View Full Version here: : Guiding woes
26-03-2012, 09:19 PM
I would be grateful if anyone has an idea or so why I am getting poor guiding in DEC only. Paramount MX with a DSI RC10C firmly clamped. Imager is a QSI 683wsg with an SBIG ST-i guider (built in shutter). My Polar alignment is excellent according to the TPoint model and I use CCDSoft as both imager and guider. I took 25 x 600 sec. images last night and all of them have oval stars, some worse than others, with the long axis in the DEC direction. CCDSoft guiding calibrated fine and the X axis was reasonably stable, varying about 0.6 pixel pp. (ST-i pixels of course) but the Y axis error wandered in excess of 4 pixels pp. The effect was as if the MX has backlash in the DEC axis so I ask is this possible? And if so, does anyone know how to test it so I can ask Bisque how to fix it? Thanks.
26-03-2012, 09:26 PM
Sounds like DEC was over correcting. Are you using relays or direct guide ?
Suggest you try direct guide if usign relays. Much better results with Direct Guide. Try tone down max move time and min pixel movement. ie make sure it does not chase seeing.
Backlash is almost impossible on these mounts with belt drive and spring loaded worm. You can test using illuminated reticle and a barlow on a star. move dec one way at 1x speed then other. should move both directions immediately. If it does not. Pop off dec cover and do the same test but watch the belt. If you see slack in the belt thats and easy fix to tighten. Just watch the video I posted a while back. It explains how to tension the belt.
26-03-2012, 09:48 PM
As I have the same scope and mount, but different camera (ST11000M) I may be able to help.
First try guiding with Dec corrections turned off. If your polar alignment is good, then you should not need to guide in Dec. In fact with the MX, a big model, PEC you may not need to guide much at all, though in practice guiding always gives better results.
Try taking shorter exposures, a couple of minutes should be enough to show any Dec guiding problems. Keep your eye on the dec error in CCDSoft (or whatever you're using to image with), if it's not much and you still have oval stars then the problem is more than likely optical.
Are you at the magical 6" back focus, if not the off-axis stars will suffer. Can you email me one of your subs to look at?
I have found with this scope, if the collimation isn't perfect you get oval stars in one area of the frame and nice round ones elsewhere. I have started to think about going for a smaller CCD as the edges of my FOV are really difficult to get right, I think that the STL11000 chip is pushing the corrected circle for this scope.
27-03-2012, 12:37 PM
Chris, I am using Direct Guide and the Calibration distance is 60 a.s. in X & Y. Minimum Move is 0.01 a.s. and Max is 1 a.s. Aggressiveness was 10 but had no effect increasing it to 13. I imagine if I was chasing seeing (a sensible suggestion) it would be apparent in both axes. I just noticed that the CCDSoft Drive Calibration Results show a slightly greater pixels per second movement for Y (1.2) vs. X (0.7) but I'm not sure how to alter that.
Stuart, nice to find another RC10C / MX user! My camera should be pretty much at the 6" back focal point. The setup is Camera, SX-AO-LF and a home made adapter. Corrections were applied for the AO glass and the Filters. I have been concerned that the Collimation might not be perfect so I bought a Tak. Collimation 'Scope but the blurb in the box said it wasn't useable for a Corrected Scope like ours! DSI keep promising to provide a Collimation Procedure so I will have to chase them a bit. I will send you a sub from Sunday via PM. since I can't get an image small enough to post here. 200K maximum!
Thanks guys for your suggestions, SkippySky suggests Wed/Thurs may be clear so I will give them a try and post again.
27-03-2012, 01:03 PM
4 pixel guiding errors is atrocious and would never give round stars with any scope.
It needs to be more like .2 to .4. With my PME recently at 3 metres focal length I was getting errors
up to 1.5 (very high) and still getting round stars with PEC on. It surprised me. But once it settled down
it was more like .1 to .5.
If in fact your PA is good (double check with drift aligning if any doubt) then next most likely error is imbalance.
If your scope is heavily out of balance you will get that or if there is bad cable drag.
