SHARC II User's Manual


Help Topics:

setting up for the night
special considerations for the first night
observing procedure
shutting down for the night
cryogen servicing
miscellaneous notes

Setting Up for the Night

Recycling the 3He Fridge

NOTE:  If this is the first night of a SHARC II observing run, please read below in the "warm detector" section.

This procedure should be started at least 2 hours before observing. Recycling is normally handled by the first-half observers, but please communicate your plans with the CSO staff and any second-half observers. The recommended starting time is 2 PM.

IMPORTANT: As of January 2011, cycling is done on sharcii again, not hapuna (which has been decommissioned).

Recycling is done electronically. The Neocera LTC-21 needs to be powered up, connected to the sharcii computer with a serial cable, and connected to the cryostat. Two processes need to be running on sharcii for the procedure to take place properly: a background server which logs temperatures and controls the Neocera, and a graphical client. Only one server should be running, but multiple clients can be running. To see what processes are already running, log onto sharcii -- "ssh sharc@sharcii" -- and type "sharcii% ps -ef | grep java". The server and client each show up as a java1.2 process. If in doubt about what's running where, kill all of the java1.2 processes: "sharcii% kill -9 {PID}". For example:

sharcii% ps -ef | grep java
sharc 9744 9743 0 10:30:41 pts/0 45:01 /opt/java1.4.0/bin/java -Xms64m -Xmx256m -Djava.library.path=./lib/native -cp .
sharc 9840 1 2 15:00:11 pts/1 0:02 /usr/java1.2/bin/../bin/sparc/native_threads/java -Djava2d.font.usePlatformFont
sharc 9849 9818 0 15:00:38 pts/1 0:00 grep java
sharc 9828 1 0 15:00:41 pts/1 0:02 /usr/java1.2/bin/../bin/sparc/native_threads/java -Djava2d.font.usePlatformFont
sharcii% kill -9 9840
sharcii% kill -9 9828

The sharcii computer can be found in the rack in the Receiver Lab (AOS room).

If the server is not running, log onto sharcii: "ssh sharc@sharcii". Execute "sharcii% cd ~/recycle" and "sharcii% cycle". Wait a few seconds, and a one-line message similar to "10 Minutes;Both; 500;One Shot;0;" should appear. (If you haven't noticed already, sharcii is running Solaris, so typical UNIX commands will work.) WARNING: Logging out of the session which started the server will abort the server.

If the client is not running, log onto sharcii. Execute "sharcii% cd ~/recycle" and "sharcii% gclient". In a few seconds, the graphical interface should appear. WARNING: Logging out of the session which started the client will abort the client.

For a standard cycle, the "3He cycle only" and "4He cycle only" boxes should NOT be selected. To begin a cycle immediately, make sure "Start immediately" is selected, and press the "Start" button. To start a delayed cycle, unselect "Start immediately", enter the number of hours to delay, and press the "Start" button.

Novice users of the cycling program may wish to compare the GUI to the following examples:

TROUBLESHOOTING INFORMATION: Cycling logs can be found on sharcii in ~sharc/recycle/logs/. Temperatures can be read directly from the Neocera front panel near the instrument; the left temperature is a "charcoal pump" temperature (3He or 4He) which normally reads 6.7 K except when a cycle is taking place. (For the pump temperature, there is a scale factor difference between the front panel and the cycle logs.) The right temperature is the detector temperature which normally reads 0.36 K in operation.

In the event that a cycle hangs in a state in which continuous heat is applied to the refrigerator, the heating can be stopped by pressing the Monitor button on the Neocera front panel.

Starting the DSOS Server

As stated in the DSOS instructions, the DSOS daemon (dsosd) and server (rdsoss) should be started on tpick prior to starting the IRC server and client.

Powering Up the Electronics

The amplifier electronics runs from a battery and is powered up manually. First, make sure that the big power switch on the analog electronics control chassis is "OFF". (The battery cable should be plugged in already.)  At the alidade power controller, switch the battery from "charge" mode to "operate" mode. On the analog electronics control chassis, you should see one yellow light and two green. If not, switch the power controller to "charge" and call for staff assistance.

