      Front B -2.325 V -8.975 V  6.65 V  -11.3 V -9.25 V  0.92
      Back A -350.0 mV -8.25 V  7.9 V  -8.6 V -9.0 V  0.84
      Back B -2.475 V -8.925 V  6.45 V  -11.4 V -9.25 V  0.91
    • Took a lamp scan vented (ci20170419lsgeeaa.001) and evacuated (ci20170419lsgeeaa.002)
    • Started IFSloop and will look at ILS.


  • Apr 14, 2017
    • Removed Jean-Francois's Spectra Physics laser and Linos beam expander.
  • Apr 13, 2017
    • HeNe laser no longer locks.
  • Apr. 6, 2017
    • Replaced the +5V coelostat computer power supply.  Additionally, did a hard reset of the Herc, power cycled the router, and filled the 50 L LN2 dewar.

March 2017

  • Mar. 30, 2017
  • Mar. 20, 2017
    • Dick could no longer log into coelostat, indicating some serious failure.  Tried various reboots and power cycles, but computer seems to be dead.
  • Mar. 19, 2017
    • Norton first noted that communications with almus stopped 3/18/17.  Matt found dome closed and there were coelostat errors:
      2017-03-19T15:39:50Z ERROR slit slit=closed command=open
      2017-03-19T16:15:53Z ERROR offset m1rot positive limit strike
      2017-03-19T16:15:54Z ERROR offset m1rot positive limit strike
      2017-03-19T16:16:02Z ERROR offset m1rot positive limit strike
      2017-03-19T16:16:13Z ERROR offset m1rot positive limit strike
      2017-03-19T16:16:27Z ERROR offset m1rot positive limit strike
      2017-03-19T16:16:37Z ERROR slit slit=closed command=open
  • Mar. 11, 2017
    • Coelostat or driver communication problems indicated by suntracker stale counter.  The m2rot mirror got stuck.  Matt was at Caltech and went up and twisted the coupler.  This did the trick and system was restarted.

February 2017

January 2017

  • Jan. 25, 2017
    • Vacuum failure
  • Jan. 24, 2017
    • Vacuum failure
  • Jan. 11, 2017
    • Coelostat or driver communication problems indicated by suntracker stale counter.  Issued a Reinit and seems fine.

December 2016

November 2016

  • November 8, 2016
    • After several days of weird behavior with the dome, we rebooted the drive amplifier by lifting the HV fuse holder and then closing it again. (See Dick's email with photos from today.)  Everythings started up after this.
  • November 3, 2016
    • Zeno failure.

October 2016

  • October 10, 2016
    • Vented FTS to optimize InSb detector.  At 11:45 am
      InSb Ref 8450
      InGaAs A 20260
      Si Ref 22450
  • October 8, 2016
    • Swapped Caltech InSb detector back into Caltech FTS just before 6 pm LT.  Since the sun is low in the sky, I will open the FTS one more time on Mon. to optimize.
  • October 7, 2016
    • Picked up Lamont InSb from JPL after 48 hours of pumping and want to test how long it will stay cool.  Vented FTS, swapped InSb detectors without optimizing, filled InSb, restarted IFSloop.  NOTE: data tomorrow will be no good!

September 2016

  • September 30, 2016
    • Around local noon, instrument was vented to maximize detector signals:
      RT-InGaAs DC [Internal Pos.1]; Preamp Gain A; (only?) 18700 amplitude
      LN-InSb FOV=30 <B0> DC [Internal Pos.2]; Preamp Gain Ref; 7500 amplitude w/filter
      RT-Si Diode DC [Internal Pos. 4]; Preamp Gain Ref; 22000 amplitude w/red filter and iris
  • September 29, 2016
    • Put IFSloop into hold, vented the instrument, removed the Lamont InSb and installed the Caltech InSb again. Waiting for local noon to adjust the InSb, InGaAs and Si detector counts.
      RT-InGaAs DC [Internal Pos.1]; Preamp Gain A; 19600 amplitude
      LN-InSb FOV=30 <B0> DC [Internal Pos.2]; Preamp Gain Ref; ~6500? amplitude w/bandpass filter
      RT-Si Diode DC [Internal Pos. 4]; Preamp A; ~33800 amplitude w/red filter; inserted iris and closed down to ~20050 amplitude.
      NOTE: there is no aperture in the source chamber.
  • September 27, 2016
    • Noted that the M1 E/W motor turned easily with the wheel, so we asked Dick to enable this mirror again.  While Dick issued commands to move the mirror, we watched the progress in the dome.  Through iterative E/W commands and applying grease to the treaded center rod and the two metal strips that the mirror slides on, we were able to get the mirror to move +/- 30" from center.  It did get stuck once and it may again, but for now it is operational.
  • September 26, 2016
    • Resumed testing the Lamont InSb detector using OPUS (ci20160926sgeeax.003 - w/ filter). Vented the instrument, removed the Caltech InSb detector again, put in the Lamont InSb, cooled the InSb at  ~9:30 a.m (local time) and optimized the signal strength:  
      LN-InSb FOV=30 <B0> DC [Internal Pos.2]; Preamp Gain Ref; 32000 amplitude wo/filter
      LN-InSb FOV=30 <B0> DC [Internal Pos.2]; Preamp Gain Ref; 10400 amplitude w/filter
      RT-InGaAs DC [Internal Pos.1]; Preamp Gain A; 21500 amplitude
  • September 23, 2016
    • Stopped measurements and switched over to OPUS.  Obtained vented measurement, ci20160923sgeeax.001.
    • Swapped in new 5mm, 50:50 BS from Thorlabs
      RT-InGaAs DC [Internal Pos.1]; Preamp Gain A; 19250 amplitude
      LN-InSb FOV=30 <B0> DC [Internal Pos.2]; Preamp Gain Ref; 18850 amplitude wo/filter
    • Removed Caltech InSb detector to make master/slave cable for Lamont InSb; sun was too low in the sky to optimize once the cable was finished, so closed up FTS, pumped down and put into Hold.
  • September 20, 2016
    • No measurements were taken this morning and I could not ssh into almus. Went down to the lab, could not run IFSloop on almus, neither get a ping. Restartet the router, nothing happened. Restartet the Hercules computer and restarted IFSloop. FTS started to take measurements. Could't find any hints in the log file why IFSloop crashed?!
    • Put the Norhof pump back into the tank, LN2 fill works again.
  • September 19, 2016
    • PStat parameter (IFSloop/LN2+) was indicating an "unknown" status and LN2 fill did not start. I did a SaveLog and reinitializing the instrument set the PStat parameter to "sleep" again.
    • Pressure created by the Norhof pump was too low to fill the InSB detector. The Norhof software indicated a frozen exhaust. Removed the pump out of the tank to warm it up and let it dry over night.
  • September 1, 2016
    • Noticed that the dome is not closing in the middle of the day due to the coelostat mirror shadowing.  Dick says, "Strange.. it must be the sweet spot where the sun just passes over M2, but it doesn't make sense why it doesn't do that when the sun is higher."

August 2016

  • August 30, 2016
    • The InSb detector was not cooled below the threshold temperature and only InGaAs measurements were being made. After cooling the detector manually for a second time, InSb measurements started.
    • InSb measurements stoped after ca. 10 minutes. It seems that the temperature sensor line was not deep enough in the dewar. Pulled out the fill and sensor line and pushed it in again. InSb measurements are running again.
  • August 29, 2016
    • Discovered that someone had unplugged the Zeno from the wall outlet.  Data stream started up immediately upon plugging in, however, now the dew sensor says that it is wet causing the dome not to open.  RH is 50 and it was not foggy this morning, so the sensor may just be dirty.  Sensor made need to be unplugged and/or replaced. UPDATE: The wetness sensor switched to 0 and the dome opened ~11:20 am.  
    • The InSb was not cooled, however, and only InGaAs measurements were being made.  Spent a good bit of time taking the fill and sensor lines outs.  The blue sensor line bends easily and gets hung up when pushing it into the InSb dewar.  The though is that it was not emerced in the LN2.  It is working again now, but this is an ongoing problem.
  • August 28, 2016
    • Received the error message at 11 am: Zeno failure. Indicated by Zeno stale counter.  Tried to reintialize, but to no avail.  Did a Savelog and sw Status Shutdown Quickly to start overnight analysis.
  • August 26, 2016
    • The LN2 inlet hose, which inserts into the InSb, had become disconnected and the InSb did not fill properly this morning.  Replaced the fill hose, but had trouble inserting the temperature sensor (blue) line into the InSb dewar.

July 2016

  • July 22, 2016
    • Again attempted to re-immerse the pump system into the LN2 and found that there was a blockage.  I swapped in the original pump system, which has been sitting on my desk since it was returned by Norhof.  It worked immediately.
  • July 21, 2016
    • Filled the LN2 and a water bubble got trapped in the pressure line. Let the pump sit out to warm up and tried twice to re-immerse it in the LN2 with the same result.  Will leave over night to dry.

June 2016

  • June 21, 2016
    • Noted that InSb detector has been warming up in the late after (Heading to Play_InGaAs: !InSb_Chilled) and the FTS has gone into SolarInGaAs measurements.
  • June 9 &10
    • FTS in AC mode after LN2 fill; did a reinit without a change then a sw shutdown quickly.  Forgot to restart.

May 2016

  • May 18
    4 years, 178 days, 6 hours, 31 mn since Thu, 01 Sep 2011 17:51:22  
    • Reset the laser on with IFS125 webpage
    • Started mis-alignment experiments
      • ci20160518lgeeax.001   under vac, 1 mm aperture, 0.01 cm-1 res
      • ci20160518lgeeax.002   under vac, 1.5 mm aperture, 0.016 cm-1 res (LINEFIT result.)
      • ci20160518lgeeax.003   (starting point) vented, 1.5 mm aperture, 0.016 cm-1 res (LINEFIT result.)
      •  1/4 turn CW on vert. cube corner screw (or down 1/12 mm) ci20160518lgeeax.004 (forgot to tighten locking screws)
      • tightened locking screws and repeated (just to make sure) ci20160518lgeeax.005
      • 1/2 turn CCW on vert. cube corner screw, ci20160518lgeeax.006 
      • 1/4 turn CW on vert. cube corner screw (returning to starting point) and 1/4 turn CW on hor. cube corner (toward the inlet tube), ci20160518lgeeax.007
      • 1/2 turn CCW on hor. cube corner, ci20160518lgeeax.008
      • LUNCH
      • With cube corner still off in the horizontal axis (CW 1/4 turn from starting point), will go CW on the horizontal entrance mirror screw (currently at -2.170 V).  Reduce by 50 %, then compensate - ci20160518lgeeax.009
      • Went equal distance in opposite direction on same axis (first going back first to the starting point) ci20160518lgeeax.010
      • Went back to the starting point on this axis (first changing the compensator and then the hor. entrance mirror screw.  Then went CW on vertical entrance mirror screw, reducing 50% and compensating - ci20160518lgeeax.011
      • Went back to the starting point on this vertical axis (first changing the compensator and then the ver. entrance mirror screw.  Then went CCW on vertical entrance mirror screw, reducing 50% and compensating - ci20160518lgeeax.012
      • Went back to the starting point on this vertical axis (first changing the compensator and then the vertical entrance mirror screw.)  Now in the starting position for the field stops, but off on the cube corner - ci20160518lgeeax.013
      •  With the feild stops now back to the starting position, we want to pull the cube corner a bit further (or 1/4 CCW on horizontal axis) - ci20160518lgeeax.014
      • Finally, back to start - CW 1/2 turn on horizontal of cube corner - ci20160518lgeeax.015 (LINEFIT result.)
    •  With external laser alignment kit, optimized cube corner (APT = 10000, ZPD) and field stop mirrors at MOPD. Cleaned the rods and peaked small mirror.  Took vented spectrum, ci20160518lgeeax.016 (didn't look great, but ILS_vented ok).  Pumped and repeated measurement - ci20160519lgeeax.001.  (Lineshape looked better; ILS_evac ok.)

April 2016

  • April 6, 2016
    • Norton questioned whether the IFS125 temperature error's were real and suggested that we run the Herc auto-calibration routine, which he did:
      "I could do this after bedtime (or before sunrise). We'd want to shut down the instrument and then run the two programs in my home directory. The first one is DSCADAutoCalVerify, which just reports how bad the current calibration is. The second, DSCADAutoCal, performs the calibration and then reports the new results. What exactly the output means, I'm not sure. It reports an offset error and a gain error. I'm guessing that the units for both are mV, and the gain value then refers to the full scale error. The last time we ran it, the gain errors were between 150 and 600 before auto cal. Afterwards, the numbers were in the low single digits."
    • (See email from 4/7/16) - there should be a significant improvement.

