60cm Deltagraph for Korea Astronomy Observatory
The 60cm Deltagraph has been designed for search for Asteroids and Comets. It was optimized for a field diameter of 84mm, the focal ratio is f/2.8. We made cost effective tube with Carbon Truss tubes to avoid a focus drift due to temperature changes.
The Focuser can carry a load of 10kg and is controlled with a servomotor. The position display is down to 1 micron. The focus can be controlled via RS232 and computer, the focus dll is compatible with MaximDL and other programs.
Testing the Deltagraph
||LOMO as usual made an interferogram of the main mirror. They did not
test the system with the corrector, because in this case they would have needed a
provisorial tube and corrector holder which would have increased the costs to much. So we
promised to test the system here at our workshop.
The main mirror was specified with surface accuracy of L/8 . This is more than sufficient for a pure astrophotographic system because the diffraction disc is only 3 microns at f/3 and therefore, the system would not need to be diffraction limited.
So we installed the optics in the tube (main mirror and the 3-lens corrector) and
where very interested to see the result. We have already built 3 such 60cm systems and the
ccd exposures have been fine in any case but this was the first to be tested in front of
the autocollimation mirror.
We used an artificial star with a diameter of 15 microns and the LOMO 60cm plane Sitall-mirror which we have in our workshop for testing the larger systems.
We use a beam-splitter that reflects the light of the star in the beam. The light enters the Deltagraph as parallel light (as would be from a star), is reflected by the plane mirror and gets focused by the deltagraph again. So the test is very sensitive because every error in the system (and air turbulence) is doubled.
First of all, the system was collimated with a laser.
There is a good thing and a bad thing about adjusting very fast astrophotographic systems. The bad thing is, that the tolerance against a lateral offset of the prime-focus corrector cell is about 0.5mm. Its hardly impossible to adjust a system with a laser so accurate. The good thing is, that if you use the mirror tilt as a compensator, the tolerances get very large. This means that you can have a quite strong tilt in the corrector unit or a lateral offset of even 5mm ! and you can still get the system to a good performance just by collimating the primary mirror by examining the out-of focus images of a star with a ccd or video camera.
|I used a TouCam (a simple webcam) that I installed after the
beam-splitter. There is a slight spherical aberration of L/4 wave due to the beam-splitter
that causes some error at f/3 (don't even think about testing this system with an
This is the way the out-of-focus image looked like after the
laser-adjustment. The size of the out-of-focus image is here about 0.5mm. Indeed some
decollimation can easily be seen and also a lot or turbulence.
As with nearly every system the collimation procedure is such that you tilt the main mirror to make the star move in the opposite direction where the central obscuration is offset. In this case the central hole is clearly offset to the left so you have to find an adjustment screw at the main mirror that moves the star to the right. We had to move the star about 5mm on the chip to get the central hole roughly centered.
|As you can see to the right maybe it was a little to much because the hole
is now slightly offset to the right. But since we have to demount the telescope in smaller
pieces the final collimation will have to be done anyway by the customer. But keep in mind
that if you have reached a point there you have pin-point images you may not need to spend
hours in searching for the perfect collimation.
To get the through-focus sequence we used the computer controlled focuser to get smoothly through the focus. Please note, that no chromatic error can be seen here, so our calculation of the corrector seems right (surprise, surprise).
|Local Seeing and air turbulence
is a big problem during testing. But it also shows the high sensitivity of the test. Indeed it was clear that the performance of this system was limited by the local seeing in the test workshop. This problem gets bigger with the size of the telescope. But imagine, that we have only the seeing here that appears in the 1.8m tube length. There is not 10km air mass between us and the star but only 1.8m x 2 (since we have a double pass autocollimator).
|Here you can see the 15 micron hole in 4x magnification. Keep in mind,
that the error has been doubled but even due to that you can see a clear image of the
hole. So at the real Sky you could easily resolve 7 micron with this system, probably
The light of the artificial star had been dimmed not to saturate the TouCam CMOS sensor.
It was also possible to perform a ronchi test. I could not get the complete aperture in one frame so I moved the TouCam around a little in front of the ronchi ruling. The ronchi ruling had 10 Lines per mm.
You can see that I used my arm in front of the aperture to shift the TouCam to scan through the aperture. Not very professional but since the artificial star was so perfect I knew that the Ronchi-Test would be only for fun.
Most of the error you can see is due to the warm air rising from my arm. If you look at the video (it may take some time do download) you see that if you average the turbulence the lines are straight as they can be.
To conclude, these tests are much more demanding than taking a real image at the night sky. Seeing effects and guiding will limit this system in all cases.