Assuming you see one yellow and two green lights, wait approximately 15 seconds for the cold electronics to pre-heat. Then switch the big power switch to "ON". You should now see one yellow, two green, and two red lights. If not, move the switch to "OFF", switch the power controller to "charge", and call for staff assistance.

The remaining configuration steps are done with software control.

More detailed information concerning the SHARC II electronics is available.

Programming the Instrument Electronics

Log into sharcii: "% ssh sharc@sharcii". Execute "sharcii% cd ~/fpga" and "sharcii% go". This configures the amplifier electronics and multiplexer, defaulting to high-gain mode. To switch to low gain (to observe the Moon, for example), execute "sharcii% all_gain lo". To switch back to high gain, execute "sharcii% all_gain hi".

The instrument will power up with an incorrect hardware level, which results in increased current draw from the batteries. See the hardware leveling instructions to correct this (after starting up IRC).

Checking the Optics

SHARC II has an internal filter wheel operated by a black knob on the top of the cryostat. The following table shows the conversion of counter value to filter setting:
000 = future 350 micron pupil imager (no longer blank; don't use this)
250 = 350 microns
500 = 450 microns
750 = 850 microns
The filter wheel is adjusted BY HAND ONLY. The IRC software does NOT change the setting.

Although the default bias voltage setting (255) will work acceptably at 850 microns, a setting of 100 will give greater responsivity.  To change the bias setting, log into sharcii -- "% ssh sharc@sharcii" -- and execute "sharcii% cd ~/fpga" and "sharcii% all_bias 100 100".  To restore the bias voltage to the optimal value for 350 microns and 450 microns, use "all_bias 255 255".  Be sure to use the same bias setting for your target and flux calibrator.

After checking the filter setting, check the SHARC II optics to insure nothing is blocking the beam and that the tertiary mirror directing the beam to SHARC II is in place. Tie up loose wires which could get in the beam. Assuming the weather is dry, you should also be opening the shutter (looking east if the Sun is still up) to let the telescope cool.

Resetting the DSP Code

At the beginning of each night, the DSP code should be reloaded into the DSP hardware in the sharcii computer. To do this, log into sharcii as user sharc. Execute "sharcii% cd ~/DSP".  Next, execute "sharcii% sharcDSP", which takes tens of seconds to execute. If this doesn't work after a few tries, try "sharcii% conf4290" first.  If this still doesn't work after several tries, as a last resort try using a small tool to press the reset button ("Rst") half-way down the Pentek board in sharcii, then execute "conf4290" and "sharcDSP" again.  Please use precaution to prevent static discharge onto the Pentek board.

Optional, but recommended:  Status of the DSP can be monitored by the "s2" program in the same directory.  To view frame status, execute "sharcii% s2" in the DSP directory, then at the prompt "SHARC2>status". You may wish to compare to an example status screen.  Note the values of 12 and 36 in the sample size field, and note the incrementing of sequence numbers and universal time between the status commands.  If you see no 12's, then the multiplexer is probably off or not configured (which is done by "fpga/go").  If the time is not updating, try the reset procedure again.  Execute "SHARC2>quit" to leave the DSP monitor, or type "SHARC2>help" to see other menu options.

NOTE CONCERNING DSP TIMING: The DSP code reads the time from the WWV/IRIG-B board at the time of executing "sharcDSP".  After that, the code counts A/D samples to keep track of the time. If for some reason samples are missed (e.g., due to pulling out the fiber optic cables), then the DSP time will be incorrect, and bolometer samples will not be timestamped correctly. If you suspect a timing problem (which usually results in multiple images of a source or a missing source), then reloading the DSP code is one possible solution.