March 2016

  • March 30, 2016
    • The slit is stuck again (11:40 am LT).  Resolved by switching to manual, closing it, and switching it back to auto.
  • March 29, 2016
    • Rebooted the dome computer inside the control cabinet upstairs by unplugging and pugging it back in.  It seems to have worked and the dome is operational again.  The crash was caused by a full disk.
  • March 17, 2016
    • Made changes in the IFS ANA15 editor:
      for board code "0x40C1", editted the string from "RT-InGaAs AC" to "RT-Si Diode DC" and 
      for board code "0x40C2", editted the string from "RT Si Diode AC + RT-InGaAs AC" to "RT-Si Diode AC+DC"
    • RT-InGaAs DC [Internal Pos.1]; Preamp Gain A; InGaAs voltage = 3.733, 23600 amplitude
      LN-InSb FOV=30 <B0> DC [Internal Pos.2]; Preamp Gain Ref; InSb voltage = 1.005, 7440 amplitude
      RT-Si Diode DC [Internal Pos. 4]; Preamp A, B, C, and Ref; >33000 amplitude (stopped optimizing!)  Put aperture back in to complete the optimization, but by this time the light intensity had also dropped.  Found that with Preamp Gain A, amplitude ~31700. 
    • The aperture was removed and several spectra were taken with the InGaAs/InSb arrangement.
    • Finally, FTS was closed, put under IFSloop control, pumped, and reinitiated.
    • IFSloop was not running and therefore there were no datasets saved, 160309-160316, on almus.  The logs on /home/tccon/Data_Verification/caltech_ifs9 are also empty.
  • March 16, 2016
    •  RT-InGaAs DC [Internal Pos.1]; Preamp Gain A; InGaAs voltage = 4.318, 27250 amplitude
      LN-InSb FOV=30 <B0> DC [Internal Pos.2]; Preamp Gain Ref; InSb voltage = 0.702, 5200 amplitude
    • Noted that solar beam was not well centered on inlet tube, so made fine adjustments on the hanging mirror above it. Of course, then also had to ensure that beam was centered on the input field stop and exit aperture as well, using IR alignment technique.
  • March 15, 2016
    • Had an brief period of sunny skies, after the mirror's shadow passed, but found that light intensity dropped with everything I tried (12:45 PDT).  Finally opened source region and found that there was a line right through the center of the image in the entrance mirror.  Turns out that the M2  mount arm continues to block the portion of light we send into the spectrometer well after the "image" we see on the first floor shows full sun.  It eventually passed, but the skies clouded as well.
  • March 14, 2016
    • Remounted the Si detector, its OAD, and red filter. Electrically connected Si detector and see: 
           RT-Si Diode AC [Internal Pos. 4]
      using OPUS.  Still need to change the software to reflect the change to DC.
    • The InSb OAP mount still requires some work.  The two OAP mounts were not exact mirror images apparently and one of the springs holding the optic to the vertical back plate does not line up properly.  This slight angle causes the whole optic to sit askew.  The kinematic screw system, of course, cannot work this way and it is nearly impossible to adjust the optic.  I have asked Mike to open a horizontal slit on the vertical back plate.
  • March 10, 2016
    • Partially cloudy skies continue to complicate any alignment and it will probably be next week before I can really optimize the system.
    • Modified the Si_ADC board, changing the jumper from "ACB" to "DCA" on  J3 and from "AC" to "DC" on JL.
  • March 9, 2016
    • The mounts consist of a base plate (InSb's is longer), a vertical back plate, and the kinematic optic mount.  The InSb optic mount was flipped 180 deg on the back plate to take light in from the opposite side in the new detector position.  The 3-screws of the kinematic mount did not match perfectly with the back plate, however, and the OAP adjustment was unstable.  It worked as a temporary fix to test the system and to maintain data collection while waiting for the new dichroic.  Mike Roy switched the back plates and rotated the Si optical mounts as show here:

      InSb_final_arrangement.JPG<- InSb 

  • <- Si


RT-InGaAs DC [Internal Pos.1]; Preamp Gain A; InGaAs voltage = 4.291 V, amplitude = 27060
Fighting the clouds that have moved in; will have to come back to alignment.  Instead, tried to modify the ADC board of the Si detector.

  • March 8, 2016
    • Ran Savelog, put FTS into Hold, ran SW Status Clear to Vent, and  finally ran IFS Direct VAC=2.  When vented, ran IFS Direct VAC=0 and sw Status shutdown Quickly to end IFSloop so could take solar runs with OPUS.
    • So with LN-InSb DC + RT-InGaAs DC (Internal Pos. 1) Preamp Gain B, took solar scan.  (Also see Bg Preamp gain B?) Res = 0.01.
      Amplitude 7400 (rising); Position 141145; 12:06; ci20160308saaaaa.001
      (will change later to sgeeaa) Amplitude (7580), 12:15 PST; ci20160308saaaaa.002 (will change later to sgeeaa)
    • Now with LN-InSb FOV 30  deg DC (Internal Pos 2)  Is this right?
      Preamp Gain B (NOT RIGHT); Res 0.01
      Amplitude  2221; Position 141146 - ci20160308sgeeac.001.0
      Amplitude  7410; - ci20160308sgeeac.001.1
    • Returned to LN-InSb DC + RT-InGaAs DC, Preamp Gain B
      Amplitude (7340), 12:42 PST; ci20160308sgeeax.001.0
    • Selected InGaAs detector alone and set Preamp Gain to B.  Amplitude = 23600.  Voltage = 3.72
      Then went to the dual selection and set the Preamp Gain to Ref. Amplitude  = 7130.
      Disconnected dual cable to measure voltage = 0.935; took scan ci20160308sgeeax.001.1
    • Now once again ...
      RT-InGaAs DC [Internal Pos.1]; Preamp Gain B; InGaAs voltage = 3.660, 23150 amplitude

      LN-InSb FOV=30 <B0> DC [Internal Pos.2]; Preamp Gain Ref; InSb voltage = 0.902, 6720 amplitude
      Then, LN-InSb DC + RT-InGaAs DC [Internal Pos.1] ; must change Preamp Gain to B; amplitude stays at 6750 (or that of InSb), but voltage on InGaAs still at 3.610 V.
    • Replaced 1" dichroic with new 2". 
      RT-InGaAs DC [Internal Pos.1]; Preamp Gain A; InGaAs voltage = 4.180, 26450 amplitude
      LN-InSb FOV=30 <B0> DC [Internal Pos.2]; Preamp Gain Ref; InSb voltage = 0.985, 7330 amplitude
      Then, LN-InSb DC + RT-InGaAs DC [Internal Pos.1] ; must change Preamp Gain to A; amplitude ~7350.
    • Removed the OAP mounts in front both InSb and Si detectors.  

February 2016

  • Feb 8, 2016
    • Jean-Francois noticed that the NTP server was unreachable. The time drifted by +1.33 seconds, and was reset on Feb 7 at 9 pm PST by switching to four servers from
  • Feb 5, 2016
    • Realized that yesterday's data was not copied to ccycle and that there had been no emails from the caltech herc since Feb 3.  Rebooted the router, the FTS, and eventually the Herc; J-F reinitiated the software and the system commenced scanning.

January 2016

  • Jan 18/19, 2016
    • Dick came to repair the coelostat slit, which gets stuck, and to do general maintenance.  Click here for his site visit report.
  • Jan 11, 2016
    • Dome was opened again and the FTS was reinitialized.  There was a problem cooling the detector and discovered that the black fill tube was not connected to the silicon tube.  Reconnected and commenced measurements at 9:38 am LT.
  • Jan 9, 2016
    • The skies were again cloudy, so a decision was made not to open the dome over the weekend and the FTS was put into hold.
  • Jan 8, 2016
    • Given the break in the clouds, the dome was opened at 11:20 am and sw Status reinit was run. 

December 2015

  • Dec 29, 2015
    • Coelostat or driver communication problems indicated by suntracker stale counter problem occurred again (10:50 am).  Paul came in and closed it manually
  • Dec 26, 2015
    • Coelostat or driver communication problems indicated by suntracker stale counter (10:50 am); After speaking with Dick, I put switched to manual control and opened and closed the slit.  Then, switched back to automatic and watched as the computer opened the slit.  With skies clear, we opted to let it run.

November 2015

  • Nov 24, 2015
    • InSb wasn't cold when IFSloop went into Play, causing a scanner failure. Filling InSb resolved the problem.
  • Nov 23, 2015
    • Dick significantly changed the location of the solar image, so major adjustments to the entrance flat mirror and OAP were required to fill the entrance OAP and the 4mm field stop. This will likely change again when Dick comes to town to replace or fix the M1EW slide motor.
    • Pumped down the IFS and set the Norhof timer to 22 hours (so that tomorrow, it will fill at 7:30 am). Will need to change time to 12 hours.
    • MXY1: 0.157 MNY1: 0.015 MXY2: 0.395 MNY2: 0.056
      MXY1: 0.085 MNY1: 0.085 MXY2: 0.249 MNY2: 0.248
      Earlier in the morning (closer to when this was aligned) I saw values of ~0.096 in InSb.
  • Nov. 18, 2015
    • Opened the FTS to clean/replace the CaF2 input window, carefully noting the orientation of the window. (Marked a small arrow on the outer rim of the window at the topmost location and pointing toward the spectrometer.)  The window diameter is 3.5" and not the same size as the overhead window in Park Falls, Darwin, and Lamont (115 mm). The window cleaned up perfectly with MeOH and has been reinstalled. The FTS is still vented and in hold until alignment of solar beam can be done.

October 2015

  • Oct. 23, 2015
    • Increased gain settings in IFS.tma to better match the measurement requirements:
      • -     > IFS Set Preamp Gains 0 0
        +     > IFS Set Preamp Gains 1 3
      • This doesn't entirely regain the signal we previously saw, but the skies are variable today, and the 1" 50/50 beam splitter does clip some light.
      • XY1: 0.081 MNY1: 0.079 MXY2: 0.238 MNY2: 0.232
        MXY1: 0.079 MNY1: 0.076 MXY2: 0.232 MNY2: 0.223
    • Opened yet again to attempt to peak detectors with bright sunshine:
      InSb w/filter, Preamp gain REF - 4230 amp, (goes up to 22200+ wo/filter)
      InGaAs, Preamp REF - 11300 amp,
      InSb w/filter, Preamp gain A - 967 amp, (goes up to 4900+ wo/filter)
      InGaAs, Preamp A - 13110 amp
  • Oct. 22, 2015
    • Oops - not the red filter, but the long pass filter placed in front of the InSb to avoid non-linearity was needed.  So opened the FTS and installed this filter, reoptimized light on InSb detecor (0.130 V), closed up and pumped down again.  Solar measurements commenced at 10:56 am LT.  IFScheck at 11:05 am yielded:
    • MXY1: 0.018 MNY1: 0.018 MXY2: 0.284 MNY2: 0.284
    • MXY1: 0.018 MNY1: 0.018 MXY2: 0.284 MNY2: 0.284
    • MXY1: 0.033 MNY1: 0.003 MXY2: 0.453 MNY2: 0.061
    • Measurements made on the InGaAs detector prior to 10:56 am LT have significant negative ZLOs and induced ghosts because I2S was asked to calculate LSEs using the InSb data, which spanned both halves of the alias. The data prior to 10:56 am LT should *not* be used or made public (this is currently spectra up to and including ci20151022sgeea?.032). This problem will be fixed in the next version of I2S, which checks the fraction of power in one half of the alias and if it's less than 0.1%, it does not attempt an LSE correction.
  • Oct. 21, 2015
    • Measurements started up fine this morning and LN2 fill worked with the timer.  Debra suggested that the filter needed to be in place due to saturation/linearity issues.  So opened the FTS at 3:43 pm LT to install, reoptimized light on InSb detecor (0.625 V), closed up and pumped down again.  Solar measurements commenced at 4:14 pm LT.
  • Oct. 20, 2015
    • Source solar - InSb, Preamp gain Ref, 3.445 V; InGaAs 1.688 V, measured solar spectra InSb DC + InGaAs DC (ci20151020sgeeax.002)
      Tried to optimize again and measured solar spectra RT LN-InSb DC + RT InGaAs DC (position 1)
           Preamp gain Ref, (InSb - 3.485 V; InGaAs - 1.698 V) - ci20151020sgeeax.004
    • Switched gain to
           Preamp gain B, (InGaAs - 2.254 V; InSb - 3.390 V), and repeated, ci20151020sgeeax.005.
    • Finally, removed input aperture and repeated measurements; first with
           Preamp gain B (InSb - 4.854 V; InGaAs - 2.907 V) - ci20151020sgeeax.006. 
      Interestingly, these two spectra are nearly on top of each other.  Finally, repeated with
           Preamp gain Ref (InGaAs - 2.070 V; InSb - 4.688 V) - ci20151020sgeeax.007.  Compared these results, red/green curves (InSb/InGaAs) with spectrum of InSb alone (blue curve) from 20130830.
      Preamp gain B, wo/ input aperture looks the best.
    • Evacuated the FTS with IFSloop and measurements commenced; NOTE - input aperture removed.
      og shows: InGaAs preamp resistor set to 3090
                        Si preamp resistor set to 7680
      but am not sure how to relate this to the OPUS (preamp gains of A, B, C, and Ref).
      IFScheck yields: MXY1: 0.104 MNY1: 0.103 MXY2: 0.239 MNY2: 0.238
      Inserted the Norhof tubing into the InSb detector, connected the Norhof itself to the laptop, initiated a fill while the FTS was pumping down and set timer to 12 hours.
    • WP_20151020_001.jpg
      While waiting for an answer from Thorlabs on the 2" optic, started to think about how to mount the new dichroic.  One possibility is to modify the current "good" mount provided by Bruker (left).  It fits in Bruker already and is adjustable.  The height of the "old" dichroic is 1.5".  Modifiying this mount would mean removing the front face plate (and dichroic and Teflon screws, of course), drilling out a round 2" diameter hole extending downward from the top of the opening, and tapping two new holes for Teflon screws, wider apart.  It might be easier just to have a new front face plate!
  • Oct. 16, 2015
    • ci20151016lgeeac.001 - InSb, lamp, Preamp gain Ref, Bruker amp = 5650! (0.864 V)
      ci20151016lgeeac.002 - InSb, lamp, Preamp gain A, Bruker amp = 1255
      ci20151016lgeeac.003 - InSb, lamp, Preamp gain B, Bruker amp = 1675
    • Found that light into InGaAs was now being clipped, so had to move the dichroic and start again.  Using Preamp Ref (favoring InSb), got opt. amp = 3650 on InGaAs and 8910 on InSb.  Using Si DC + InGaAs DC detector setting in Bruker, with InSb in the Si detector position, Preamp Gain = Ref, tried to take spectra. 
  • Oct. 14, 2015
    • Opened the FTS to reinstall the InSb detector and base plate in the current Si detector position.  Si detector removed!  The OAP mirror orientation for the InSb detector was rotated 180 deg and a slave/master cable was made.  Needed to move the pegs (in the base of the detector compartment) to their back position to install the InSb detector.  (May need to find another 1", M5 allen head screw to place Si base plate in current InSb position.)  Swapped in the beamsplitter for the current dichroic and removed the red filter. 
    • Operating with the NIR (little to no sunlight), opt. voltage on InSb = 0.020 (Bruker amp = 118; InSb DC, Preamp gain B, pos. 2, 0.01 cm-1 res, 1 mm aper.)  Opt. voltage on InGaAs = 0.807 (Bruker amp = 5409; InGaAs DC, Preamp gain B, pos. 1, 0.01 cm-1 res, 1 mm aper.)  Took spectra:

