Reynolds Observatory 0.30 meter
Operating Procedures and Best Practices
Clarkson University
Potsdam, NY 13617
315-268-6677

Abstract: This document is meant as a reference for infrequent, overly tired, or forgetful observers who want a reference or who need reassurance that they've taken care of the equipment.

1  Note

Contents

• 1  Note
• 2  Startup Procedures
• 3  Observing Procedures
  • 3.1  Solar Observing
  • 3.2  Imaging
    • 3.2.1  Focusing
    • 3.2.2  Position Limits
    • 3.2.3  Autoguiding
    • 3.2.4  Calibration
    • 3.2.5  Flat Field Procedure
  • 3.3  Time-Delay Integration (TDI), aka Drift Scan
• 4  Computing Facilities
• 5  Analysis
• 6  Shutdown Procedures
  • 6.1  Imperatives
• A  Appendix
  • A.1  Location and Directions
  • A.2  Instrument Specifications
  • A.3  Site Limits
  • A.4  Contact Information

2  Startup Procedures

• The key to the dome is stored in ROARS ¹; there is a key to ROARS on a bolt above the door to the dome.
  
  There is a crowbar in ROARS which may be useful if the door to the dome freezes shut in the spring.
  
  
• Sign in to the logbook.
  
  
• Call the cops: (9-265-2121): Heightened security, don't you know??
  
  
• Open the dome slit by moving the toggle switch to the right. It will automatically stop when it is fully opened.
  
  
• Remove all lenscaps.
  
  
• Power on the scope using the toggle switch on the far left of its control panel.
  
  The GPS subsystem is sometimes unable to sync to a satellite, apparently because of the metal dome. If alignment fails, or if you cancel (by pressing “Mode”), then you can still use the previously found GPS information; you will, however have to enter the current time.
  
  Meade's alignment procedure is apparently too complicated for us Clarkson folk, so we do our own, by simply syncing on an object.
  
  The Autostar's “Mode” button escapes from the current menu and enters the menu one level up. “Enter” selects the current item.
  
  When pointing at an object, hold then press “Enter” to update the telescope's coordinates to those of the current object (“sync”).
  
  
• Align the dome slit with the telescope. There is exactly one power outlet which powers both the dome slit control and the dome rotation. Unplug the dome slit and plug in the dome rotator. Wrap the power cord for the dome slit around twice such that it doesn't catch on anything while the dome is rotating.
  
  
• Power on the PC. Launch The Sky and link to the telescope with “Telescope - Link Establish”.

3  Observing Procedures

• To point to an object using the PC:
  • Click the object on the sky map, then click the green telescope icon on the object's information window; or
  • Use “Edit - Find” to query for an object by name; or
  • Use “View - Move To” to enter coordinates.
  
  
  
• To view low altitude objects, it is necessary to close the dome slit, then latch the bottom slit the main slit and reopen the slit. To observe high altitude objects (near the zenith), it is then necessary to close the slit, unhitch the bottom slit, and reopen the slit.

3.1  Solar Observing

• Do not point anywhere near the sun (within 90^∘) unless the filter is attached. Ideally, the dome slit is closed until it is so. Before removing the filter, either close the dome, rotate it 90^∘ from the Sun, or slew the telescope 90^∘ from the Sun. Ensure that there is always a shadow on the front lens of the telescope when the filter is removed.

3.2  Imaging

• The f/0.63 focal reducer is strongly recommended for imaging. Focusing and guiding are much harder at the full focal length. The focal reducer screws into the back of the 'scope. Attach the electric focusser to the focal reducer, being careful to ensure that the three screws are actually in the ridge on the ring.
  
  If you add or remove the focal reducer, it will be necessary to also move a protective ring. With the focal reducer, the setup is: telescope - focal reducer - ring - electric focusser. Without the reducer, the setup is: telescope - ring - electric focusser.
  
  
• Counterweights may need to be added or repositioned when the CCD is in use.
  
  
• The CCD is approximately aligned when it is square with the mount and its wires are on the same side as the Telrad.
  
  
• If images show a strong gradient and no stars are visible, it is possible that the filters are not aligned. Sometimes the filter wheel fails to put a filter where it should be, but instead positions the wheel between two filters. The solution to this is to fuss around with it until it happens to work. You may manually reposition the filter wheel while the camera is unplugged from the power source; this may help solve the problem. Also, the filter wheel is supposedly reinitialized when the camera is repowered. This does not necessarily solve the problem.
  