Starting the Acquisition Software (IRC)

IRC requires a server running on sharcii and a graphical client running on another machine. To start the server, log onto sharcii (user sharc) and execute "sharcii% cd irc" and "sharcii% sharcServer". To start the client, log onto the appropriate machine (currently sharc@kilauea, sharc@champinux, or guest@champinux) and execute "kilauea% cd irc" and "kilauea% sharcClient".  (The preferred machine for running the IRC client is currently kilauea, although champinux can also be used for the client.)  See the readme file on both machines for alternate versions of the client and server.  We sometimes use the beta versions "sharcServerTest" and "sharcClientTest", if there is reason to believe they are better.

The IP number of champinux is

The sharcServer shows up as a j2re1.4.1 process on sharcii, visible by "sharcii% ps -ef | grep j2". The sharcClient shows up as many j2sdk1.4.2 processes on kilauea, visible by "kilauea% ps -ef | grep j2".


Procede to the observing section.

Special Considerations for the First Night

Window Ice

If the instrument was cooled down in humid conditions, there may be ice or water on the white, translucent window on the underside of the instrument. Use a Kimwipe to gently wipe away ice or water. In the case of thick ice, you may try isopropyl alcohol on the Kimwipe to remove the ice faster. Please DO NOT use stronger solvents or a heat gun.

Normally, a small fan blows onto the window to prevent moisture accumulation. If this is not installed, ask the staff for help.

Warm Detector

Following the intial cooldown, it takes many days for the detector to passively cool to 4 K.  (This will be noticed as a "detector" temperature between approx. 10 K and 100 K on the Neocera front panel or the cycling gclient display.)  You will want to actively cool the detector instead. One approach is to execute a "3He only" cycle cycle to force detector cooling. After the "3He only" cycle, wait for the detector temperature to drop below approx. 8 K before proceeding. If one "3He only" cycle was not sufficient to drop the temperature below 8 K, try another one. Once the temperature is below 8 K, follow with a standard cycle.

In August 2003 and January 2004, the first standard cycles ran succesfully, but the detector took several hours to cool below 0.4 K, and the liquid 3He ran out prematurely.  This problem disappeared by the next day.  The best explanation is that there is a component which is poorly heat sunk and causes a transient load on the 3He system.  Performing the first 3He cycle as much as a day in advance is a suitable way to address this problem.

Setting Up the champinux and sharcii Drives (for CSO/Caltech staff) -- THIS NO LONGER APPLIES, AS OF 2009; BACKUP DRIVE ON LAPAKAHI/KILAUEA

The transfer (removable backup) drive can be "hot swapped" into champinux. Insert the drive, use the key to lock it in place and turn it on, then log in to champinux. First execute "champinux% sh ~sharc/bin/", then "champinux% mount /home/transfer".  Data reduction accesses the transfer drive.

To mount the sharcii drives, continue the root login to champinux and execute "# mount /home/data0" and "# mount /home/sparc". THIS PARTICULAR PROCEDURE IS NO LONGER RECOMMENDED, BECAUSE IT FACILITATES ACCESSING THE SHARCII DRIVES DURING DATA ACQUISITION, WHICH CAN CAUSE DATA DROPOUTS.

To mount the transfer drive on sharcii -- which is necessary for automatic data backup -- log in as root to sharcii, then execute "# mount /home/transfer". It may be necessary to execute the following sequence as root on champinux: "# cd /etc/init.d", "# ./nfs stop", and "# ./nfs start".

Configuring the Data Backup Directory (for CSO/Caltech staff)

Edit the following file to change the directory for data copying:
The parameter of interest is "sharc.copy_directory".