      ci20151014lgeeaa.001 - InGaAs, lamp, Preamp gain B, Bruker amp = 5409
      ci20151014lgeeaa.002 - InGaAs, lamp, Preamp gain A, Bruker amp = 4580
      ci20151014lgeeaa.003 - InGaAs, lamp, Preamp gain Ref, Bruker amp = 3968

      InSb, lamp, Preamp gain B, Bruker amp = 118
      InSb, lamp, Preamp gain A, Bruker amp = 91
      ci20151014lgeeac.004 - InSb, lamp, Preamp gain Ref, Bruker amp = 371
      (NOTE: Later realized that these InSb values were not optimized; beam clipping OAP.)
    • Unfortunately, the dual detector cable is much to short at ~ 8.5"; it probably needs to be more like 12-13".  I attempted to take a solar scan, but the light is waning.
  • Oct. 7, 2015
    • Opened FTS again to remove InSb base plate and OPD. The orientation of the InSb's OAP is rotated 180 deg on it's mount (so that it is the mirror image) and I think it will work.  That said, the mount is NOT exactly symmetric.  There are only three metal pads for the kinematic screws to turn against during adjustment and, of course, the missing one is one we need.  Also, the mirror is held to the back of the mounting plate by one screw and the two spring assemblies.  The screw no longer aligns with the hole in the back of the mounting plate and cannot be used with the current orientation.  Again, this is caused by the fact that the pieces are not exactly symmetric.  I am able to adjust the mirror on the mount, but I do not know how  critical the distance is between the OAP and the detector.  I guess we just need to try it.
    • Also checked on the slave/master cable connection between the InGaAs and Si detectors.
  • Oct. 6, 2015 -
    • Open FTS to again try and optimize light into Si detector.  Amplitude (at start) = 135, Position 141401, 1 mm aperture, Si DC, PreampGain=Ref.  Amplitude (end) = 5650! (For future reference, InGasAs amplitude =15600.)  By noon, the amplitude on both detectors had increased further.
    • Took OPUS spectrum with Preamp Gain = Ref (ci20151006saaaaa.001) followed by Preamp Gain = A (ci20151006saaaaa.002) for comparison.
    • Removed InSb detector for Debra to take to J-F.  IFS.tma modified to run InGaAs_Si only.
  • Oct. 1, 2015 - Opened FTS to explore usage of (BSW511) 50:50 CaF2 beamsplitter between InGaAs and InSb detectors. (InGaAs V= 1.389 V w/current TCCON settings)  First measured amplitude and position with OPUS on the various detectors:
        InGaAs - amplitude- 10898; position- 141395
        Si - amplitude - 60; position - 141390
    Removed filter before InSb
        InSb (position 4, wo/filter) amplitude - 7216; position 141389
    Removed Si and moved InSb over to previous Si position
        InSb (position 2, w original dichroic and red filter) amplitude -215; position - 141385
        InGaAs - amplitude - 10774; position - 141385
    Removed original dichroic and red filter, installed new beamsplitter, and optimized light into InSb
        InSb - amplitude - 2062;  position 141385
        InGaAs - amplitude - 4442; position 141385
    Obtained lamp spectra w/OPUS -  0.02 cm-1, 4 scans - for InGaAs, then InSb; then recorded solar spectra for both.


September 2015

  • September 2, 2015 - Work on lab above FTS, and work on telescope image. Much of the data today will be garbage.
    • The centering of the solar disk has changed. Will likely need to modify image on pinhole.

Back to top

July 2015

  • July 27, 2015 - Foam placed back on IFS.
  • July 24, 2015 - Realigning IFS. No data. See the Alignment History page for details. Not sure whether the alignment was successful. May be poor quality data July 24 - 27. Foam covers were left off throughout the weekend. NOTE: the ILS should be assessed with a 10kHz scanner speed. We know that the 7.5 kHz speed is noisy and will make the final quality assessment difficult.
  • July 17, 2015 - Drilling in the lab above the IFS.
  • July 10, 2015 - Realigned the IFS and re-minimized ghosts today. Mostly cloudy day. See the Alignment History page for details.
  • July 4, 2015 - Coelostat failed to open after noontime dance. And on an OCO-2 target day...

Back to top

June 2015 

  • June 23, 2015 - Laser modulation amplitude errors... the laser may be failing.
    • Nope, the scanner cable was broken. Replaced at ~2pm with Camille's help.


  • June 15, 2015 - Dick Treffers was in town for coelostat maintenance:
    • "I washed the primary (poorly), cleaned the rails, replaced some of the drip pads, lubricated the ways, checked the limit switches and most importantly made measurements so I can design the increased torque drive for M1NS.  I tried to get the slide to work better, but failed (by the time I got to work on that, it was so HOT in the dome). Nora had gotten someone to wash the floors a while back and they looked pretty good.  I also removed some bizarre debris from the projector debris." - DT

Back to top

May 2015

  • May 19-20, 2015 - The M1 N/S mirror was not high enough on the slide, so the image was being cut off. Not clear whether the IFS aperture was getting appropriately filled.
  • May 12, 2015 - The Paroscientific, Inc. Model 765 Barometric Standard (Digiquartz) arrived yesterday, and I recalibrated the Setra pressure sensor:
    • Zeno: 983.4 hPa; Digiquartz: 983.671 hPa; pout_cor = +0.271 hPa
    • Zeno: 983.5 hPa; Digiquartz: 983.716 hPa; pout_cor = +0.216 hPa
    • Should check again later in the day, if it clears up, but it looks like ~0.2 hPa is about the right answer. Camille agrees that adding 0.2 hPa to the Zeno value is the right answer.
    • Will reprocess all data recorded to date using pout_cor = 0.2 hPa. I will also average FWD/REV scans to mitigate the scanner noise problems described below until April 20, 2015.

Back to top

April 2015

  • April 20, 2015 - Replaced motor/encoder assembly and scanner cable. The noise bursts on the interferograms went away, and the MVD went from ~16 to ~1.7.


  • Unfortunately, the above plot was for two 5 kHz scans. At 7.5 kHz, the noise was there (not bursty, but large noise), and not at any other speed I tried (5, 10, 15, 20, 30, 40).


  • Put container back in play, but changed velocity to 10 kHz until we understand this problem better. I did not commit this change via CVS, but I saved a copy of the previous file (named
  • April 14, 2015 - Took audio of the scanner while scanning. The noise in the audio matches up nicely with the interferogram (plotted below as a function of approximate time).
  • interferogram_7.5kHz_vs_seconds.jpg
    • Here's the audio file: 
    • Seems clear that we need to swap out the scanner/encoder assembly.
  • April 13, 2015 - Pumped back down and put into play before dome opened in the morning
  • April 11, 2015 - Recorded a bunch of (non-coadded) noise-only spectra. Strangely, we don't see noise bursts in the noise-only interferograms, but maybe we need to coadd a few interferograms to see it?! Noise bursts still visible in single lamp interferograms.
  • April 10, 2015 - Upgraded firmware to try (unsuccessfully) to get rid of the noise bursts in our interferograms. Down for most of the day.
    • Changed HCl cell for old Lamont #5 cell. Original cell to be shipped to Frank for analysis. (Returned to Caltech Dec., 2015
    • Cleaned rails (quite dirty with Delrin residue).
    • Running 100 coadded scans of darks to look for noise bursts.
      • Coadded interferograms see no noise bursts - maybe because we're averaging them out?
    • No data today; probably none over the weekend.

 Back to top

March 2015

  • March 30, 2015 - Ran some lamp tests in the morning, and forgot to turn off the lamp after putting the system back in play. Turned off lamp at 9:19 am PDT.
    • Dick Treffers is in town for coelostat maintenance. He will be cleaning the mirrors this morning. He also sprayed some lubricant inside the dome.
    • Mirror cleaning finished at 12 PDT.
  • March 15, 2015 - In response to an email sent to Gregor and Axel regarding periodic noise on the 125HR interferograms, Gregor asked me to check the CAN bus 5V supply (5.172V measured), and the ramp input. This was my response. The instrument was shut down for ~an hour to take these measurements between 10 and 11.
  • March 9, 2015 - Modifed several settings on the new Norhof pumping system, which will hopefully take care of the "LN2 Pump did not go active in InSb_Fill" error message.  The feed height or height difference between the pump exhaust and the InSb input line was set to "52 cm" and the repeat time was set to "off".  These values were confirmed by connecting the old pumping system.  While at it, I inserted the old pumping system back into the LN2 and re-established that the LN2 level sensor is indeed broken.  Upon drying out completely, the new pumping system was re-inserted into the 50 L Dewar.
  • March 4, 2015 - Added some LN2 from the newly delivered large low pressure tank in lab.  The large tank had nearly no pressure and the fill was taking forever, so only about 10 L were added.
  • March 3, 2015 - Swapped the entire pump assembly from the new "spare" Norhof system for the old pump assembly.  The LN2 level immediately jumped up to 9.4 cm, which seemed about right.

February 2015

  • February 27, 2015 - Removed LN2 pump from LN2. Dried out the system and blew dry air through the tubes. Reassembled and the system seems to be working (for now).
  • February 20, 2015 - LN2 pump stopped working. Presumably ice in the system.

Back to top

January 2015

  • January 27, 2015 - Dome closed when it ought to be open (~10 am LT).  Logs show "too many guide errors", so contacted Dick.  Dick found that the zero moved again and did "the business" again.
  • January 21, 2015 - Dick and I just tried to move M1 N/S.  The controller in the old-fashioned control box displayed "---" and Dick said it was overloaded.  I watched the rotation of the motor while Dick commanded it to turn.  It seemed to be ceased up.  I tried to turn the motor manually and then at the same time Dick was commanding it.  We got mirror to move up ~1.5", but then the motor stopped moving.  It should be ok for a couple of weeks, but Dick is going to schedule a trip down.  (Dick, please correct me if necessary.)
  • January 21, 2015 - Dick (remotely) and Coleen tried to move M1 N/S.  The controller in the old-fashioned control box displayed "---" which Dick said meant it was overloaded.  Dick commanded M1 to turn, but Coleen observed that it did not rotation freely and seemed to be ceased up.  Coleen tried to turn the motor manually, while Dick commanded the roatation.  The mirror moved up ~1.5", but then the motor stopped moving and moved no further when commanded.  It should be ok for a couple of weeks, but Dick is going to schedule a trip down.
  • January 19, 2015 - Dick noticed that the Mirror 1 North/South axis is stalled and needs to be investigated.
  • January 15, 2015 - Instrument reset in software. Measurements after reset are in AC mode. Very clear day, however.

Back to top

December 2014

  • December 29, 2014 - Not really sure what I did, but the LN2 pump pumped LN2.
    • I blew dry N2 into the pressure sensor tube at the top of the pump under the cap, and noted that the the tubing wasn't firmly on the pressure sensor, which I fixed after drying it with N2.
    • Put the pump back in LN2, but the depth and pressures were not reading correctly. Took the pump out of LN2 but left it plugged in.
    • After it warmed up, took the clear, hard plastic tubing off of the pump sensor cables and turned it around. It kept falling off, so I thought that it was perhaps installed backward.
    • Put the pump back in LN2. The depth reading was still wrong, but the pressures looked correct, and the system filled the InSb. We'll see what happens tomorrow.
  • December 13, 2014 - December 29, 2014: LN2 pump unable to generate pressure. This may be related to the rain, but many attempts to dry the pump have been unsuccessful. No InSb measurements.
  • December 5, 2014 - December 8, 2014 - After heavy rains, the rain gauge appears to be broken. Dick has tried to fix the problem by ignoring the rain gauge and using the leaf wetness sensor instead, but this is slow going and some data have been lost.