  
• A homemade eyepiece parfocal with the CCD is stored in the CCD box.
  
  
• TheSky5 (not 6) with CCDSoft5 is recommended; the two packages work together. The 'scope's position is written to the FITS files, and the autoguider knows the declination coordinate.
  
  

  3.2.1  Focusing

  • It is recommended to lock the mirror while focusing the CCD, especially at low altitudes. The lock knob doesn't need to be tight; just twist it enough to show *some* tension. Be sure to unlock it later.
    
    
  • Moving TheSky focus “Out” has the same optical effect as rotating the focuser clockwise.

  3.2.2  Position Limits

  • When taking CCD observations with the focal reducer and the electric focusser (the usual configuration), the bottom edge of the camera will run into the base of the telescope at high declinations. The upper limit is δ = +74 ^∘.

  3.2.3  Autoguiding

  • Connect the autoguide cable (RJ-11 phone cable) between the telescope electronics and the autoguider.
    
    
  • Ensure that the autoguider uses dark frame calibration; it suffers from much worse dark current noise than the primary CCD. What might look like a blank frame can become a veritable starfield after dark subtraction.
    
    
  • The autoguider software requires a matrix of coefficients to properly track a star. After one successful calibration of the autoguider, I obtained the following coefficients (which are useful at least in CCDSoft, and probably any other guider software):
    
    

    Axis	Motion (pixels per second)	Angle (degrees)
    +X	5.7492	181
    −X	5.1402	1.84
    +Y	3.0735	269
    −Y	3.2147	92.1

    
    
    
    The motion may need to be multiplied by a factor related to cosδ, where δ is the declination. The above data was obtained with δ=24.299. When using software other than CCDSoft, the above angle may be measured from a different reference, causing, for example, “181” to become 179 or 91.

  3.2.4  Calibration

  • I have a set of calibration frames taken at 0^∘C stored on the dome computer. We should be able to cool the camera to that temperature any day of the year, and if you can't take your own calibration data, these should suffice.
    
    
  • The camera appears to be largely immune to electronic noise; don't worry about slewing the telescope or the dome while taking calibration frames. However, dark frames appear to vary, not only with temperature, but also with the power applied to the thermoelectric cooler. Supposedly it will shut itself down during readout, but it appears to affect the image acquisition as well.
    
    The moral is to take multiple darks and statistically combine them.
    
    
  • Flat fields are highly dependent on the optical path. As such, flat fielding is largely impossible because the camera is not permanently mounted. Twilight flats are the only possibility, and calibration would be meaningless if the camera were removed between the taking the flats and taking data.
    
    If you intend to image beginning at dusk, or until dawn, then flat fielding is possible and appropriate. The uniform illumination spot is supposed to be opposite the Sun in azimuth and at 70degrees altitude.
    
    Flat fields must have more than 10000 and less than 25000 counts/pixel to be useful. Any more and the CCD has nonlinear response. Any less and the relative error on the flatfield is greater than 1%: the whole point of flatfields is to get accuracy better than 1% .
    
    It is important to realize that the Max counts statistic is not the important number when taking flats. Ideally, the Max never exceeds 25000 (this is tricky). More importantly, though, is that the images “typical² ” value is sufficiently high to be useful. Few than 10000 counts per pixel on a frame is marginally useful.
    
    When the sun is high, the typical pixel has only slightly fewer counts than the max pixel. But, as it gets lower, the different between the max pixel (which may be the max pixel only because it is “hot”) and the typical pixel increases rapidly. I have sometimes found myself gauging the exposure times on the “Max” pixel, only to realize that the typical pixel has much fewer than 10000 counts, and the flat field therefor contains only minimal information.
    
    

    3.2.5  Flat Field Procedure

    For sunset flats, arrive on site 20 minutes before sunset. For sunrise flats, arrive on site 1 hour before sunrise.
    
    Open the dome to cool everything down, and cool down the camera to 30 degrees Celsius below ambient temperature. Point the telescope at 60 degrees altitude and 180 degrees opposite the sun. This is the field of uniform illumination.
    
    Sleep the telescope, or set targets=terrestrial. This ensures that stars will not be in the same place on successive images (and will be removed during statistical combination of the multiple calibration frames).
    