If you are getting 4 pixel errors and your max move is 1 then your autoguider is effectively never going to do a correction as you are always outside the maximum move before it corrects.
Here is the list of guide errors:
1. Guiding is only as good as your polar alignment. So check your alignment is in fact good. Your guide errors are telling you otherwise than T-point.
2 You have bad flexure - something is very loose, guide camera, the camera itself. Is it a solid screwed adapter fixing to the scope or is it a slide in eyepiece holder type arrangement?
Is your OAG tight and no loose parts?
3. Balance. Scope must be very evenly balanced at the angle you are going to be imaging at. If the centre of gravity is badly skewed it can be
nicely balanced horizontally yet badly imbalanced at 35 degree angle. Check this.
4. Cable drag. Power cables, USB cables, can get tangled up and pull on the guide camera and cause it to be off.
5. I take it your mount is relatively level.
6. Turn PEC off in case that is not setup properly. Turn Protrack off as well.
7. Make sure your Sky X is running on the correct time and location.
8. Sometimes selecting a different star can get guide errors down heavily.
9. Are your image frames from the STi clear of artifacts? Lodestar can get a white line or a hot pixel down one side
and CCDsoft does not autodark from libraries (as far as I can tell and I have tried several times) when autoguiding.
Click on the auto button in the Autoguiding tab and see what it selects with the white square. Maybe its chosing
a hot pixel. The clue to that is the guide error does not change (being the same distance in the image every time)
and often a large error like 8 or 15.
27-03-2012, 04:05 PM
Just send Rich an email. He'll want your serial number, then he'll send one to you. It's comprehensive. The Tak collimation scope will get you in the ball park, but it doesn't work on my scope to get fine collimation.
Join the DSI Yahoo group as well.
28-03-2012, 10:47 AM
Thanks again Guys.
Greg, I should be able to do more checking tonight and will look at all your excellent points. I have a Starrett Machinists Level and got the mount base very level. I'll check the drift alignment as it seems I have been maybe too trusting with the TPoint Polar Alignment. I noticed a peculiar effect with the guide image box. The initial 'large' image from STi was noisy but the Dark frame fixed that. Image reduced down to 30 x 30 pixels with the selected star centred and guiding started. During the guiding session for the 10 min sub, even though CCDSoft indicated it reused the dark frame, only about 1 in ~10 images lost their hot pixels which then came back for the next (or so) guide image. Most peculiar and I can't describe it any better!
I always balance pretty well before a session since the mount only is 'permanently' mounted and I have to re-fit the OTA. I have no trailing cables, all power and USB cables go through the mount and then to the cameras, OA, 2x USB Hubs and the telescope from the Versa Plate. I have just looked at the mount and there may be a small problem in this area. The setup of the MX describes how to get extra cables through the mount, up to the Versa Plate. I did that with 2 USB cables and a 9 wire flat cable and just found they (or something) is acting slightly like a spring, winding up and releasing as the mount turns. I'll check this further.
Camera and OAG are firm to the scope. My eyes are no good anymore for viewing so my setup is solely for imaging. The camera is clamped to the AO which is screwed to a spacer screwed to the scope.
Stuart, I will email Rich and see what he suggests. I put an image into your Dropbox and will join the DSI group. I hadn't thought about there even being a DSI group!!
28-03-2012, 11:18 PM
I've cast an eye over your sub of OC.
Doesn't appear to be a collimation problem.
Focus isn't quite right, these scopes are very finicky on focus I've found, watch out for cool down issues, the focus changes.
It does look a bit like a guiding issue, but not in DEC. Are you sure your DEC axis is horizontal in the sub? Either way, with AO and an OAG, your mount's guiding should not be being utilised much if you're PA is good.
Which AO unit are you using? Try not using it and see if the star shape changes. Does the mount need to be "bumped" when the AO runs out of travel?
29-03-2012, 03:25 AM
Interesting problem. Dec related issues don't suggest a mount problem. Most likely this is a software problem and I bet it is CCDsoft at play here. I never once found it to work 100% of the time for me. Do you have a copy of Maxim? If so try that and see what transpires.