Back to top

November 2014

  • November 18, 2014 - Added Dick Treffers to container emails so he is aware when the dome is opening and closing too frequently. This should be temporary, so I didn't do it through cvs.
    • $ cd /home/citco2/src/config/caltech_ifs9
      $ cp
      $ edit
      $ diff
      <           Richard Treffers <>,
      $ make distribution 

Back to top

October 2014

  • October 28, 2014 - Dick Treffers is in town. Coelostat will be taken down for maintenance.
    • Mirrors cleaned using rubbing alcohol.
  • October 4, 2014 - Discovered that the source of the problems with the LN2 filling system was a water droplet trapped in the tube housing the pressure measuring device.  This droplet freezes upon putting the pump in the LN2, of course, and prevents the device from measuring.  Most of the "pressure measuring" tubing is inside a larger diameter black tubing and hence the water was not visible.  To get at it, one must remove the pastic tube shielding the printed circuit board (PCB) and fill tube.  Locate the pressure measuring device on the PCB and carefully remove it from the silicon tubing.  Blow dry N2 into the silicon tubing (need 1/8" tubing to insert).  Will see water move up the tube toward the heating unit.

Back to top

September 2014

  • September 15, 2014 - IFSloop in hold while dome opening issue is dealt with. Taken out of hold at 11:25 am PDT.
  • September 14, 2014 - Dome was failing to open - contacted Dick for suggestions. No good data recorded.
  • September 4, 2014
    • Filled 50 L LN2 dewar, but was unable to get the pump to work normally.  Norhof software reports that there is a blockage in the exhaust, left the pumping unit out of the LN2 to thaw and dry out. (9/5/14 - works fine again.)
  • September 3, 2014
    • Scanner failure with error messages in the log:
      17:30:53 [ERROR] 17:30:51:207:F:CANDRV.CPP:CanFatalError:Device communication error: CANBUS transmit not possible

      17:30:54 [DEBUG] TMA: Entering State Scanner_Failure
      17:30:54 Srvr: TMA: Email Error Message Scanner failure.
      17:30:54 [DEBUG] TMA: Entering State Failure
      Reinitiated software
  • September 1, 2014
    • Scanner failure. Reinited software.
    • InSb signal levels are very low (~0.06, instead of ~0.2 at ZPD).

Back to top

August 2014

  • August 20, 2014

    • Connected the Norhof to the laptop and used the Norhof software to monitor the values.  Initial values appeared reasonable: LN2 level = 22 cm, InSbT = -5 C, and LN2P = 31 mbar. Using the controls on the pumping system itself (the green button), filled the InSb dewar.  The LN2P went up to ~ 90 mbar as expected, the InSbT dropped, and after ~1.5 min at -191 C the pump shut off.  The InSbT immediately started climbing and was at -127 C within 2 min. Next, topped off the InSb dewar manually to see how long it would hold - ~ 23 min. There clearly was LN2 in the dewar, as was evidenced by 1) frost on the black filling tube and T sensor when the puck was removed, 2) the fact that there was no boil off or steam exiting the dewar, and 3) inserting the T sensor alone into the dewar and observing InSbT= -191 C.  So, decided to feed the T sensor further into the InSb, so it is now ~ 5.5" in length into the dewar.  Reconnected Norhof system to the Herc and ran "Status InSb fill".  Temp dropped rapidly, but also came up quickly once pumping stopped, so filled InSb dewar by hand.  T of -193 C was maintained for just over 2 hours, then T shot up to -138 C, indicating that the sensor is no longer in the LN2 (which means < 3" LN2 in dewar.)

  • August 19, 2014

    • With more time this morning, I repeated the process:
      16:09 UTC (9:09 LT): InSbT = -110 C; LN2T = -194 C; LN2P = 18.5 mbar; Depth = 23.8 cm
      Filled manually
      16:12 UTC (9:12 LT): InSbT = -193 C; LN2T = -193 C; LN2P = 18.5 mbar; Depth = 23.8 cm
      16:34 UTC (9:34 LT): InSbT = -193 C; LN2T = -194 C; LN2P = 17.9 mbar; Depth = 23.2 cm
      16:36 UTC (9:09 LT): InSbT = -123 C (and rising); LN2T = -194 C; LN2P = 17.9 mbar; Depth = 23.8 cm. 
      Looking in the IFSloop log, this coincided timewise with the "Enter State InSb_Starting".  The LN2 pump had not started.  I did not try to fill the InSb dewar using the Status InSb Fill command today.

  • August 18, 2014

    • Pressure inside InSb detector = 1.29e-5 mbar; removed from pumping system and replaced in FTS system.  FTS pumped down with no problems.  Replaced and plugged in the LN2 filling system, but filled the InSb dewar manually.  The pressure looked good (at least for the 10-15 minutes I looked at it) but the LN2P (=-4 mbar)  and Depth (1.0 cm) made no sense.  Confused and not trusting it over night, I removed the pump from the 50 L and unplugged.

  • August 15, 2014

    • Vented system, removed InSb detector, replaced cap on detector cover, replaced cover, and restarted IFSloop

    • Started pumping out InSb detector using procedure outlined by Bruker, pumping station set up by Jason, and vacuum line connector (CPC-CRYOLAB VO8-082T) borrowed from Jean-Francois.  Will pump over the weekend.

  • August 7, 2014

    • After letting the LN2 the InSb detector evaporate, used a N2 line inserted into the detector to dry out the detector - same results.  Also took a few measurements: internal depth of InSb detector = 8.5", length of black fill tube inside dewar = 6", and length of T sensor inside dewar = 4.5".

  • August 5, 2014
    • Removed the pump from the 50 L LN2 dewar and the puck assembly from the InSb dewar for several hours to see if there is some ice buildup somewhere causing the problem.
  • August 4, 2014
    • The 50 L LN2 level had dropped just below 15 cm, which shouldn't be a problem, but first filled this and then attempted to fill the InSb dewar.  Shortly after fill, IFSloop went into, "Entering State InSb_Not_Chilled".
    • Pulled the black puck up slightly and confirmed that the black filling tube was still in place and again tried filling the InSb dewar with the same result.  
    • Removed the puck completely to check the pointing direction of the outlet hole in the black fill tube.  Made sure that it was away from the temp sensor, carefully reinstered the two tubes and puck, and again restarted the InSb fill.  The temp dropped (~-191 C) was maintained as long as the pump was on, but started to go up (to ~ -135 C)when the pump shut off and IFSloop went into "Entering State InSb_Not_Chilled".  (Repeated this several times.)
  • August 3, 2014
    • Received the error message - LN2 Fill not finished after 15 minutes.  Suspect that the black tube that attaches to the black puck and feeds LN2 down inside the InSb detector has fallen out.  Will check tomorrow.
  • August 2, 2014
    • No data recorded - something happened during wakeup that prevented IFSloop from running. Was a cloudy day, so probably not much data loss, but there is no 140802.1 saveset.
    • From Norton:
      • I don't see anything obviously wrong.

        From citco2.log, we get this sequence early Saturday:
        Memo Starting: Sat Aug  2 02:49:48 2014
        02:49:48 TMbfr: Startup
        Running script interact
        02:49:48 Herc: /dev/huarp/citco2/Hercules Initialized
        02:49:48 TMbfr: Shutdown
        02:49:48 Col: Startup
        02:49:48 [FATAL] Col: Unable to open TM/DG: 2
        06:51:17 Herc: /dev/huarp/citco2/Hercules Shutting down
        Memo Terminating
        IFScol (aka "Col") died because it could not access /dev/huarp/TM/DG, which is created by TMbfr, but TMbfr apparently shut down just before IFScol started. I have no idea why it would do that.

        I'm going to try restarting IFSloop.
    • Restarting IFSloop worked, and the system is in play for August 3, 2014.

Back to top

July 2014

  • July 18, 2014
    • 10:20 am PDT: Maintenance work on fixing rails up in dome.


June 2014

  • June 23, 2014
    • 10:10 am PDT: On Geoff's advice, increasing the resolution of the InSb measurements to 0.01 cm-1. Also want to match the InSbCell measurements to the same resolution, so we can monitor the ILS at 0.01 cm-1 resolution. Only 2 InSb scans were recorded at 0.02 cm-1 resolution before the switch was made.
      • $ pwd
        $ diff -c 
        ***	Mon Jun 23 10:01:53 2014
        ---	Thu Sep 20 16:04:51 2012
        *** 141,147 ****
            SolarInSb => {
              SFM => 'SolarInSb',
              DTC => "16577",
        !     RES => "0.010000",
              NSS => "2",
              APT => "1000",
              AP2 => "1000",
        --- 141,147 ----
            SolarInSb => {
              SFM => 'SolarInSb',
              DTC => "16577",
        !     RES => "0.020000",
              NSS => "2",
              APT => "1000",
              AP2 => "1000",
        *** 186,192 ****
            InSbCell => {
              SFM => 'InSbCell',
        !     RES => "0.010000",
              DTC => "16577",
              AP2 => "1000",
              APT => "1000",
        --- 186,192 ----
            InSbCell => {
              SFM => 'InSbCell',
        !     RES => "0.012000",
              DTC => "16577",
              AP2 => "1000",
              APT => "1000",
        $ cd ..
        $ make clean
        $ make
        Shut down IFSloop
        $ make distribution
        $ IFSloop start
  • June 13-18, 2014
    • Dick Treffers in town cleaning, oiling, general maintenance on the coelostat. Swapped out the USB frame grabber for an ethernet one. May expect some S-G shift changes during this period.
  • June 13, 2014
    • Added Paul Wennberg to caltech_ifs9 email recipient list.
  • June 12, 2014
    • Scanner failure. Reinited software, but software still sent another reset. The IGMS being recorded after that were in AC mode. Shut down IFSloop, rebooted the scanner and started IFSloop again. Back in DC mode.
    • One of the error messages was about the LASB modulation too small... need to keep an eye on the laser.
  • June 8, 2014
    • Scanner failure. Reinited software and restarted IFSloop.
  • June 2, 2014
    • Installed USB hub to see if we could circumvent the ccd expose errors (suspected to be caused by a bad USB port on the computer). This seems to have worked, at least temporarily.
    • Dick has also purchased an ethernet version of the sensoray to replace the USB-driven one, and will configure and send down to Caltech for installation for a more permanent solution.
    • The LN2 tank sounded a general alarm this morning while trying to fill. Not sure why. I have unplugged it and taken the pump out of the tank to warm it up. It has ~15-20 cm of LN2, which I would have thought would be enough, but perhaps not.
    • Coleen filled the dewar and everything appears okay for now.

Back to top

May 2014

  • 2014-05-27
    • Tried rebooting the sensoray (by unplugging and plugging back in). The zeros are off again, in both declination and ha.
  • 2014-05-26
    • Coleen put the system into hold because the coelostat was still throwing ccd expose errors.
  • 2014-05-23
    • There are ccd expose errors causing the error count to go up on the coelostat and the dome to close. I don't understand what's going on. Seems like the guider has stopped guiding.
    • No solar image was appearing on the first floor. Tried Dick's suggestion from last time. Fingers crossed:
      • /home/coelostat> tx offset ha=0.75
        done offset
        /home/coelostat> tx offset ha=0.75
        done offset
        /home/coelostat> tx offset ha=-0.75
        done offset
        /home/coelostat> tx offset ha=-0.75
        done offset
        /home/coelostat> tx offset ha=-0.75
        done offset
        /home/coelostat> tx offset ha=-0.75
        done offset
        /home/coelostat> tx offset ha=-0.75
        done offset
        /home/coelostat> tx offset ha=-0.2
        done offset
        /home/coelostat> tx offset ha=0.1
        done offset
        /home/coelostat> tx zero last
        done zero crot=0.666 ctilt=0.020
        /home/coelostat> tx offset ha=0.01
        done offset
        /home/coelostat> tx zero last
        done zero crot=0.686 ctilt=0.020
  • 2014-05-22
    • The coelostat computer was rebooted while some security holes were being plugged. Cloudy day, coelostat didn't open.
  • 2014-05-16
    • Dick asked me to rezero the coelostat tracking (send commands while seeing whether the image appears centred on the window):
      • /home/dwunch> tx offset ha=0.75
        done offset
        /home/dwunch> tx offset ha=-0.75
        done offset
        /home/dwunch> tx offset ha=-0.80
        done offset
        /home/dwunch> tx offset ha=-0.90
        done offset
        /home/dwunch> tx offset ha=-0.85
        done offset
        /home/dwunch> tx offset ha=0.5
        done offset
        /home/dwunch> tx offset ha=0.5
        done offset
        /home/dwunch> tx offset ha=0.5
        done offset
        /home/dwunch> tx offset ha=-0.2
        done offset
        /home/dwunch> tx zero last
        done zero crot=0.633 ctilt=-0.071
      • Seems to have temporarily worked, so I've put the IFS back in play. For now.
  • 2014-05-15
    • IFS in hold due to lack of coelostat tracking.
  • 2014-05-14
    • Coelostat doesn't seem to be able to track the sun. Not sure why.

Back to top

April 2014

  • 2014-04-16
    • LN2 failed to stay cold after the morning fill - the black tube had disconnected from the puck. Plugged the tube back in and sent sw status InSb Fill. This seemed to work.
    • The IFS pressure was also already up at 4 hPa by 11 am. Don't know why. Sent sw status pumpdown.