Observing Procedure

Telescope/UIP setup

The default UIP configuration for SHARC II can be obtained with the following commands:
UIP> cat user:[SHARC]
The second line is needed only if you do not log in via the SHARC account. Please see the CALIBRATION web page for more information on this IMPORTANT catalog command. Note that "INST SHARC2" wipes out any focus refinements ("FOCUS /OFFSET={offset}") and pointing refinements ("FAZO" and "FZAO") that you may have made. A typical SHARC II antenna monitor screen looks as follows:
 DATE   SEP/01/2004  DOY 245 MJD  53249  AAZ      +216.8346  AZA       +59.7708
UTC 07:49:57.6 DAT 32 DUT -0.450 RAZ +216.8359 RZA +59.7702
LST 20:11:18.6 EAZ(") -4.8 EZA(") +1.9

GLO +0.0 GBO +0.0 RAEP 17:45:40.04 DECEP -29:00:28.1
AZO +0.0 ZAO +0.0 RA 17:45:57.62 DEC -29:00:43.8
FAZO -124.6 FZAO +47.3 HA +02:25:20.97 P ANGLE +40.0077

CHOP AZO +0.0 AIR MASS 2.006 BAROM 627.6
LCAS APA +90.1993 RCAS APA -132.1957 TEMP +3.2
LCAS RPA +0.0000 RCAS RPA +0.0000 SWP AZO -9.6 HUMID 12.2
LCAS OFF +0.0000 RCAS OFF +0.0000 SWP ZAO -56.2 TAU 225 0.079
X POS -10.80 THETA +90.00 WIND PK 0.0
Y POS +11.87 FOCUS -4.03 FOC MODE CONSTANT AT 00:00:00.0
Some possible deviations and their solution:

Observing Modes

Acquire the target with "UIP> OBSERVE ..." or "UIP> PLANET ...".

For SHARC II, it is always necessary to have a scan pattern (SWEEP or BOX_SCAN) running in order to collect useful data.  Some suggested scan patterns are available.

Preparing and Starting and Integration

One of the important actions you will need to do with IRC is to level the hardware. If the bolometer signals are greater than 1 V = 1000 mV in absolute value, it is recommended to run "Level Hardware", since the A/D converters saturate at +-3.5 V = +-3500 mV.  To examine the mean bolometer signal, look at the raw bolometer array readout, which has units of mV, or the bolometer frame medians display (graphical display and bottom line of table), which has units of V.  Re-leveling is necessary when the gain is changed (high to low, or low to high), or when the optical loading is changed. While tracking a source, it is typical to hardware level every 20-60 minutes.  Hardware leveling should not be done during an integration.

The integration parameters are set with the SHARC controls window. The most common integration type is "Single, Preset Time".  As of this writing, the true integration time is approximately 20 seconds less than the requested time.   Before starting the integration, you may wish to set some of the documentary features, such as Filter, Observer, Project, and DSOS status.  Begin an integration with "Start".  Selecting "Copy Observation Files" performs automatic backup to kilauea or lapakahi (currently /halfT).  An integration is complete when the integration file name appears in the sharcServer monitor window, and, if automatic backup is selected, when "Copy complete" appears.

Pointing and Focusing

A quick-look reduction facility "sharcpoint" is available to evaluate the image and facilitate pointing and focusing.  To initiate sharcpoint, log into kilauea as sharc:  "kilauea% ssh sharc@kilauea".  From the home directory, execute "kilauea% sharcpoint #####", where ##### is the five-digit scan number to reduce.  At the completion of the data reduction, ds9 is launched to show the resultant map, and the results of an attempted Gaussian fit to the center of the image are shown.  If the signal-to-noise is sufficient, you will receive useful suggested pointing offsets for the telescope (FAZO and FZAO, which can be adjusted using the UIP commands of the same names).  The focus can be adjusted with "UIP> FOCUS/OFFSET=#.##"; you may wish to review a focus record from November 2002.  In unusual circumstances, it may also be necessary to adjust "UIP> Y_POS/OFFSET=#.##" and "UIP> X_POS ##.##". Note that after an XPOS command, the focus mode will have to be changed to constant again: "UIP> FOCUS /CONSTANT".

