Possible RASA 8 Mirror Shift Issues on Long Exposures

14 replies307 views
Tommi Liinalampi avatar
Hello all RASA users!


I recently bought a RASA 8 that had been in use for several years. After the first tests I didn’t notice any problems, but when I moved to 15-minute exposures, the stars turned into streaks. It feels like the guide scope is not following the main scope’s movement. In an hour the drift is about 60 microns, i.e. 1 micron per minute. I checked the guide scope’s mountings, but the problem wasn’t there. In the end the only remaining option was mirror shift in the primary mirror. However, I wasn’t able to eliminate this movement in any way. When I exposed the focuser knob shaft, I could move it up and down quite freely. I use EAF, but I have removed without any change.

Here are a few points related to the situation:

- The problem is not in tracking, as tracking stays mostly at 0.3–0.5" RMS on both axes without large jumps.

- The guide scope’s attachments have been checked and also tested with another guide scope.

- The telescope has been kept for a long time in the same position it will be used for imaging, so that the mirror would settle into place. This has had no effect. The movement is still here after 3 hours!

- The movement is, in the long run, steady or at most slowly decelerating. Still, in the short term, occasional jumps are visible.

- When the scope rotation is adjusted so that, when looking at the image, the RA axis is horizontal and the Dec axis vertical, the movement is largely sideways (but not totally) when imaging near the southern meridian.

- The movement observed near the southern meridian is the same as the up/down movement of the focuser knob shaft. 

In the end I disassembled the back of the scope and noticed that not only does the focuser knob move, but also the part attaching behind the mirror rotates relative to the mirror holder. Is the focuser knob’s movement or the above-mentioned rotation normal? Any ideas what might be wrong?

It is worth noting that with 5-minute exposures the stars showed only slight elongation, but at 15 minutes it became very pronounced. The telescope is excellently collimated and tilt has been carefully adjusted to be nearly eliminated. I’m using a small-pixel monochrome camera (ASI183MM), which also makes errors more apparent. I know that many people use this scope in city environments and typically shoot with exposure times of no more than 5 minutes, so they may not even notice if there are issues. I, however, image under very dark skies and want to get the most out of the scope’s optics. With short exposures everything looks fine, but with longer ones the problems appear.

Does anyone have an idea what could be done?

I have integrated 4 x 15 minutes exposures in this image without dithering so you can see the movement. 



And image from focuser knob shaft which moves up and down without resistance.

Helpful Engaging
AstroRBA avatar

15 minute exposures are a bit extreme for a RASA; what are you shooting ?

Tommi Liinalampi avatar
15 minute exposures are a bit extreme for a RASA; what are you shooting ?

This is relative. It's extreme long in bortle 7, but not in bortle 3. I calculate all my exposure times carefully using light and bias frames to analyze read noise level. Actually according last test shots exposure time could be maybe 10 minutes but 5 minute is not enought to swamp read noise when using 6nm narrowbands. If I would use 3nm filters then optimal exposure time could be even 20-30 minutes even with RASA! That sounds crazy but it's true in dark areas when using little old Asi183MM and relative small pixel scale. Also F-ratio 2 is not comparable in speed to refractors because of camera obstruction.

But have you any ideas to fix the problem?
Tony Gondola avatar

Also keep in mind that the read noise on that camera in certain modes can be as high as 8 ev. I’m not a fan of super long integrations but in this case, it’s probably needed.

andrea tasselli avatar
To know for sure and fix it you'll need to open up the thing, in my experience. Sounds like a loose clamp.
Tommi Liinalampi avatar
andrea tasselli:
To know for sure and fix it you'll need to open up the thing, in my experience. Sounds like a loose clamp.

I have already opened it, but I didn't find the answer. Example this mirror holder had slight play when I try to rotate it because of a black fork below which allows little movement. But I think it's normal. I also find that the focuser knob shaft had little amount of side play, but this direction is not the same like movement in the image. Backlash can sometimes also be intentional. But would this side play to be indicator that the focuser knob rotates too easily? So what is normal an what is not...?