Back to top

March 2014

  • 2014-03-18
    • Dick Treffers in town for coelostat maintenance. Replaced the faulting motor controller. Left before tuning it, so it may be suboptimal until 2014-03-20.
  • 2014-03-13
    • Coelostat running again.
    • After a software reset of the IFS, the Bruker was taking data in AC mode. This lasted for over an hour before I went downstairs and reset the IFS with the reset button. Back in DC afterward.
  • 2014-03-12
    • Coelostat failure again. Replaced and programmed the EPM chip, but Dick does not have easy access to the internet to restart the coelostat.

Back to top

Februrary 2014

  • 2014-02-28
    • Starting raining heavily last night and the overnight transfer seems to have failed due to a network error.
    • Tried rebooting the router, but this only fixed the problem temporarily (~15 minutes). Rebooted the router several more times, then tried rebooting the herc. We'll see if that helps. It did not.
    • Swapped out the router for a new netgear...
  • 2014-02-24
    • EPM chip blew again in the coelostat dome. Replaced chip and reprogrammed.
  • 2014-02-17
  • 2014-02-03
    • Coelostat dome rotation died again. Replaced the EPM chip and reprogrammed (for the 5th time).

Back to top

January 2014

  • 2014-01-17
    • Swapped the coelostat netgear router for a new one to see if that stops the period network connectivity problems.
  • 2014-01-16
    • Glendora fire broke out at 5:50 am.  (Glendora is ~17 miles due east of Caltech.)
  • 2014-01-14
    • Coelostat's netgear router was acting up. Rebooted the router (by unplugging). This caused an IFSloop failure, because the zeno data wasn't being transmitted. A reinit solved that problem, but we lost ~90 minutes of data.
  • 2014-01-10
    • Fixed dome rotation problem: 220V was not reaching the rotation controller.
  • 2014-01-08
    • Dome rotation failure.
  • 2014-01-07
    • Vibrations in the tank below the IFS due to some maintenance work on the pumps. The work will continue over the next couple of weeks.
  • 2014-01-02
    • Filled up the LN2 tank from 16.5 cm to 38 cm in prep for some work being done in the water pit below the IFS.
    • Ran sw status InSb Fill afterward, since the software had defaulted to Solar scans only (i.e., not alternating between Solar and SolarInSb).

Back to top

December 2013

  • 2013-12-17
    • Norton ran the herc's autocalibration program to see if that would fix the erroneously high temperature readings. It worked!
    • Sensor Before Cal (9:08 am) After Cal (~9:30 am) Comments
      MtrCT 29.0 26.2  
      ScnCT 29.4 26.3  
      IntCT 28.1 24.7  
      SrcCT 28.0 24.2  
      InpTT 24.0 21.0  
      LaserT 30.6 26.7  
      ElecT 30.1 26.8  
      PCDiskT 29.2 25.3  
      PC_CtrT 37.5 33.1  
      ZenoT 24.5 21.3  
      AirT 23.5 21.1  
      ScBlkT 27 27 not expected to change
      SrcT 24 23 not expected to change
      IFS_P 0.70 hPa 0.62 hPa not expected to change
    • The procedure was:
      • Shut down IFSloop
      • cd /home/nort
      • sudo ./DSCADAutoCalVerify
      • sudo ./DSCADAutoCal
      • Restart IFSloop

 Back to top

November 2013

  • 2013-11-19
    • Dick has shut down the coelostat to investigate the oscillation problem.
  • 2013-11-17
    • Coelostat is down because the sensoray (tv guider) is not connecting. Rebooting twice solved the problem, but not until the morning of 2013-11-18.
  • 2013-11-01
    • Replaced the silicone tubing on the LN2 autofill system as the previous one had developed a crack.

Back to top

October 2013

  • 2013-10-29
    • Some really bad oscillation of the solar image. Not sure what's going on. The S-G shifts will be off.
  • 2013-10-28
    • Cleaned residual water out of the LN2 fill tube (the hard plastic part that sits inside the 50L tank).
    • Put pump back into LN2 - now pumps with sufficient pressure.
  • 2013-10-26
    • LN2 fill system isn't generating enough pressure to pump LN2. Took pump out of LN2 to warm up.
  • 2013-10-18
    • Pump failure in the morning. Issued a sw status pump down, and that seemed to work. Maybe IFS direct VAC=0 didn't fully close the valve.
    • I'm seeing pretty high MXY values on InGaAs - 0.8+ in the early mornings. This seems to be transient, like we're not really centred on the sun properly, and if we were, we might be saturating... not sure what to do about this.
    • InSb isn't filling correctly. There are several cracks in the input fill puck system - I will have to take it apart to see if I can fix it.
      • Took apart the puck - the black tube that delivers LN2 into the InSb dewar had fallen off. I don't think the crack in the puck goes all the way through (though it's difficult to tell). There is a crack in the small tube that mates the input tube to the puck - I will try to get the specs of the tube and purchase replacements.
      • Took off the velcro from the puck, cleaned the glue off and put new velcro on in places that don't seem to get as cold (on the side and near the edge).
  • 2013-10-14
    • 50 L LN2 was too low to fill (10/13 and today).  Once the large lab replacement dewar arrived, the 50 L and subsequently the InSb detector dewar were filled.  Software was reinitialized at 17:41 (10:41 LT) and InSb scans initiated at 17:47.
  • 2013-10-03
    • 11:30 am - Replaced jumper that forces HeNe shutter to stay open with a proper jumper soldered onto a mating LEMO connector.
  • 2013-10-02
    • Taking the opportunity while the skies are cloudy to transfer over data.
  • 2013-10-01
    • Dick Treffers is in town to clean M1 and perform maintenance of the coelostat. Expect periods of no/unstable sunlight.

Back to Top

September 2013

  • 2013-09-24
    • Repeated scanner failure due to the HeNe shutter jumper failure; new piece is being ordered.
    • Smoke from the Azuza fire easily discernible at Caltech this morning.
    • Began automated nightly transfer of data from the herc to ccycle.
  • 2013-09-23
    • Scanner failure. Discovered that the jumper for the HeNe shutter had again failed.
  • 2013-09-22
    • Scanner failure. Seems as though the jumper that forced the HeNe shutter to stay open failed. Must purchase a mate for the cable to make this jumper more permanent.
  • 2013-09-15
  • 2013-09-09
    • After a IFS125 Software Reset, the interferograms were recorded in AC mode. Rebooting the IFS while IFSloop was in hold did not change to DC. Exiting out of IFSloop and then resetting made the first InGaAs/Si pair in AC, the subsequent InSb in DC, and the subsequent InGaAs/Si in DC.
      • However, the first slice of the InGaAs/Si pair (after both reboots) was in DC, which will make for very confusing interferograms.

Back to Top

August 2013

  • 2013-08-30
    • Changed out the input tube for one without a quadrant sensor (to ship the original one to Lamont)
    • Adjusted the 45-degree mirror to best match the previous position of the solar image. Did fine tuning with the OAP inside the source compartment. Pins providing resistance on springs on three adjustement screws on the 45-degree mirror were loose and kept falling out.
    • Changed the 35mm aperture stop for a 45mm aperture stop
      • Increased InGaAs signal from ~10000 counts (Gain A) to ~17000 counts
      • Increased InSb signal from ~6000 counts (according to the wiki) to ~9000 counts
    •  After pumpdown:
      • MXY1: 0.209 MNY1: 0.047 MXY2: 0.560 MNY2: 0.088
      • MXY1: 0.274 MNY1: 0.023
  • 2013-08-29
    • Rebooted almus since transfer to ccycle stalled out three times in a row.
      • Exited out of IFSloop
      • sudo shutdown
  • 2013-08-23
    • Drilling in the floor beneath the IFS this morning at 7:40 am local.
  • 2013-08-22
    • Drilling in the floor beneath the IFS today - should be done by 12:30pm.
  • 2013-08-19
    • Work began in the sub-basement pump area. Some sporadic vibration is expected.
  • 2013-08-16
  • 2013-08-13
    • Used LINEFIT14 to calculate the ILS from 2012-10 through 2013-07. You can see that changing the laser made a difference in the ILS. Figure attached.
    • J-F noticed that the set points for the IFS are set too high. This needs a new resistor inside the herc to fix (and explains the periodic high T errors), but not urgent.
  • 2013-08-07
    • Both InSb and InGaAs scans seem to be in DC mode...
  • 2013-08-06
    • Scanner error just before 11 AM (LT). Simple re-init of IFSloop re-started measurements, but with ADCs stuck in AC mode. Went to Hold. Waited until end-of-scan. Then issued an IFS125 soft reset. After IFSloop re-init, system took data in DC mode, starting at 18:34:02 UT. But subsequent Solar InSb seems to be in AC again...

Back to top

July 2013

  • 2013-07-20&21
    • Overnight IFS compartment temperature errors. Again, on nights when the lamp runs are being taken. The temperature is only reaching 30C, but must check to ensure that the lamp is being turned off when it should.
  • 2013-07-02
    • Guider failed due to lack of communication with Zeno and guider camera. Rebooted sol and plugged in Zeno cable at ~9:30 am PDT
  • 2013-07-01
    • Dick Treffers is in town to install a new M1 rotation motor. Coelostat is unavailable, so no measurements.
    • New rotation motor is much smoother than the previous one. Hopefully less jerky solar positions from now on.

Back to top

June 2013

  • 2013-06-24
    • IFS125 temperature error occurred because the DI water in the lamp chiller was low. Filled it back up.
  • 2013-06-20
    • Dome failure due to declination angle being off by ~0.5 degrees. Dick fixed it after a couple of hours, but look for S-G shifts in the data that are reduced after 2:30 pm PDT.
    • Dome slit blocked part of the beam in the later afternoon - should show high S-G shifts.

Back to top

May 2013

  • 2013-05-22
    • Put Spectraphysics laser back into the Caltech instrument
    • Lineshape still looks terrible, but ME from spectraphysics laser looks okay
    • Back in Play by 1:30 pm. Only an hour or two of data lost, since it was cloudy this morning.
Location Laser Zero Max Min ME
Front A -8.95V 4.075V -8.225V 0.89
Front B -9.275V 0.075V -8.925V 0.93
Back A   3.975V -8.075V 0.87
Back B   0.0V -8.75V 0.89
  • 2013-05-13
    • Dick walked me through resetting the defaults on the new chips for the dome rotation controller, so they could be programmed properly. Back up and running all day.
  • 2013-05-10
    • Dome rotation failure. Looks like another chip in the rotation controller is corrupt. Unfortunately, the replacement chips don't seem to be the same as the originals.
    • Will run mornings through 2pm, so the dome doesn't have to rotate.
  • 2013-05-06
    • Went down to the lab to see if switching the laser detector board helped with the alignment problem. Couldn't reproduce the alignment problem, so I did not switch out the board. Cloudy day, so no data lost.
  • 2013-05-01
    • Welders in the lab - clear spikes seen in C2H2 as a result.

Back to top

April 2013

  • 2013-04-30
    • The data starting this day indicate that the laser is failing - the noise level increases dramatically. (Data were processed on May 16.)
  • 2013-04-17
    • Between 11am-12:30. Coelostat down for mirror cleaning.
  • 2013-04-02
    • Replaced the Spectraphysics laser with the refurbished SIOS laser (for good or for ill)
    • No visible ghost.
  • 2013-04-01
    • LN2 is coming out the sides of the black "puck" lid on the InSb detector - there may be a blockage in the fill tube inside the InSb detector itself.
    • Indeed, the black plastic tube inside the InSb detector that should connect with the main fill tube had disconnected and fallen into the InSb detector. Reconnected, and the system seems to be running.

Back to top

March 2013

  • 2013-03-30
    • InSb isn't staying cold - it seems as though there is an obstruction somewhere in the system.
  • 2013-03-12
    • Took off the M1EW motor cover and found a metal part jammed under the teeth. Removed part and the motor functions properly again.
  • 2013-03-11
    • Coelostat M1EW motor seems to be jammed - can't move the mirror toward the west.

Back to top

February 2013

  • 2013-02-19
    • C2H2 spikes - probably related to welding in the lab.
  • 2013-02-14
    • C2H2 spikes likely related to welding in the lab.
  • 2013-02-12
    • Dome opened and rotated.
    • We opened up the IFS to look at the scanner - noticed that there was pretty awful noise from the encoder motor at 7.5 kHz, but this did not seem to affect the ILS. It started the day looking good, but after several spectra, degraded. We didn't touch any of the mirrors. We did try the alignment kit and despite what we saw in OPUS, the Haidinger fringes were nicely centred.
    • Mounted a square (48" diagonal) piece of foam to the four pickoff mirror tubes (for the sunlight chandeliers) over the IFS to darken the room a bit and prevent dust and dirt from raining down on the IFS. The foam is mounted using bungee cords that wrap around the pickoff mirror tubes.
  • 2013-02-11
    • It seems as though the EPM for the rotation controller is corrupt (it says CF on the screen in the upper dome box). Following Dick's instructions, I replaced the EPM in the rotation controller with the one in the M2EW controller (which is unused), and changed the appropriate parameters. The dome then turned in azimuth when commanded.
    • We had some trouble putting the system back into its daily operational schedule, so we'll see tomorrow if it tries to wake up.
  • 2013-02-08
    • Dome is failing to turn in azimuth. Of course, Dick is away on holiday.