If the source is well off the center of the map from "sharcpoint" (i.e., the pointing was bad), the suggested pointing offsets may be incorrect.  (Look at the fit position, amplitude, and FWHM to see if they are sensible.)  You can bring the source back into the center by trial and error:

To move the source to the right, decrease FAZO.
To move the source to the left, increase FAZO.
To move the source up, increase FZAO.
To move the source down, decrease FZAO.
If sharcpoint does not seem to work, check the code (it is a shell script), and carefully make any necessary changes.

DSOS Operation

If you haven't done so already, set up the DSOS according to the instructions on the web page.

Additional Notes on IRC (for CSO/Caltech staff)

The server and client logs are available in:
/home/sharcii/sharc/irc/v*/app/SHARC_IRC/IrcLogs/     (server)
irc/ircData/sharc/IrcLogs/ (client)
Many configuration settings are defined in:
The "work space" files which save the state of IRC can be found in:

Data Reduction

Data reduction at the observatory is currently performed on kilauea.



Pointing and Calibration References

The following planets and moons are available in the UIP (through 2003 at least) and can be used for pointing and flux calibration: MERCURY, VENUS, MARS, JUPITER, CALLISTO, GANYMEDE, EUROPA, IO, SATURN, TITAN, NEPTUNE, and URANUS. Caveats: Mercury is usually difficult to observe, and little is known about it as a flux calibrator. Venus, Mars at opposition, Jupiter, and Saturn are not ideal for calibrating point sources due to their extent. Europa and Io are difficult to observe due to their proximity to Jupiter; Titan, Ganymede, and Callisto sometimes have the same problem.

Positions of the planets and their satellites can be located with the JPL Horizons System. For observer location, select "Mauna Kea" (code 568).

"Current" positions of the planets (occasionally updated):
Date:  2008 Sep 15

    Source        RA     Dec      Ang. Diam. (")
--------------   -----  ------    --------------
(Sun             11:32    3:01)
Mercury          13:05  -10:11         7.49
Venus            13:10   -6:49        11.34
Mars             13:04   -6:23         3.82
Jupiter          18:54  -23:10        41.76
Callisto                               1.40
Ganymede                               1.54
Saturn           11:01    8:10        16.09
Titan                                  0.xx
Uranus           23:26   -4:31         3.69
Neptune          21:38  -14:32         2.34

Some asteroids can be used for pointing and, if good models are available, flux calibration. The following asteroids are available in the UIP, at least through 2005: AMPHITRITE (29), BAMBERGA (324), CERES (1), DAVIDA (511), DESIDERATA (344), EGERIA (13), EUNOMIA (15), EUROPA52 (52), HERCULINA (532), HYGIEA (10), INTERAMNIA (704), JUNO (3), PALLAS (2), PROKNE (194), PSYCHE (16), VESTA (4). Some asteroids are only useful when near closest approach to the Earth. See Redman et al. (1995, AJ 109:2869) for dates of opposition and predicted 1.1 mm fluxes for major asteroids.

"Current" positions of the asteroids (occasionally updated):
Date:  2008 Sep 15
    Source        RA     Dec      Ang. Diam. (")    Est. Flux (Jy)
--------------   -----  ------    --------------    --------------
Amphitrite 29   10:00   15:19        0.087               
Bamberga 324    16:35  -31:01        0.128
Ceres 1          8:59   22:14        0.404              *
Davida 511       8:26   19:10        0.142               
Desiderata 344  14:43  -22:03        0.080
Egeria 13        8:34   30:59        0.100               
Eunomia 15      10:56   -0:28        0.090               
Europa 52        0:07   -8:00        0.188              *              
Hebe 6          12:05    7:27        0.066
Herculina 532    4:25    4:38        0.115
Hygiea 10        5:02   25:36        0.172              *
Interamnia 704   9:37    5:30        0.107               
Iris 7          14:05  -15:12        0.076
Juno 3          17:10  -10:08        0.106
Pallas 2         5:06  -11:36        0.347              *
Prokne 194       6:48    5:52        0.079               
Psyche 16 M     15:27  -15:58        0.100
Vesta 4          3:01    6:45        0.417              *

The brightest blazars are useful for pointing, but only useful for flux calibration if they have been recalibrated vs. a standard within the last few days. Here are some useful links for identifying bright blazars:
SMA compilation -- try this first
BIMA list
IRAM list

Evolved stars, (proto)stellar condensations, and some galaxies can be used for flux calibration. I recommend the following: CRL618, OH231.8, IRC+10216, CRL2688, GL490, HL Tau, IRAS16293-2422, G34.3, and ARP 220. Most of these sources can be found in Sandell, MNRAS 271:75 (1994), although note the larger 350 micron beam in that reference.

