apo-wiki

The ARC 3.5m telescope is controlled using the TUI software, which is software written in Python using a Tk graphical interface.

The telescope can also be controlled with command line commands and scripts, which can be run through TUI. TUI includes some built-in scripts, but users can also provide their own.

Steps 1 and 2 are advance preparation before your observing run.

1) Get raw ephemerides from JPL Horizons. These are more accurate than our in-house ephemeris program can provide. Go to the JPL Horizons web interface and get these data for each target of interest: timestamp, astrometric RA and Dec (airless, in decimal degrees, J2000.0), RA and DEC rates. You can get additional info if you want, but those are the ones that we really need. Example settings for JPL ephemeris:

• Ephemeris Type [change] : OBSERVER
• Target Body [change] : Saturn [699]
• Observer Location [change] : Apache Point [705] ( 254°10'45.9E, 32°46'49.8N, 2891.2 m )
• Time Span [change] : Start=2008-04-24, Stop=2008-04-25, Step=1 h
• Table Settings [change] : QUANTITIES=1,3; time digits=SECONDS; angle format=DEG; refraction model=REFRACTED; skip daylight=YES; extra precision=YES
• Display/Output [change] : default (formatted HTML)

Then Generate Ephemeris, which outputs something like:

Date(UT)HR:MN:SS R.A._(J2000.0)_DEC. dRA*cosD d(DEC)/dt
$$SOE
..... Daylight Cut-off Requested .....<
2008-Apr-24 02:00:00 C 154.3057447 12.5867617 -2.52 0.61
2008-Apr-24 03:00:00 A 154.3050264 12.5869290 -2.52 0.60
2008-Apr-24 04:00:00 154.3043103 12.5870917 -2.51 0.58
...
..... Daylight Cut-off Requested .....<
$$EOE

2) Write a script to convert the ephemeris numbers into the correct command syntax. The BIG, IMPORTANT difference is that JPL gives dRA and dDec in arcsec/hour, whereas our software wants deg/sec. That means YOU HAVE TO DIVIDE THE SPEED BY 12960000 or you will get no data and track the telescope into the ground. It's useful to make a file with multiple commands, each one marked by a timestamp so as to pick a command depending on what time one actually issues the slew command. Here's some example slew commands:

tcc track 136.72666,  17.43467, -0.00000146,  0.00000046 Fk5=2000.0 /Rotangle=0.0 /Rottype=Object
tcc track 136.72666,  17.43467, -0.00000146,  0.00000046 Fk5=2000.0 /Rotangle=108.5 /Rottype=Object
tcc track 136.72478,  17.43524, -0.00000146,  0.00000046 Fk5=2000.0 /Rotangle=0.0 /Rottype=Object

where the first four numbers are: RA(decimal degrees), Dec (decimal degrees), dRA (degrees/sec), dDec (degrees/sec), and rotation can be specified with the /Rotation=rotation, where rotation should be calculated as 90 - PA, where PA is measured N thru E.

From the JPL ephemeris example above, one line would convert to:

2008-Apr-24 04:00:00 154.3043103 12.5870917 -2.51 0.58
tcc track 154.3043103, 12.5870917, -0.000000194,  0.000000047 Fk5=2000.0 /Rotangle=0.0 /Rottype=Object

3) Slew to the planet with desired rate and rotation. Notify the observing specialist that you are about to issue a non-sidereal tracking rate to the telescope and get their permission to do so before doing the following. If the slew rate is wrong it can cause the telescope to slew too fast, and the observing specialist can stop the telescope before it goes out of control. The telescope will not automatically check your numbers! If in doubt, ask the observing specialist to double check your numbers and they can issue the tcc track command for you as well.

Paste your pre-prepared commands into the TUI Log window. When in doubt about anything (if your offsets get tangled up, for example), reslew. This is the motivation for making all those pre-prepared commands. They will come in useful.

4) Take images often to see what's going on. This is just generally good policy. It should be safe to run the guider even if you have a non-stellar tracking rate.

Rough IDL exposure time calculator

Other exposure time calculators for the 3.5m?

[root@astronomy pages]# vi start.txt [root@astronomy pages]# vi 3.5m.txt [root@astronomy pages]# vi tui.txt

Some (probably obsolete!) information about building TUI from source on a Linux machine.

TUI has strict python dependencies. Many linux users have problems displaying guider or slitviewer images because they have the wrong version of one or more packages. Even if TUI is built correctly, running it from the wrong directory might result in a different package being loaded. Since individual linux and python setups can vary widely, the same solution might not work for everyone. But here are some solutions that have worked for some people.

Download a copy of TUI pre-built for linux as of December 2017 at

http://www.apo.nmsu.edu/35m_operations/TUI-images/files/tui-centOS.tar.xz

Future TUI versions will be linked from the TUI-images web page so be sure to check there also.

1. install miniconda2 using the usual installer
2. in miniconda2/bin do ./conda install python=2.7.12 to revert to python 2.7.12
3. create a file “pinned” in miniconda2/conda-meta with the line python ==2.7.12
4. in miniconda2/bin do “./conda install numpy=1.8.2” to install numpy 1.8.2
5. add line “numpy ==1.8.2” to “pinned”
6. in miniconda2/bin do “./conda install matplotlib”
7. in miniconda2/bin do “./conda install Pillow”
8. in miniconda2/bin do “./conda install astropy”
9. in miniconda2/bin do “./pip install pygame”
10. in miniconda2/bin do “./pip install RO==3.6.9”