Back to top

January 2013

  • 2013-01-29
    • Added washers to the four locking screws of the fixed cube-corner.  This cured the problem by which the locking screws were bottoming out and could not prevent sideways motion of the cube-corner.  We did not change any of the adjustments, but we did manually move the cube-corner while testing for tightness.  The cube-corner should have returned to its original position on its own.  We'll see what the lamp runs reveal tonight.
    • Notice a "feature" of IFSloop whereby it enters failure mode after entering hold and venting the instrument. To avoid this error, put the container in hold, then issue a SW Status Clear to Vent command, followed by IFS Direct VAC=2.
  • 2013-01-28
    • Put the pump back into the 50L dewar. Need to keep an eye on this to ensure that the heater near the pump hasn't failed. Also, Paul suggested we fill the InSb every day to prevent humid air from building up inside the InSb detector and then flowing back into the 50L dewar.
  • 2013-01-26
    • InSb failed to fill.
    • It appears as though there is a significant ice blockage in the top part of the tube directly underneath the pump. It has been very humid here for the past few days and the system didn't fill yesterday, so I guess this isn't surprising. I read the manual, which only seemed to infer that we may need to increase the height distance set in EEPROM between the pump and the InSb detector to avoid this error in the future. The pump has been left out of the 50L dewar and unplugged.
  • 2013-01-10
    • Dick Treffers back for continued coelostat maintenance.
  • 2013-01-09
    • Dick Treffers accidentally left the cover over the coelostat entrance hole. All recorded spectra will be noise.
  • 2013-01-08
    • Dick Treffers is in town for maintenance on the coelostat.
  • 2013-01-07
    • No InSb cell scan were run during bedtime story state. Norton was updating the algo distribution to solve a few problems related to the coelostat and LN2 drivers, which caused the gain setting on the InSb to remain unset:
    • I'm afraid I interfered a bit in bedtime stories. I decided to do the
      algo distribution before bedtime, since we were napping with the dome
      closed anyway. Only have restarting did it occur to me that we would be
      stuck in InSb_filled (indicating the dewar is full but the algo has not
      set the detector gains) rather than InSb_chilled (gains set), so the
      Monday night InSb cell scan will not occur.

Back to top

December 2012

  • 2012-12-22
    • Norton has worked out most of the bugs in the LN2 fill system, and it seems to be running smoothly. The software waits for the coelostat to open (good_weather) before filling the InSb to avoid filling for no reason, so we get a couple of InGaAs scans before the InSb is ready. We should check the InGaAs scans taken while the InSb is filling to ensure there are no ill effects from this.
  • 2012-12-18
    • Dick in town (2012-12 18-20). Replaced ESTOP relay, so the slit now opens through software. There are other problems with the coelostat, but not so pressing (see Dick's blog:
  • 2012-12-12
    • Tried using the new alignment kit. The Haidinger fringes looked relatively well-centred all the way from ZPD to MOPD. We only slightly tweaked it. No change to the poor line shape.
    • Then noticed that the spectra physics laser wasn't far enough forward on in the laser holder beneath the IFS, so we moved it forward and recentred the beam throughout the system. We then used the alignment kit again and retweaked everything, and the fringes looked nicely centred. No change to the poor line shape. If anything, it is now worse.
    •  MEs on the laser looks fine. Doesn't explain the terrible lineshape.
  • 2012-12-06
    • A workman drilling holes in the coelostat electronics cabinet to install a fan accidentally broke the ESTOP relay (on the dome slit controller). The slit can be opened manually, but not through the software.
  • 2012-12-05
    • Between today's data and data from 2012-12-07 the MOPD changed from 45.01 to 45.00 cm. Was the laser altered?
  • 2012-12-04
    • Vented system to set up alignment kit.  Removed the HCl cell and holder temporarily to measure area for a plate to hold the laser speckle reducer.  Replaced cell and holder, pumped down system, and reinitialized software; it is cloudy, however, so no data was lost.

Back to top

November 2012

  • 2012-11-29
    • In preparation for the AGU next week, I'll set up the screenshots to last a couple of weeks:
      • C:\Users\TCCON\Downloads\nircmd\nircmd.exe loop 4000 300000 savescreenshot C:\Users\TCCON\Dropbox\Screenshots\current.png
      • C:\Users\TCCON\Downloads\nircmd\nircmd.exe loop 330 3600000 savescreenshot C:\Users\TCCON\Dropbox\Screenshots\scr~$currdate.yyyyMMdd$-~$currtime.HHmmss$.png
  • 2012-11-09
    • I've kludged together a monitor program to take screenshots of the Norhoff app every minute to determine whether the InSb detector is cold. I used an executable called nircmd.exe on the Windows laptop ( and ran a loop to save a screenshot every 60 seconds - this image is saved to a dropbox folder as current.png. I am also running it to take a screenshot every hour, also saving to dropbox, and that image will have a dated name, so they are not replaced. This will allow me to see whether the InSb detector is cold, so I can send sw status InSb Chilled. These will run all weekend, but not much longer:
      • C:\Users\TCCON\Downloads\nircmd\nircmd.exe loop 70 3600000 savescreenshot C:\Users\TCCON\Dropbox\Screenshots\scr~$currdate.yyyyMMdd$-~$currtime.HHmmss$.png
      • C:\Users\TCCON\Downloads\nircmd\nircmd.exe loop 4000 60000 savescreenshot C:\Users\TCCON\Dropbox\Screenshots\current.png
  • 2012-11-07
    • The autofill system starts counting when it is finished filling, so a 1-day repeat would actually change by ~+10 minutes every day. I've set the repeat time to 23h45m to move it back a bit, to handle the weekend, so we're sure it's cold by 9am. Then we can manually run the fill system on Monday to restart the clock (I think).
  • 2012-11-05
    • Put LN2 autofill on 1-day repeat at 8:40 am PST. We should ensure that it has worked consistently all week, and then we can log in and start InSb scans manually on the weekends.
  • 2012-11-02
    • J-F put lock washers under the LASA and LASB detector assemblies to better centre the laser signal on the detectors.
    • The line shape looks okay (not great, see this figure, but not as bad as it has been recently).
    • My opinion: slowly back away from the instrument and put it back into Play. Continue alignment when the alignment kit is assembled.
    • Coelostat was oscillating - may be poor data. Dick solved the problem late in the afternoon.

Back to top

October 2012

  • 2012-10-29
    • The coeolostat is having its UPS installed, so it will be down for several hours today.
  • 2012-10-15
    • Si data was peaked and improved significantly. The InSb was peaked as well, but did not improve by much (the change to peak InGaAs did not affect the InSb).
    • The dome has been having significant trouble staying open since October 5.
    • Opened up the instrument to investigate the vacuum leak. Found a hair across the beamsplitter o-ring. Removed it and cleaned the o-ring. Put vacuum grease on the o-rings that seal the lid onto the InSb detector coupler. The combination seemed to do the trick.
  • 2012-10-12
    • J-F and I looked at this again. We are not able to improve the line shape, even by making the field stop smaller.
    • Tested to see if this could be a phase error problem by adding a phase offset in opus-ipp: the line shape cannot be corrected that way.
    • We peaked the InGaAs signal by walking the parallel beam in the detector compartment.  But this left the Si misaligned (it is missing its DC test point output which would have made peaking both detectors easy).
    • IFSloop back in play.
  • 2012-10-09
    • Jean-Francois and I looked into the stability of the interferometer. Found nothing amiss except:
      • The rails were not level (we leveled them)
      • The screws locking the fixed cube corner seemed didn't seem to actually lock anything
      • We tried to realign, but got nowhere. IFSloop off.

Back to top

September 2012

  • 2012-09-26
    • The lineshape looked pretty poor in the morning. Went to lunch, came back and looked at the InSb detector. Looks okay (not good, but not horrible). Tried InGaAs. Now looks okay (not good, not horrible). Pumping down instrument and putting in Play. Will order a telescope alignment kit from Frank Hase and hopefully we'll get to the bottom of this.
  • 2012-09-25
    • Laser is clipping on the scope for this beamsplitter (-9V to clipping at 11.75V - the max amplitude is 20V). To reduce the gain, go to:
      • IFS125 webpage, Service, Edit Hardware Configuration, Edit Beamsplitter Eeprom. Uncheck the Laser amplifier on: box and restart IFS.
    • Replaced original beamsplitter into the IFS - no change in the laser amplitude from before.
  • 2012-09-24    
    • Shut down IFSloop in the afternoon to try a new beamsplitter. So far, this beamsplitter is not any better.
  • 2012-09-20
    • The alignment is not great using 1.0mm APT, but it's terrible on 1.3mm APT. So I will change the InSb settings to 1.0mm APT while we figure out the alignment issue.
      • APT changed to 1.0 mm for InSb at 4 pm PDT. The signal levels were ~0.4 MXY, and now are:
      •  0.19 MXY.
    • Use today's data with caution: Coleen was installing the foam on the IFS and may have moved the IFS slightly.
  • 2012-09-18
    • Not happy with the alignment, but will try to start IFSloop again. Installed the bandpass filter in front of InSb and have increased the APT to 1.3 mm (up from 0.5 mm). Gain remains the same. Recompiled the software. The new spectra look nice in the zeros, but have a bit of channeling.
  • 2012-09-07
    • Have been trying to align the IFS with the replacement SIOS laser from Bruker, but with no end of trouble. See the Alignment History page for more details.
  • 2012-09-04
    • Turned off laser. Waited 30 minutes. Turning it back on resulted in it locking and then losing lock repeatedly. Will change out laser tomorrow. While the instrument is vented, set up the optical filter stand for the InSb detector. Will install after the laser is replaced and the system is realigned.
  • 2012-09-03
    • 8:20 am: Scanner error. The laser was in error. Tried rebooting twice but the laser would not achieve lock.

Back to top

August 2012

August data appears to be poor due to the laser slowly failing...

  • 2012-08-26
    • 1:20 pm: Scanner error. Went into lab to find laser "error" light on yellow instead of green. Restarted the laser, and when it attained lock (the "error" light went green) restarted the IFS. Went into play.
  • 2012-08-24
    • 9 am: Shut down software and plugged router directly into UPS. Hopefully this will avoid problems when we have power glitches.
  • 2012-08-22
    • 10:15 am: Installed new Bruker firmware for the InSb detector. First scan after upgrading on the InGaAs/Si was in AC mode, second scan in DC mode. After cooling InSb detector and reiniting, the first InGaAs/Si scan is in DC, InSb is in DC.
    • Also punched a hole in the router for nc.
  • 2012-08-21
    • Scanner error, after which the scans were recorded in AC mode. Rebooting the IFS got the second scan after booting in DC. Should probably install the patch for the InSb detector.
  • 2012-08-20
    • Bypassed the herc's external ethernet port and plugged the router directly into the motherboard ethernet port. No more dropped packets and the transfers to ccycle are about 1.5 MB/s.
  • 2012-08-18
    • Brought a mac and a windows laptop to the lab for continued testing of the slow transfer speeds. Learned that both windows and mac drop ping packets when connected inside the router to almus. However, the windows machine was able to scp (via cygwin) much faster than the mac. Solaris machines also seem to transfer quickly.
    • Pinging the IFS125HR does not drop packets, nor does pinging ccycle.
    • Switched port on router and no change to the symptoms.
    • Looks like the problem is the ethernet cable either between the motherboard and the external connector on almus, or the motherboard itself.
  • 2012-08-17
    • Transferring data to ccycle by: transferring from almus to the laptop, transferring from the laptop to ccycle. Almost as bad as shipping disks across the country!
      • Worse: even though I have the "Preserve Timestamps" function enabled in WinSCP, the timestamps are all wrong.
  • 2012-08-16
    • Attempting to figure out why scp'ing data from almus to ccycle is so slow (~4-12KB/s)
    • Symptoms:
      • scp from almus to ccycle is slow (whether logged into almus or ccycle): ~12KB/s
        • slow also for lamont, nicolet
      • scp to the laptop (via WinSCP) is fast ~2.5MB/s
      • scp to my laptop at home (WinSCP) is pretty fast ~800KB/s - over slowish wireless
      • scp to Norton's laptop (WinSCP, I believe) is reasonably fast ~200KB/s - across the country
      • scp to a mac pro laptop (scp) is slow
      • Seems like there's a mismatch between scp on linux/mac and QNX, but not between WinSCP and QNX.
    • In WinSCP, we get this error:
      Command 'groups'
      failed with return code 126 and error message
      sh: <stdin>[30]: groups: cannot execute - No such file or directory.
    • But WinSCP copies quickly, from everywhere
  • 2012-08-10
    • Maintenance thought it would be a great idea to lean on the IFS while accessing panels in the solar tower. Look for misalignment starting sometime around 10:15-11:00 am.
  • 2012-08-07
    • Plugged the LN2 autofill system into the power strip that has the IFS and laser on it as well. This in turn plugs into the UPS. We have been having brownouts and I'd like to avoid freezing the LN2 pump.
  • 2012-08-06
    • ifsreduce.log file format has changed:
    • $ cvs log QNX6/anal/ifsreduce
      revision 1.5
      date: 2012/08/07 00:03:54;  author: blavier;  state: Exp;  lines: +1 -1
      Changed the way the remaining disk space is reported to better match
      the QNX4 output:
        1) Use 1024-byte units instead of 512-byte.
        2) Report only the mounted partitions, not the raw devices.
        3) Eliminate trailing spaces at the end of each line.