SHARC II measurements of 350 micron secondary calibrators

Shutting Down for the Night

Following is a shutdown checklist:

Cryogen Servicing (for CSO/Caltech staff)

Normally, the CSO staff handles cryogen servicing. It's a good idea to keep in contact with them about the schedule and to plan for weekend servicing. Please service cryogens yourself only if you are trained by the staff and wear the proper safety equipment.

Liquid Nitrogen

The liquid nitrogen reservoir needs to be topped off every day. I prefer to fill it early in the day, or to fill it with the cryostat moderately tipped. Otherwise, a fair bit may spill on the electronics and floor if you observe a high source early in the night. The nitrogen fill port is in the middle of the cryostat, and the reservoir capacity is 7 liters.

Liquid Helium

SHARC II needs liquid helium every other day during active observing. The helium level should be at least 3 inches (measured with the cryostat vertical) at mid-day before 3He fridge recycling and observing. The helium fill port is the one closest to the edge of the cryostat. The helium is easiest to fill at a zenith angle near 35 degrees. The reservoir capacity is 11 liters.

In the case of a suspected ice plug in the SHARC II helium reservoir, CALL STAFF IMMEDIATELY FOR ASSISTANCE. Ice plugs can eventually result in a cryostat explosion. The distance from the bottom of the reservoir to the exit hole is 20.5 inches, and the reservoir is 7 inches deep.

Please replace the downward-pointed black rubber hose assembly after filling the liquid helium. This helps prevent ice plugs.


Source suddenly disappeared or signals suddenly went haywire

Check the 5 LED's on the analog electronics control chassis. If you see something less than 1 yellow, 2 green, and 2 red, then a fuse probably blew. Turn the big power switch to "OFF", switch the power controller to "charge", and call the staff for help.  NEED A LINK TO PHOTOS HERE.

Bolometer signals aren't updating on the IRC Sharc Client display

First, check that signals are passing through the DSP.  This procedure is explained elsewhere.  If the "s2" program shows incrementing sequence numbers, then the DSP and A/D are OK, and the problem is related to the IRC software.  If the sequence numbers are NOT incrementing, the problem is related to the DSP and A/D.

If the sequence numbers are not incrementing, first try restarting the DSP code.  If that doesn't help, the next suspect is the A/D hardware.  Check the Topward A/D power supply to insure that the voltages are set at -7.0 and +6.0 volts, and that the currents are 2.5+-0.4 amps.  Check that the orange fiber optic cables are connected between the A/D box and sharcii computer -- 8 on each side, leaving 2 spare cables out of the bundle of 10.  Finally, press the small, red "reset" button on the A/D box.  If none of these work, then the problem probably needs help from the CSO staff.  If you have managed to get things working again by fiddling with the hardware, restart the DSP code again to make sure the time synchronization is correct.



Miscellaneous Notes

Observing Log Sheets

Here is a paper log sheet template provided by Colin Borys:

Estimating Weather Conditions with SHARC II

If the default bias (255 digital setting = 1 V) is used in combination with the 350 micron filter, and assuming a hardware level has been successfully run since the IRC client was started, then a real-time "SHARC Tau" can be derived from the bolometer median window.  The Bolometer Power Median value (top line of the table) can be converted to atmospheric opacity with the following graph:
bolometer V vs. tau, Jan. 2003
The average bolometer voltage is related to the radiation load on the bolometers, which is related to the atmospheric opacity.

This file last updated on