    The goal is to take multiple short (<60 seconds) exposures in each filter, such that the typical pixel has between 10000 and 20000 counts. Too few counts mean bad statistics (relative error is roughly √(counts), and flat fields must have a relative error <1%); too many counts means that the CCD response may be nonlinear. Also, the exposure time must be much greater than 0.12 seconds, because that is the minimum, and we don't want to have variations in shutter timing confusing the issue.
    
    For sunset flats, take 15 second exposures in the dimmest filter you will use until there are <10000 counts, or you have more than 10 flats in that filter. Double the exposure time when necessary, and move on to the next most-dimmest filter. Continue taking exposures with successively brighter filters while periodically modifying the exposure time such that there are always >10000 counts per pixel. Stop when you have more than 10 flats per filter, or the exposure time gets to be greater than 60 seconds, or you see stars on every image.
    
    For sunrise flats, start with 60 second exposures in the brightest filter you have used. Continue exposing until there are >20000 counts per pixel. Adjust the exposure time as necessary while moving to successively dimmer filters, always keeping the typical counts/pixel between 10000 and 20000. Stop when you have 10 flat fields in each filter, or the exposure time gets to be shorter than 15 seconds.
    
    It almost never makes sense to go “backwards” in the filter order; you will just waste time, and lose data on the later filters, with likely no extra data on those filters.
    
    The order of the photometric filters, from brightest to dimmest, is: clear, blue, visual, red, infrared.
    
    When you have finished taking flats, you need to take biases and darks. Do not change the camera temperature. Take at least 10 biases, and then switch to dark frames until some sufficiently bright object is visible, allowing one to sync the telescope on it and begin observing.
  
  

3.3  Time-Delay Integration (TDI), aka Drift Scan

• Parallel port of camera points to the East.
  
  
• If stars trail to the left as time increases, rotate the camera counterclockwise.
  
  
• I had to add 23 to the declination to get the right speed in WinScan. This is likely a result of imprecision in our knowledge of the pixel size, as the same factor of cos23 appears in my own software.
  
  
• Disable tracking with “Sleep scope”

4  Computing Facilities

5  Analysis

• SBIG software records the start of the exposure time;
• The dual serial ports on the LX200 are not independent; communication through one can interfere with communication through the other. The left port is the only one allowed to upgrade the 'scope software.
• CCDs can be highly nonlinear; our chip includes the “antiblooming” gate, which is bad for photometry. At 1x1 binning, (fullframe) it seems to clip values at around 32000 counts. Values over 25000 are probably nonlinear.

6  Shutdown Procedures

• Send the telescope to its home position. Select “Item: Utilities: Park Scope”. If the telescope has lost its position, manually put it at home by loosening the RA and Dec knobs. Remove any auxiliary counterweights.
• Replace the lenscaps and cover the telescope with the garbage bag.
• Power off the Telrad.
• Close the dome. This may require rotating the dome such that the slit control power cord reaches the outlet.
• Put away the CCD; the ends of the CCD cables should not be left to dangle in the basement of the dome; instead, the ends should be visible and accessible from the dome floor. This is for ease of use and to prevent damage to equipment. The signal cable is quite fragile, and the contacts will break.
  
  
• Power off the PC.
• Turn down the thermostat in ROARS.
• Ensure that the radio antennae are pointed at a low elevation angle; this prevents against wind and snow damage.
• Call the cops.
• When locking the outside gate, the observers' padlock should be lock the airport padlock to the gate such that either key can open the gate.
• Don't speed while leaving Potsdam; the police here are really bored.

6.1  Imperatives

• Close the dome. Otherwise the equipment will get wet.
• Turn off the 'scope. Otherwise it will continue moving at a sidereal rate and will run into itself by the next night.
• Call the cops. Otherwise we'll be accused of terrorizing the airport.
• Put away the camera; it costs 2000 U$D; 'Nuff said.

A  Appendix

A.1  Location and Directions

A.2  Instrument Specifications

A.3  Site Limits

A.4  Contact Information

1

Reynolds Observatory Astronomical Research Shed

2

Typical here means something like the statistical mean or mode. In practice, it is sufficient to look at a small sample of pixels. They should all be the same to within a tolerance of 1%; mentally reject any chance outliers you find.

This document was translated from L^AT_EX by H^EV^EA.