Back to top

July 2012

  • 2012-07-30
    • Pressure seems to have held in the IFS all weekend. I used the LN2 autofill to fill the InSb dewar - this seemed to work without increasing the IFS pressure.
  • 2012-07-27
    • Pressurized the InSb detector with dry N2 and used snoop to find the location of the leak. A leak is definitely present at the interface between the dewar and the sleeve that fits in the dewar (the bellows attaches somehow to this sleeve). Cleaning o-ring on underside might help.
      • Installed the "Crounse Kludge" which involves stacking o-rings. The metric o-rings that came with the Bruker seem to be slightly thicker than the standard o-rings we have in the lab - stacking them will hopefully do the trick.
      • Put the InSb back into the system and pumped down without cooling the detector to test the top seal with the IFS lid. After 2 hours, the IFS only leaked from 0.25 hPa to 0.47 hPa.
      • At 2:30 pm, very slowly filled the InSb detector to try to avoid freezing the o-rings. The pressure rose to 0.5 hPa (from 0.47) almost immediately. Will keep an eye on it. At 3:30, the pressure is still 0.49 hPa. At 5pm it's at 0.6 hPa.
  • 2012-07-26
    • Indeed, the IFS is holding pressure now (increasing by ~0.2 hPa in 3 hours). The InSb assembly is to blame.
  • 2012-07-25
    • The IFS leaks have continued to get worse, so I've removed the InSb detector (around 4 pm PDT) and replaced the lid with a blank. This is a test to confirm that the leak is coming from the InSb part of the system and not elsewhere.
  • 2012-07-10
    • Today was a mostly sunny day, but it's probably the first day we've measured that hasn't been completely clear. I'm only noting this because we don't really have an FVSI value to work with.
  • 2012-07-09
    • 5:15 pm PDT: Installed new Bruker firmware update to try to solve the problem of the Bruker losing its detectors. Now we wait to see if it happens again... or if a new problem turns up!
  • 2012-07-05
    • 12:30 am PDT: 747.4 mmHg at 22 C in the lab
    • IFS lost its detectors again today. Rebooted IFS and sent sw status reinit seemed to do the trick. The first InGaAs/Si and InSb scans were in AC mode.
    • Reset PGN=3 for InSb (PGN=1 was the wrong setting and had been used up until now.)
  • 2012-07-04
    • 9:35 am PDT - 746.4 mmHg and 22C in the lab
    • Norton/I fixed some problems with the InSb scans. Ran reduce around 2:30 pm PDT to split the data up for today. Here are the differences in the and IFS.tma files (IFS*.20120704 is what was previously running):
      • in /home/citco2/src/QNX6/TM/Config
      • $ diff -U 6
        ---  2012-07-04 13:36:49.000000000 -0700
        +++ 2012-06-24 12:12:48.000000000 -0700
        @@ -141,17 +141,17 @@
           SolarInSb => {
             SFM => 'SolarInSb',
             DTC => "16577",
             RES => "0.020000",
             NSS => "2",
             APT => "500",
        -    AP2 => "500",
        -    VEL => "7500",
        +    AP2 => "1000",
        +    VEL => "10000",
             LPF => "10000",
        -    GNS => "1"
        -#    PGN => "1",
        +    GNS => "1",
        +    PGN => "1",
         #    PHR => "1.0",
         #    HFW => "7898.0000"
      • $ diff -U 6 IFS.tma IFS.tma.20120704        
        --- IFS.tma     2012-07-04 14:09:18.000000000 -0700
        +++ IFS.tma.20120704    2012-07-01 14:45:17.000000000 -0700
        @@ -38,14 +38,12 @@
         State Init_Choice "Startup.tmas" {
              Hold until ( SWStat == SWS_TIME_WARP ) or 2:00;
           +1 > IFS Time Synch
         #    > Sun Tracker Synchronize Time
              > IFS Set Preamp Gains 0 0
        -     > IFS Direct DTC=0x40C1
        -     > IFS Direct PGN=1
              > IFS Read Status
           +2 { if ( Convert(Sol_ele) < 0. ) Validate Init_Close; }
           +2 depending on ( Bad_Weather ) { Validate Nap; }
          +10 Validate Ready_To_Play;
            # These two will never occur, but are here
            # to improve the visualization

 Back to top

June 2012

  • 2012-06-30
    • 10:10 am PDT: 21C and 746.7 mmHg in the lab.
  • 2012-06-29
    • IFS lost its detectors again. Rebooted the IFS and that "fixed" the problem.
    • 9:15 am PDT: 748.6 mmHg at 21.5 C in the lab
    • Norton found a bug in the Pump_P reader, and has fixed the digital noise problem.
  • 2012-06-28
    • In the lab at 7 am to let the electrician into the tower to try to fix the central pump. The problem seemed to be down underneath the floor, and not in the tower, but the tower floor panel was lifted up and some work was done in the junction box on the N side of the tower. (A waterproof wire in the pit was sliced, causing a short. They're pulling the wire out and will have to purchase a replacement.)  If the interferograms aren't good this morning, it may be due to the activity in the tower (though they tried not to interfere with the instrument).
    • 2:24 pm PDT: 747.8 mmHg at 22 C
  • 2012-06-26
    • The vacuum leak rate was about 0.5 hPa per hour, which is ridiculous. I've vented the instrument and cleaned the o-rings on the source, interferometer and detector compartments again. I saw a small speck of dust on the InSb o-ring, which may (?) have been the cause, but we'll find out.
    •  9:45 am PDT: 21.5C and 745.8 mmHg in the lab
    • It's really humid in the tower. I can hear water sloshing around.  I've opened most of the doors to try to dry it out a bit while pumping. I probably should have had the doors open when the lids were off to prevent moisture from getting into the IFS.
    • When I topped up the InSb detector with LN2, the pressure (which is still pumping down) jumped up by ~0.1 hPa. This did not happen when I first started filling.
    • 3:50 pm PDT: scanner error. Everything looked okay when I got down to the lab (no error LEDs), but when I tried to send the scanner into front adjust mode or back adjust mode, it went into error. I'm venting the instrument to try to figure out what's up. Hopefully just a reboot of the scanner will help...
    • I'm not sure why the scanner was claiming it was in error - it did touch the second optical switch briefly, but reset itself and did both front short adjust mode and back short adjust mode without problem.
    • Norton implemented an algorithm change that isn't so sensitive to these minor errors.
  • 2012-06-25
    • 8:45 am PDT: Came into the lab to beeping. The herc was recording HTML slices.
    • Sent sw status shutdown quickly. Rebooted the IFS.
    • From /home/citco2 sent ./ifsdoit -sa (without the wait); sw status time warp
    • Taking InGaAs/Si scans now. First scan was in AC mode. Second scan was in DC mode.
    • Will finish filling InSb with LN2 in an hour or so. (It's only partially filled right now.)
    • 10:00 am (ish): filled InSb, put software into hold, ran sw status pump down
    • once back in play, issued sw status InSb Chilled. Now alternating between Solar and SolarInSb.
    • 10:30 am PDT: 21.4 C and 748.0 mmHg in the lab.
  • 2012-06-24
    • Norton implemented the InSb-InGaAs/Si switching option in the QNX6 software! Once the InSb detector is filled, just type sw status InSb Chilled, and the algo will begin switching between InGaAs/Si and InSb for the next 12 hours. In the evening, the IFS will automatically go into InGaAs/Si idle mode, which will move the mirror off the InSb detector, in order to protect it from having direct sunlight when warm.
    • The diag_DTC webpage contains a NOT READY/READY field for InSb that switches over to READY when the detector is cold. This will eventually be used to determing whether to take InSb scans, but for now, manually entering the InSb Chilled command is the way to do it.
    • 12:30 pm PDT: 746.7 mmHg and 22 C in the lab
  • 2012-06-22
    • First attempt at automated measurements. Currently just running InGaAs/Si, but it seems to be working!
    • I've asked Dick to permanently switch over to 7-day per week coelostat observations.
  • 2012-06-21
    • 4:15 pm PDT: 22.5C and 744.9 hPa in the lab
    • Starting measurements at 4:15 pm.
    • Pressure is not holding in the IFS - it was up to 9 hPa since last pumpdown 24 hours ago. Probably need to clean the o-rings again.
  • 2012-06-20
    • Didn't get to measurements until 3:40 pm PDT. Stopping at 5 pm today for Norton.
    • 3:40 pm PDT: 744.0 hPa and 22C in the lab.
    • It looks like the detectors were found.
  • 2012-06-19
    • No measurements in the morning - filled InSb over lunchtime. LN2 dewar in the lab was empty. Had to pump down IFS twice (it was 20hPa after the week-long break, and up to 3.6 hPa after the morning was through.)
    • 1:30 pm PDT - readjusted cistern height; 22 C and 745.1 mmHg in the lab
    • Image of sun on 1st floor screen seems to be in a different location - look for S-G shifts in spectral residuals.
    • First attempt to take measurements through OPUS: InSb-DC worked, InGaAs/Si-DC did not. Upon rebooting the IFS, neither InSb-DC nor InGaAs/Si-DC worked. A second reboot brought both sets of detectors back online.
  • 2012-06-08 through 2012-06-18
    • No measurements as no one is in town to take them.
    • Pump left switched off with mechanical outlet switch but plugged into pump optical switch
    • Bowl left over the input tube to prevent sunlight from entering instrument
    • Left under vacuum - 0.411 hPa at 5:15 pm June 7.
  • 2012-06-07
    • 3pm PDT - 22C and 743.7 mmHg in the lab.
    • First solar and cell spectra recorded using the QNX6 system!
  • 2012-06-06
    • Removed (and later replaced) tube to measure diode box to make a blank for the Caltech instrument (in order to have a complete solar tracker spare)
    • Replaced resistors in InSb detector so it doesn't saturate:
    • Preamp no longer saturates on any of A, B, C with a 1mm aperture, however, there is significant detector nonlinearity.
    • Lots of noise seen below 2000 cm-1 (on all three detectors)
    • Using new preamp gain C with 0.8 mm aperture shows better nonlinearity (though the zero levels still slope up a bit) and the noise level below 2000 cm-1 improves.
    • The best seems to be REF+0.5 mm pinhole. Will run with those settings today.
    • Starting with ci20120606sgeeac.007 & ci20120606sgeeaa.002; 008/003 after pumpdown.
    • 12:18 pm PDT - 21 C and 743.6 mmHg in the lab.
  • 2012-06-05
    • 9:50 am PDT - 21.5C and 746.4 mmHg in the lab.
    • Regular measurements stopped at 1 pm to try to measure Venus.
    • Venting instrument and removing HCl cell - will require re-alignment of input and output optics: will carefully mark the cell before removal.
    • Venus appears on limb at 10:11 pm UTC.
    • (Note that ci20120605s0eeav.007 has the neutral density filter left in.)
  • 2012-06-04
    • At 9:22 LT - 21.2 deg C, 746.5 mm Hg in lab.
    • No measurements recorded today since the coelostat mirrors and telescope lens were being cleaned.
    • Checked whether solar beam needed to be recentred on the pinhole. Adjusted only slightly.
    • Measurements of InSb detector:
      • 63.6 mm diameter of dewar
      • 12 mm threaded through-hole
      • 8.44 mm inner diameter of threaded through-hole
      • 7 mm inner diameter fitting that screws directly into InSb detector
  • 2012-06-03
    • In lab in late afternoon (5-6pm) checking out feasibility of Venus transit measurements.
    • De-centered coelostat.
    • Vented IFS, cooled down InSb, closed down pinhole to 0.5 mm and moved iris out of the way.
    • Replaced iris and pumped down IFS before leaving.
    • Took two spectra: centered on a sunspot w/ 0.5 mm pinhole and no iris: ci20120603sgeeav.001; recentered on sun with iris and 1 mm pinhole: ci20120603sgeeac.001. Spectra attached.
  • 2012-06-01
    • 1:15 pm PDT - 22C and 743.9 mmHg in the lab.
    • 2:05 pm PDT - added the DC detector option to InSb. Started measurements. AC until now, DC after, starting DC with ci20120601sgeeac.033.
      • To add DC option:
        • Webpage --> Service --> Edit Harware Configuration --> Edit ANA15 board data
        • Hit next board until we got to InSb, copied config to pdf (attached). Changed name to include AC.  Checked boxes (left channel and simultaneous right+left channel). Saved config.
        • Copied values into "next board" fields and changed detector name to include DC.
        • Saved and rebooted IFS computer. DC InSb option appears in OPUS.
      • InSb DC peak-to-peak IGM values (~6000 counts) are lower than InSb AC (~8000 counts). Can probably open up iris a bit.
      • The large noise below 1850cm-1 is reduced.

