This web-page describes the theory behind the tip-tilt system.
See the following pages for instructions on using the system and the software.
- A tutorial of how to run the tip-tilt system.
- The manual for the tip-tilt software (Otto).
- The manual for the guide stage and positioning software (Atlas)
- Tip-Tilt / Autoguider / XY Stage Usage Notes.
Contents
What is Tip-tilt?
When the light from a star passes through the earth's atmosphere, it is refracted many times by the turbulent atmosphere. Consequently, the image of the star at the telescope will seem to move around and be distorted. In addition, the star will seem to move around because the telescope itself is vibrating from the wind and imperfect mechanical couplings to its drive motors, the dome, and other sources of vibrations. Exposures of the star that are several seconds or longer will show a blurred star due primarily to the image motion of the star. The tip-tilt system on the University of Hawai`i 88 inch telescope is used to compensate for the motion of the image caused by the atmosphere and telescope vibrations. Much higher resolution images than would normally be possible are the result.
Diffraction-limited images in K (0.25 arcsec FWHM) of several minutes exposure have been obtained, when the guide star is close to on-axis. The tiptilt system successfully stabilized an image through the mag 4.5 earthquake in May 1995. It also stabilized an 8th mag star directly into a 35 MPH wind.
The correction of atmospherically-caused image motion is best for the guide star and degrades as a function of angular distance from the star. The region over which there is improvement over the static seeing conditions is known as the isoplanatic patch. The size of the isoplanatic patch is not well understood at this time, but it clearly will vary with atmospheric conditions. A nominal size of 40 arcsec radius can be assumed, with the degree of image improvement a function of the distance from the guide star. At this time it is not clear how much motion is due to common mode vibration (wind shake, dome movement, micro-tremors, etc.) vs. atmospheric distortion. 0.4 arcsec resolution has been possible even with guide stars up to 5 arcmin away, but that is not normally the case.
The tiptilt system was successfully used to obtain some of the highest quality ground-based images of the Shoemaker-Levy 9 comet impact on Jupiter in July of 1994. It also took high resolution images of Saturn prior to the ring-plane crossing of May 1995.
Observing in I and all filters available for QUIRC is possible. The beam splitter reflection cuts off above 720 nm. Consequently, the best guide for magnitude limits is the V magnitude of the guide star, although its color will also have a strong effect. A V=13.4 star was the faintest object that worked with the system. Realistically, V=13.0 is the limit for most observing. A new CCD camera is being constructed which should allow fainter guide stars.
How Tiptilt Works
Below is a diagram of the complete tiptilt system at the UH 88-inch telescope. The system consists of a secondary mirror mounted on a platform that can rapidly tip and tilt, a wavefront sensor which determines where the star is, and a control system which repositions the tip-tilt platform to recenter the star image on the wavefront sensor.
The basic operation of the control system is easy to understand. An image of the star is formed on the wavefront sensor-and the control system tries its best to keep the star image at the center of the wavefront sensor.
|
The visible light is diverted to the quad cell, while the infrared passes through the dichroic to the science camera. |
|
The light from the guide star falls onto the quad cell. The secondary mirror is moved to center the star on the quad cell. |
|
The star will move from the center because of atmospheric motion or because of telescope vibration, so the secondary moves may times per second to keep the star in the center. |
In the case of our system, the wavefront sensor is a CCD camera. Part of the CCD is used as a quadcell-the simplest kind of wavefront sensor. The CCD camera is mounted on an X-Y stage in the autoguider box. This allows it to move around and locate stars at various positions relative to the center of the imaging field of the science camera. This can be viewed as moving the quad cell around the sky.
The dichroic is mounted at a 45 degree angle to the incoming light and reflects the visible light to the CCD camera. Its mounting point normally comes in from the East and everything to the east of the mounting point will be obscured by the mount. Thus one would normally want to use guide stars to the east of the center of the field. The entire instrument package can be rotated, however, allowing the use of stars at all angles.
Observational Restrictions
Magnitude Limitations
The Tiptilt system is currently capable of fast guiding on star as faint as 13.5 Vmag on a dark night. With a full moon, depending on how close one gets to the moon, the faint guide star limit is about 12.5 Vmag. With a 100x ND filter, it has been used to guide on 5.0 VMag guide stars. In the faint limit, the quality of the guiding degrades, as the integration time must be increased to obtain sufficient signal to noise. The update rate is thus decreased. At V=13.0, the loop rate is only 25 Hz or so-meaning that it can correct for seeing of only 2.5 Hz or slower.
Guide Star Position Limitations
One of the biggest restrictions with the current system is how far one can dither. Right now, that means only 9" East, 40" West, and maybe 20" N-S, from center. Beyond those distances, the dichroic will vignette the edge of QUIRC-the edge of the dichroic will appear in the image. This range should be expanded with a new, larger dichroic.
It is now possible to rotate the instrument package from its nominal 270 degree position. At the 270 degree rotation, the dichroic probe comes in from the East. It can now be rotated either to the 0 degree position, where the dichroic probe comes in from the North, or to the 180 degree position, where the dichroic comes in from the South. It is not possible to rotate to 90 degrees. Soon, any rotation between these 90 degree steps will be possible.
Software
The following software is used to control the tiptilt
system in the SDSU controller. All three parts are
combined to form the tiptilt control software. Please
note that this is software which is still being refined,
and is rife with false starts and errors in documentation.
Also note that if you don't know what
MPY Y0,Y1,B Y:GAIN,Y0
means, this will be of little interest to you.
The actual control software |
|
List of allowed commands |
|
The clocking waveforms and memory storage |
Who's Responsible?
A lot of people contributed to this instrument, including:
Len Cowie,
Buzz Graves,
Kevin Jim,
Tim Keller,
Reni Kupke,
Gerry Luppino,
Mark Metzger,
Malcolm Northcott,
Wes Nakamura,
Andrew Pickles,
Ed Sousa,
Alan Stockton,
Hubert Yamada,
Tony Young
Kevin Jim has been doing the system integration, so he's the person to contact for questions about actually observing with it.
Last modified on May 26, 1995.
