#1  
Old 06-02-2013, 12:24 PM
pvelez's Avatar
pvelez (Pete)
Registered User

pvelez is offline
 
Join Date: Aug 2006
Location: Sydney
Posts: 1,250
Odd guiding behaviour

Spotted something very odd last night.

Looking to refine my polar alignment on my PMX over the last few days, I have undertaken a few TPoint runs. With about 150 points, I am out by a few tics on altitude which I need to fix up. Still, TPoint gives me RMS of 29 arcseconds which is ok for now.

I wanted to take a few shots to see how Protrack impacted on tracking. As you might expect, there was some minor drift in 120second images - I'm imaging at 1600mm f/8 using a QSI 583. Binned x2 I have an image scale of 1.4 arcseconds/pixel.

So I turned on my guide camera - an Atik Titan on a short tube Orion ED80 and banged out another image. Still had some drift?!

I reasoned that there is an issue with my guide cable and then used the manual move function. No dramas there - ask the mount to move and the target star duly moves.

So then I looked at the graph - no major movement at all. It just tracks very nicely at about 0.1 pixel variations. I'm using Maxim for guiding.

I'm stumped on this. The Maxim graph shows not only the movement in the target star but also the guide instructions issued to the mount. It doesn't suggest that there is any issue. I thought I might have a cable snagged or something similar but that shoudl have shown up in the graph.

But the pic clearly has elongated stars.

What could be doing this? I'm wondering whether mount is not receiving the move instructions in the ordinary course - but that still should show up in the graph.

I did notice that opening the TCS Bisque window in TSX that the guider relays are not illuminated when a guide instruction was sent - I didn't check it when the manual move instructions were sent. I'll try that this evening.

I am guiding using guider relays and the ST-4 cable. Guiding rate is 0.5.

Has anyone else seen this?

The funny thing is, its a new development. Go back tio December and everything worked fine.

Pete
Reply With Quote
  #2  
Old 06-02-2013, 04:51 PM
Geoff45's Avatar
Geoff45 (Geoff)
PI rules

Geoff45 is offline
 
Join Date: Jul 2006
Location: Sydney
Posts: 2,631
Are you using a guidescope? If you are, I would guess that it's differential flexure. In which direction are the stars elongated? If it's not E-W, then it's almost certainly flexure.
Reply With Quote
  #3  
Old 06-02-2013, 08:41 PM
pvelez's Avatar
pvelez (Pete)
Registered User

pvelez is offline
 
Join Date: Aug 2006
Location: Sydney
Posts: 1,250
Thanks for the tip Geoff

so what is happening is that the guidescope thinks its tracking well but there is some movement on the main scope or connection with the camera so that the guide camera's response doesn't match what the imaging camera is seeing.

That makes sense.

I have tightened everything up but there is probably some slop somewhere in the system.

Have a look at this screenshot - this is a 5 minute exposure in R. The graph reports no movement (other than when moving from about 1 arcsecond to the zero line) above about 0.25 arc seconds. I've zoomed in 200% to capture this.

If you are right - which I strongly suspect you are - then I do need the new Lodestar so I can use the QSI with the OAG rather than with the guidescope. Still not sure why this issue has arisen lately.

Pete
Attached Thumbnails
Click for full-size image (Screen Shot 2013-02-06 at 9.38.48 PM.png)
117.6 KB62 views
Reply With Quote
  #4  
Old 06-02-2013, 08:59 PM
Geoff45's Avatar
Geoff45 (Geoff)
PI rules

Geoff45 is offline
 
Join Date: Jul 2006
Location: Sydney
Posts: 2,631
Quote:
Originally Posted by pvelez View Post

Still not sure why this issue has arisen lately.

Pete
Who knows? Different part of the sky, nylon tips of the guidescope screws softening, a loose screw somewhere.
Reply With Quote
  #5  
Old 06-02-2013, 10:05 PM
Joshua Bunn's Avatar
Joshua Bunn (Joshua)
Registered User

Joshua Bunn is offline
 
Join Date: Oct 2011
Location: Albany, Western Australia
Posts: 1,459
Try orientating your camera "north up" so that if you do get any elongation, you will know whether its E-W or N-S. I would also sugest to get PA closer to the pole, But thats probably not the issue.
Can you use direct guide through the mount?