andrea tasselli avatar
A certain amount of slack is intentional to avoid binding. The source of the mirror flop isn't in the movement of the focuser screw but rather between the guiding/support sleeve and the mirror itself, if beyond the tolerances that need to be built in a coupling between materials with rather different thermal expansion coefficients. If it is mirror flop then the direction of the trailing should point downward, wherever the scope is pointing at. So the question is that what you observe? And secondly, is the mirror moving or is the camera instead?
Tommi Liinalampi avatar
Maybe it’s better if I share two images in PNG format. The target here is WR134 in the constellation Cygnus, and the imaging was started right after crossing to the west side of the southern meridian, meaning the imaging is done entirely toward the west, with no meridian flip in between. Both images have been mirrored and rotated to match the sky orientation. The RA axis is very precisely horizontal and the Dec axis vertical. Image1 was taken right after the target crossed the meridian. Image2 was taken exactly 2 hours and 50 minutes later. When you blink the images, you can see how far the frame has shifted — the image drifts toward the right edge — and I also thought I saw a slight rotation, though very minimal.

Since this shouldn’t be an issue with tracking or the attachment of the guide scope, do you have any ideas what might be causing it? Polar alignment was not done with a polar alignment camera, but rather programmatically using a regular camera, and it should be accurate enough. The target is also not particularly close to the celestial pole.

Image1.png 
Image2.png
Well Written Helpful Engaging
andrea tasselli avatar
I'll have to analyse the two to give a proper answer, so bear with me.
andrea tasselli avatar
BTW, what is the latitude of the imaging location?
Well Written
Tommi Liinalampi avatar
andrea tasselli:
BTW, what is the latitude of the imaging location?

It's 61 degrees north. I have to mention that these shared images were just 5 minute subs (and half of original resolution). I went to it, when 10-15 minutes were too bad because of the movement.
AwesomeAstro avatar
I had a similar issue when I first used a guidescope on my SCT, and also measured the number of arcseconds of motion to determine it had nothing to do with guiding. I know it might not be the solution you're looking for, but at that point I immediately returned the guidescope and bought, for a similar price, an off-axis guider (so the guide star sees the same motion of the mirror that the image does). Problem immediately fixed, and I haven't looked back in 6+ years. Stars went round instantly (excepting other unrelated issues sometimes). Additional benefits; I can use stellarium to plan exactly which bright guidestar I intend to use (no hunting for them randomly, which is very hard) and they are also very large at that focal length so PHD2 can easily work with them. If you can't beat em, join em! Granted I didn't even bother to try and diagnose it, I just went towards a different solution smile
Helpful Engaging
andrea tasselli avatar
Tommi Liinalampi:
andrea tasselli:
BTW, what is the latitude of the imaging location?

It's 61 degrees north. I have to mention that these shared images were just 5 minute subs (and half of original resolution). I went to it, when 10-15 minutes were too bad because of the movement.

The displacement I see is consistent with mirror flop/shift given your latitude and the angle at transit allowing for further differential shift due to apparent rotation of the gravity vector after nearly 3 hours.

Ratio of sin(65deg)/cos(65deg) =  2.14
Ratio of (average) Dx(AR)/Dy(Dec) = 1.81

* Rigid transformation matrix:

+0.999993 +0.003716 -26.983788

-0.003716 +0.999993 -7.163396

+0.000000 +0.000000 +1.000000

I don't have the fits files so I can hardly judge with any precision (and also not having the camera settings) but I shan't think you're loosing too much restricting yourself with 5 min imaging as I can see your background noise not too dissimilar from what I (would) get from my B7 sky (same subject, OIII filter).
andrea tasselli avatar
Tony Gondola:
Also keep in mind that the read noise on that camera in certain modes can be as high as 8 ev. I’m not a fan of super long integrations but in this case, it’s probably needed.

It does not. There are no "modes" either, just gain settings.
Tommi Liinalampi avatar
Thank you very much for the good analyze!

Yes, it's possible to use 5 minute exposures without losing toi much. When these images were taken it wasn't perfect dark sky and Oiii filter passes much more skyglow than example Sii filter. I think that I still can swamp read noise five times instead of ten when it's perfectly dark. Of course it's wise in this case go with shorter than 10-15 min exposures if there is no solution to the mirror movement.

However if it is possible to fix the problem, then I it would be nice to fix. After taken images 2.4um pixel size camera I drizzle x2 images. I'm close to pixel scale which I have when using my EdgeHD800! I bought RASA 8 for high resolution scope instead of just fast scope. And after careful collimation and tilt adjustment images are amazing sharp when using shorter exposures. Problem is longer exposures.

But I have to use 5 min exposures if there is no solution.