When you say the camera is clamped onto the AO; why is this not screwed onto the camera? The 683 is a bit heavier and that could be pulling a bit.
Also just check the screw on the WSG that could be loose. In fact check all the screws on the guider attachment. Those things are pretty tiny and the allan key might not have been the right one. Just a guess here but worth looking.
I'll think a little more and get back to you if I think of something else.
Lucky there are a few more SB owners around now and QSI owners too.
29-03-2012, 04:08 PM
You are correct Stuart, I did some more measurements last night and the errors in the image are in the RA axis. Sorry for the confusion. I tried to perform a Drift align and, following a star on the Meridian + ~ Equator, there was a significant drift to the East in RA with a very slight drift in DEC. The TPoint Model said the MX was 0.5 clicks out in Az. and 0.0 out in El. but I adjusted the MX Azimuth anyway and the star eventually didn't move in DEC over a period of 30 minutes. Moved to East and 20 deg up but got confused on how to read the EW NS drift. I crudely adjusted the MX in Elev and subsequent images were better although there was still significant X and Y variation in guiding corrections. The stars were much better but still out-of-round, slightly triangular if anything.
I agree the focus isn't spot on but I have had that concern since I bought the RC10C. I use FocusMax so if anyone has a better solution I would love to hear it.
Paul, you are correct, I was using CCDSoft and had some really wierd problems last night with the guiding camera, ST-i, so I did switch to Maxim but still had the guiding variations. They were less than the previous time and never exceeded 1.2 pixels but still more than Greg thinks should be. I have not yet installed my SQM-LE so I don't know what the seeing was like but feel that was the cause of the guiding variations.
The camera is clamped on to the empty, but still rigid, SX-AO-LF body as the innards are in the UK at the doctors. I've just heard they have been fixed and should be on the way back soon. There doesn't seem to be a way to screw the AO into both the QSI and the 'scope, one side finishes up using a clamp.
I am really grateful that there are SB users and QSI users these days. It is great to get such good advice. I must say though, last night there was very nearly one less! :(
29-03-2012, 05:48 PM
So, was AO on or off during tis testing?
29-03-2012, 06:02 PM
If stars are triangular it sounds like you have coma as well.
So you need to differentiate between coma and guide errors. Guide errors are in the same direction across the whole image usually,(sometimes they can rotate if your Polar Alignment is badly off).
Coma is comet shaped stars usually worse off centre and in the corners.
Also tilt will give out of focus stars in one side or corner of the image. If your scope has a corrector then tilt could also cause coma like effects.
Also your corrector (if using one) may have a metal back distance and you are not close to it causing coma.
If polar alignment was off and you got an improvement I suggest a brand new T-point model (delete the old one) and after say 50 point model read the polar alignment report and adjust the mount.
Have you done PEC yet?
I wonder if the PMX PEC software is bugged. Chris had a lot of trouble with his and PEC made errors worse. I did a standard PEC curve and I am not convinced it is improving guiding performance also. I get the same or better results with PEC turned off. Also the curve I made and what was saved in the mount seem different.
Chris used Pempro and got a great result.
I have used Precision PEC on my PME and it definitely helps get round stars.
29-03-2012, 07:57 PM
I think as the "innards are in the UK at the doctors" we can safely assume it was off.
30-03-2012, 07:44 AM
If you have no drift in DEC, then the PA is fine, even a small amount can be guided away, I would still trust the SB PA using a super model though, it works. Still sounds like a guiding problem.
What are your guide intervals, i.e. what's your guide exposure time? Does the roundness of the stars change when you change the guide interval? Sometimes a longer guide interval is better as you don't chase the seeing.
Of course all this is moot when you get your AO back on line.
Mk I Eyeball works for me. The stars are reasonably obviously out of focus near the edges of the field, they get elongated, simply focus until they become round, shift focus field, round the stars, shift to point of interest and ensure focus. Works OK, needs tweaking probably about an hour after initial cool down.
vBulletin® v3.8.7, Copyright ©2000-2013, vBulletin Solutions, Inc.