How to Take Guided Observations

Eric Deutsch
December 14, 1995

The document is designed as a short guide to successfully taking guided observations on the 3.5m. Please send comments, corrections, additions, etc. to me for incorporation into this document. If you are reading a paper copy of this document, I suggest that the date at the top be compared with the electronic copy at http://www.astro.washington.edu/deutsch/apoinfo/guider/howtoguide.html to insure that the latest version is being used.

All commands quoted below (in courier font) should be entered either in an MC window or a tcctalk window (see tcctalk manual by Bernadette Rodgers).

Preparations Before the Observing Run

Since the current state of the 512 Guider (no binning available, moderate read noise, moderate dark noise, high S/N requirement by software) requires at least 16th (preferably 15th) magnitude stars, suitable guide stars are not likely to fall in the Guider FOV except perhaps near the Galactic plane. This means that you will need to search for suitable guide stars near your object. (There are a reports of successful guiding on 17th magnitude objects with 60 second exposures, but I wouldn't plan to rely on this.) Remember also that the brighter your guide star, the higher your chance of successful guiding when light clouds or bright moon are present.

Go to http://www.astro.washington.edu/deutsch/apoinfo/guider/gdrgacform.html and follow the directions there to make Guider Acquisition Charts (GACs) for your targets.
Determine an appropriate rotator angle to place a suitable guide star in the Guider FOV (15th - 9th magnitude preferred.)
Note that when using DIS, if the rotator angle is non-zero anyway, it might be prudent to aim the DIS slit toward the horizon as much as possible to minimize differential refraction problems. Here's a PostScript plot of horizon direction as a function of hour angle and declination.
Also, although the annulus is pretty narrow, it is possible to acquire stars outside the annulus since the DIS slit is long, and, when imaging, it is not always necessary to have the target in the center of the chip. To acquire stars outside the annulus, you would first rotate to the appropriate angle, and then make an instrument offset in the X direction (think about the direction you would slide the annulus to make the star appear in it). For DIS, you want to make a positive offset in X for objects inside the annulus, and a negative X offset of objects outside the annulus. You can use the annulus thickness (80 arcseconds) to judge the magnitude of the offset.

Preparations as the Observing Run Begins

It is necessary to prepare the Guider instrument before guiding can be used. The Observing Specialist can assist with this if you need assistance.

Guider should be checked out that it is responding properly.

       tcc gcam init
       tcc gcam status
       tcc gmech status

Filter wheel should be set to the red filter (5) although be sure to set is back to redND (1) before doing FK5 offsets or something. If you are operating a an airmass near 1.0, you might try running unfiltered (0) for more throughput, although note that the focus is slightly different for filtered and unfiltered images!
```
       tcc gmech filter 5   
```
Guide camera should be focussed (currently best done by the Observing Specialist). Note that Remark 2.34b is capable of crontrolling the Guider in a limited fashion; you can take exposures like any other instrument, but you can't change focus or filters.
```
       (e.g. tcc gmech piston 68000)   
```
Ask the Observing Specialist to check/set the guiding exposure time in the instrument block. It should be around 10-60 seconds. This can be changed while guiding, too.
To monitor guider performance, start Bernadette Rodgers' tcctalk software and use the 'guideplot' task (see the quicky startup guide , which, however, is no substitute for reading the tcctalk manual ).

Starting a Guided Observation

Slew to the target with the pre-determined rotator angle which places a guide star in the Guider FOV (offset if necessary).
Only if pointing is bad, take a DIS acquisition image and get the star near the desired location on the chip (if pointing is good, skip this step!!) (really!)

Engage guiding.

       tcc gctrack 0,0 phys
       tcc guide on   
       (or just 'guideon' in tcctalk)

Monitor guider output to ensure proper acquistion (e.g., with ' guideplot ') and stability. Note that at present there is always an "initial jump" when guiding starts. Wait until you're past this jump and guiding is stable, then...
Take an acquisition image (with DIS or slit viewer) and make an instrument offset to get the target centered on slit if necessary; if the offset is too big, guiding could fail (10 arcseconds is safe, but don't offset the guide star off the chip!).
Start science observations.
Stop guiding.

       tcc guide off
       (or just 'guideoff' in tcctalk)

Questions and Notes

What if ' guideplot ' shows that measured centroids are consistently away for target coordinates be several pixels (i.e. DX, DY aren't bouncing about 0.0)? or what if ' guideplot ' shows that measured FHWM are very high, i.e. over 5 arcsec (unless it is known to be that bad)?
Stop guiding immediately; the guider is guiding on noise or something else is wrong, and the telescope is quickly drifting away from where it should be. You may need to set the filter to 1 and do a FK5 "Recalibrate" and set the filter back to 5 and try again, perhaps with a longer exposure time. If you are very near zenith, consistently off centroids could mean that the guider is working okay but having to make large corrections because tracking is poor near zenith; in this case you're probably best off moving away from zenith.
You can do offsets while guiding (in fact this is recommended since the guider will insure that you get the exact offset you requested) but don't offset your guide star out of the field! If you are monitoring guider performance, you can see the X,Y position of the guide star. A positive X instrument offset will cause the guide star to move positive Y on the guider chip. Likewise, a positive Y instrument offset will case the guide star to move positive X on the guider chip. Remember that guider scale is 0.14 arcseconds per pixel.
What if you just can't find a star that should be in the guider field?
There are a few options. It may be that pointing is bad, and an FK5 "Recalibrate" maneuver (available through Remark) will help (but be sure to set the guider filter to 1 first!! and then back to 5 or 0 before you try to guide!!). You can also try to use the larger-field DIS to locate your guide star; to do this: First make an instrument offset X,Y=(-780,-4) so the DIS is looking at the guiding region. Now select an appropriate guide star and do an offset so that the star is within 30 arcseconds of 435,435 on the DIS red chip. Now, make the reverse instrument offset X,Y=(+780,+4) to bring that star into the guider FOV.
What if the guider informs me that guiding failed because the star fell on a bad pixel?
This is unfortunate, particularly since there are no bad pixels on the chip (but someone has created a non-zero bad pixel map). I recommend a small offset or rotatation to move that star to a different place on the chip.
What if the Guider fails to lock onto a star?
Bad news... Any of a number of things could have gone wrong. First make sure the filter is really set to 5 (nothing worse than trying to find a faint star through a neutral density filter). You can try a longer exposure time (e.g. try 30s instead of 10s). Those sneaky clouds are always a possibility; you might check the sky. Then it is probably best to ask the Observing Specialist to snap an image and see what's there. If the star is too close to an edge or a bad pixel, try an offset to get it into a better position. If there's nothing there, try another star (at a different rotator angle). Make sure that the main instrument is really pointing where you intend it to be. If you can grab an image that the Observing Specialist has taken and identify the Guider FOV, you might be able to offset to the star. If all this fails, stop wasting time and take unguided exposures and lobby your local representative to order the purchase of good equipment and better software.