 Back to top

May 2012

  • 2012-05-31
    • 9:26 PDT - 21.0 C and 745.8 mm Hg in the lab.
  • 2012-05-30
    • 9:45 am PDT - 21C and 748.5 mmHg in the lab.
    • No "existential crisis" today. DC options were available.
  • 2012-05-29
    • At 9:32 PDT - 21.2 C and 747.8 mm Hg in lab.
    • Tried to set up the InSb detector. Was saturating on the lamp: changed the gain to REF, which seemed to help, but still required an iris in the parallel beam between the detector OAP and the detector selection OAP. Taking solar scan alternating between InGaAs/Si and InSb. InSb does not appear to be DC-enabled.
    • The DC options on the InGaAs and Si detectors did not appear upon starting OPUS. Rebooting the IFS computer did the trick.
    • Stopped measuring at 2 pm to install the temperature sensor covers and neaten up the wiring of the temperature sensors heaters.
    • Restarted measurements at 3:45 pm - 5 pm.
    • It's about 8 InGaAs/Si+InSb scans per hour.
  • 2012-05-25
    • Changed Zeno config again this morning to add a decimal place to the Pyranometer field.
    • Cloudy for most of the day, so no measurements.
  • 2012-05-24
    • No measurements this morning.
    • Jean-Francois and Jess came to caltech to finish installing the temperature sensors and heaters. The power system was put into place: batteries and UPS were connected. The laser and IFS are separately plugged into the UPS via the terminal strip inside the IFS cage.
    • Note: upon rebooting the herc (because the keyboard appeared no longer connected) the IFS and S/T on fiber switches were off. I logged in, started telemetry start and the lights turned on.
    • Now pressure inside IFS appears to be "unstable". Evacuating. After pumpdown, it appears stable.
    • Measurements began at 3:11 pm PDT. First interferogram recorded in AC mode!
    • Second scan in DC mode.
  • 2012-05-23
    • No measurements today.
    • Jean-Francois and Jess came to Caltech to install the temperature sensors and heaters. All but two of the heater paint patches are removed. The remaining two patches are on the interferometer compartment lid and appear to have been machined differently than the other lids. This is the 4th coat of paint stripper.
    • The pump control box was installed and seems to work, although the herc turns the pump on when you send the power pump off command.
    • Tape has been removed (I gave up after the 4th coat) and everything was wiped down with ethanol.
  • 2012-05-22
    • 10:23 am PDT - 21C and 745.7 mmHg in the lab
    • 9:20 am Arrived in Lab, turned on pump, booted laptop, pumped down IFS, started OPUS, turned off pump.
      • No DC detector options. Rebooting laptop didn't help so rebooted IFS computer - DC options reappeared. First scan was in AC, but subsequent scans were in DC.
      • Communicated with Norton who tried an IFS read status command last night - not likely the problem, but worth a try
      • Shut down OPUS and Norton sent IFS read status. Restarted OPUS and DC detectors still enabled. Took a scan
      • Spoke with Jean-Francois on the phone for a few minutes, and by the time I got back to OPUS, the DC detectors were no longer enabled.
      • Rebooted the IFS125 computer. DC options returned. Tried a "check status" but no signal displayed and constant beeping. Recorded a full report from the webpage. Not sure it says anything useful.
      • Attempted to scan; given Instrument Error message from OPUS. Restarted OPUS. Check signal now works and shows DC mode.
      • Taking FWD/REV scan pair: DC mode
      • Measuring 90 FWD/REV pairs to bring us to ~5pm when Norton will take over 
  • 2012-05-18
    • 3:30 pm PDT began acquisition (dome opened)
    • 3:32 pm PDT - 21C and 744.1 mmHg in the lab
  • 2012-05-17
    • 8:50 am PDT - 21C and 744.1 mmHg in the lab
    • 2:52 pm PDT - 3:07 pm PDT - Solar tracker was not tracking, all data should be thrown away. (Around spectra .096-.100.)
  • 2012-05-16
    • 9:10 am PDT - 20.5C and 745.5 mmHg in the lab
  • 2012-05-15
    • 8:52 am PDT - 20.5C and 746.4 mmHg in the lab
    • 12:31 pm PDT - 21C and 745.8 mmHg in the lab
    • 081-100 inclusive are bad because I closed the dome to change the Zeno config.
    • Changed the Zeno output configuration around 2:30 pm. Swapped around Pres and SolarRad, VBATT and BIT to match the string of the other Zenos.
    • Moved the Herc and Power supply to the rack inside the tower. The new ethernet extension cable (plugged into the ethernet port on the motherboard) seems to work.
  • 2012-05-14
    • 9:08 am PDT - 20.2C and 748.5 mmHg in the lab
    • 12:30 pm PDT - 21C and 747.8 mmHg in the lab
    • 4:33 pm PDT - 21C and 746.7 mmHg in the lab
  • 2012-05-12
    • 5:30 pm PDT - 21C and 745.7 mmHg in the lab
  • 2012-05-10
    • 10:20 am PDT - 20.8C and 745.8 mmHg in the lab
    • 3:20 pm PDT - 21C and 744.5 mmHg in the lab
  • 2012-05-09
    • Change laptop time to GMT - Done, although the header of the spectrum clearly had GMT-7 yesterday. There must be something that can be done to read that automatically.
    • Look for NTP server connection - Using
    • Figure out how to request FWD-REV pairs - just request 2 scans instead of 1
    • 9:00 am PDT - 21C and 745.2 mmHg in the lab
    • Renamed interferograms to ciYYYYMMDDsgeeaA.XXX
    • Interestingly, the FWD-REV pairs seem to have NPT=NPTVEC:
      • Data Block IgSm/2.Chn. length 3034242
        Data Parameters IgSm/2.Chn.
          DPF2: 1
          NPT2: 3034242
          FXV2: 0
          LXV2: 3034241
          CSF2: 1
          MXY2: 0.388662695884705
          MNY2: 0.061490073800087
          DXU2: PNT
          TIM2: 16:24:04.276 (GMT+0)
          DAT2: 09/05/2012
  • 2012-05-08 - Finally pumped down successfully. Coleen noticed the problem: a missing o-ring under the teflon screw cap for the InSb detector. Not sure where it went...
    • 3:05 pm PDT - 21.8C and 744.2 mmHg in the lab
    • NOTES about opus-ipp with this new IFS125HR.
      • The data block length does not equal the NPT length (NPT=NPTVEC-1)
        • Data Block IgSm/2.Chn. length 1517122
          Data Parameters IgSm/2.Chn.
            DPF2: 1
            NPT2: 1517121
            FXV2: 0
            LXV2: 1517120
            CSF2: 1
            MXY2: 0.421160578727722
            MNY2: 0.0665391087532043
            DXU2: PNT
            TIM2: 10:53:23.108 (GMT-7)
            DAT2: 07/05/2012
      • I "fixed" this in OPUS-IPP by changing the following if-statement
        •               if(npts(channel).ne.nptvec(slicnt)+1) then ! The +1 is because of the new caltech instrument that always has one more NPT value than length of the block... This ought to be fixed!
                          write(stringa,'(i11)') nptvec(slicnt)
                          write(stringb,'(i11)') npts(channel)
                          write(*,'(8a)')'Error: ',path(1:lnbc(path)),
               &           ' has mismatch between NPT=',stringa(fnbc(stringa):11),
               &           ' and ',chan(channel)(1:lnbc(chan(channel))),
               &           ' point count of ',stringb(fnbc(stringb):11)
        • This "fix" will of course break everything else.
    • Noticed when processing yesterday's data that there don't appear to be reverse scans recorded. Not sure why.
    • NOTES about running GGG:
      • The laptop time is local, so the time stamps in the header are also local time. Kludged create_runlog to add 7 hours.
  • 2012-05-07 - Solar spectra acquired, put in 35 mm aperture stop next to cloudy CaF2 window. Getting ~11700 counts on ADC using OPUS with 1 mm aperture.
    • 10:10 am - 20.5C and 747.0 mmHg in the lab
    • 12:21 pm - 20.8C and 746.4 mmHg in the lab
    • 2:02 pm - 20.8C and 745.9 mmHg in the lab
    • 2:45 pm - pump down IFS125HR
      • There seems to be a leak in the IFS. It seems to equilibrate at around 30 hPa.
      • Cleaned all lids and o-rings (again). Lots of black residue on the lids.
      • No luck. Still unable to stay stable - started at 1.7 hPa and was back up to 6 hPa within a few minutes. Vented.
    • 4:33 pm - 20.5C and 745.3 mmHg in the lab
  • 2012-05-04 - SEAK4RS overpass. Variable cloud, basically impossible to check whether the solar alignment is good.
    • 8:10 am - 19.5C and 748.8 mmHg in the lab; 994.12 hPa; Zeno=987.1 hPa
    • 10:40 am - 19.5C and 749.8 mmHg in the lab; 995.45 hPa; Zeno=988.3 hPa
    • 12:05 pm - 19.5C and 749.6 mmHg in the lab; 995.18 hPa; Zeno=988.3 hPa
    • 2:30 pm - 19.5C and 749.1 mmHg in the lab; 994.52 hPa; Zeno=988.0 hPa
    • Spiral occured around 3 pm.
    • 4:00 pm - 19.8C and 748.5 mmHg in the lab; 993.68 hPa; Zeno=987.1 hPa
  • 2012-05-03 - Minimized laser sampling ghosts. See the Alignment History page.
Time Location Temperature Princo
hPa Zeno
Zeno Pressure
10:16 am PDT Roof (~230masl) 20C 747.1 743.91 991.80 16.7 988.2 3.6 DW
10:17 am PDT Roof 20C 747.9 744.7 992.85 16.6 988.2 4.65 CR
10:32 am PDT Sub-Basement (~210masl) 20.5C 748.5 745.23 993.56 17.2 988.3 N/A DW
10:33 am PDT Sub-Basement 21C 748.5 745.17 993.48 17.2 988.3 N/A CR
  • 2012-05-01 - Aligned IFS125HR. See the Alignment History page.
    • Cap came off the pyranometer, so values after this date should be correct.

 Back to top

April 2012

  • 2012-04-20 - Coleen finished installing the mirror mounts and mirrors in the column
    • First spectra (just tweaking the input mirrors and the internal flat mirror and OAP to centre the beam on the 0.8 ppm aperture). Nothing to write home about, but attached is the first solar spectrum.
    • Recorded at 2:37 pm, clear sky.
    • Notes:
      • The encoder is making a horrible grinding noise.
      • There is no DC option available in OPUS for the InGaAs/Si detectors
      • The .xpm file is attached to this page (and is located in Users/TCCON/Bruker/ on the toughbook)
      • Source entry window is very dirty / cloudy.
    • I will leave the bruker vented over the weekend, as we will want to tweak the alignment and access the detectors and encoder next week

 Back to top

December 2011

  • 2011-12-12
    • Installation of the FTS in the solar tower will commence.
  • 2011-12-08
    • Brian Drouin at JPL filled the HCl (3.753 Torr or 500.358 pascal HCl).

 Back to top

October 2011

  • 2011-10-17
    • FTS delivered to Caltech, but installation postponed until solar tracker functioning.

 Back to top

March 2011

  • 2011-03-18
    • Started purchase order for new FTS system to be installed at Caltech

 Back to top

Tag page

Files 24

FileSizeDateAttached by 
Email sent to Gregor and Axel about 125HR interferogram noise on 20150316
1660.81 kB10:49, 17 Mar 2015dwunchActions
Response to Gregor's requests from email on 20150316
84.14 kB10:50, 17 Mar 2015dwunchActions
 7.5 kHz Scan_1.m4a
Audio file of the noisy scanner.
2.04 MB13:54, 20 Apr 2015dwunchActions
 Baro. Cr. Spread. 2000.Caltech.xls Preview
Princo Pressure Conversion Spreadsheet
51.5 kB10:17, 3 May 2012dwunchActions
One of the first solar spectra recorded.
19.58 MB13:38, 20 Apr 2012dwunchActions
Sun-centred spectrum.
15.58 MB17:37, 3 Jun 2012dwunchActions
Sunspot-centred spectrum.
15.58 MB17:37, 3 Jun 2012dwunchActions
First cell spectra recorded with QNX6
353.4 kB15:13, 24 Jun 2012dwunchActions
First solar spectra recorded with QNX6
218.76 kB15:13, 24 Jun 2012dwunchActions
Initial evacuated ILS, before mis-alignment (but same parameters)
6.05 kB21:55, 18 May 2016coleenActions
Initial vented ILS, before mis-alignment (but same parameters)
5.99 kB21:55, 18 May 2016coleenActions
ILS after returning to start position, but before doing laser alignment (vented)
5.98 kB21:55, 18 May 2016coleenActions
Final ILS of vented FTS
5.96 kB21:47, 18 May 2016coleenActions
Final ILS of evacuated FTS
6.03 kB21:47, 18 May 2016coleenActions
XPM file
37.6 kB13:40, 20 Apr 2012dwunchActions
InSb DC xpm file.
61.49 kB17:37, 3 Jun 2012dwunchActions
ILS from 2012-10 through 2013-07 using LINEFIT14
5.96 kB09:55, 13 Aug 2013dwunchActions
Detector setup page before InSb DC options added.
34.34 kB13:10, 1 Jun 2012dwunchActions
Comparison of the InSb alone versus in dual channel with InGaAs
302.21 kB14:15, 21 Oct 2015coleenActions
OPUS macro to run InGaAs/Si and InSb DC iteratively.
1258 bytes17:37, 3 Jun 2012dwunchActions
InSb dewar pumping procedure
1229.13 kB13:36, 20 Aug 2014coleenActions
Full report after DC options reappeared but check signal would not work (constant beeping).
155.62 kB09:43, 22 May 2012dwunchActions
 Sales Order_SO52389_1554166319069.pdf
New camera which was installed in the Caltech library (April 2019)
12 kB14:05, 8 Apr 2019m_kielActions
Site visit report, Jan 18/19, 2016
421.18 kB12:48, 29 Feb 2016coleenActions