Josh
Reply With Quote
  #6  
Old 07-02-2013, 12:02 AM
PRejto's Avatar
PRejto (Peter)
Registered User

PRejto is offline
 
Join Date: Jan 2011
Location: Rylstone, NSW, Australia
Posts: 1,397
You have not asked about polar alignment, but I had difficuty with the last bit of adjustment in altitude. I picked up a great tip from the forum at Software Bisque. Namely, use the camera to make sure that when you adjust altitude you actually get what you have dialed in. It seems that for very small movements in altitude the mount may or may not move exactly as planned. You end up just chasing it up and down wasting too much time collecting a million points.

Here is how I did it and got it spot on. Align the camera so the X axis is parallel to motion in RA as suggested in the post above. Let's say that T-Point PA says you need to move the mount down by 1 tic, which is also given as "X" arcsec. First slew a location near zero dec and take a photo taking note of exactly where the mount is pointing. Then use the Jog control in TheSkyX to move the mount exactly the amount you need to move the mount, but move it in the opposite direction. In other words, move the scope up by x arc sec if you need to adjust the altitude down. Now go ahead and lower the mount and take another photo. If the altitude control responds exactly as planned the scope should be back pointing exactly where you started from. It probably won't be so just keep adjusting until you are back to the beginning location. I know some people use a video camera to do this but I couldn't be bothered to take off my regular camera, refocus, etc etc. It worked out great and I nailed the final bit of alignment. The tricky bit is knowing what direction you are moving when using the jog controls. Just experiment a bit to work it out.

By the way, it's worth pointing out two other important not so well known things about T-Point. It seems that the polar reports are not very useful if you collect fewer than 40+ points, except perhaps when the mount is very far out. It's also essential to collect points on east and west of the meridian, and finally, the T-Point polar alignment rountine is not iterative. According to Patrick Wallace if enough points are collected it is theoretically possible to align the mount in one go. That would be hard to do in reality, but the point is that each T-Point run has nothing to do with the previous run; the resuts say where to put the scope to be aligned. If you succeed in putting it where it says to put it you are done.

Anyway, that is my take on this. YMMV.

Peter
Reply With Quote
  #7  
Old 07-02-2013, 05:15 PM
pvelez's Avatar
pvelez (Pete)
Registered User

pvelez is offline
 
Join Date: Aug 2006
Location: Sydney
Posts: 1,250
Thanks Peter - I'll give it a go this evening.

I do have a dumb question - which your method neatly avoids - if the TPoint PA report says to move the right az knob, is that the one on the right facing the mount from the front ie the counterweight shaft? Or is it facing it from behind ie the electronic connection side?

Pete
Reply With Quote
  #8  
Old 07-02-2013, 05:17 PM
pvelez's Avatar
pvelez (Pete)
Registered User

pvelez is offline
 
Join Date: Aug 2006
Location: Sydney
Posts: 1,250
Quote:
Originally Posted by Joshua Bunn View Post
Try orientating your camera "north up" so that if you do get any elongation, you will know whether its E-W or N-S. I would also sugest to get PA closer to the pole, But thats probably not the issue.
Can you use direct guide through the mount?

Josh
Sadly no - I am using CCD Autopilot and the pier flip only seems to work properly using a ST0-4 cable

Pete
Reply With Quote
  #9  
Old 07-02-2013, 08:47 PM
PRejto's Avatar
PRejto (Peter)
Registered User

PRejto is offline
 
Join Date: Jan 2011
Location: Rylstone, NSW, Australia
Posts: 1,397
Quote:
Originally Posted by pvelez View Post
Thanks Peter - I'll give it a go this evening.

I do have a dumb question - which your method neatly avoids - if the TPoint PA report says to move the right az knob, is that the one on the right facing the mount from the front ie the counterweight shaft? Or is it facing it from behind ie the electronic connection side?

Pete
Hi Peter,

I think SB changed the way these reports read in an update not to long ago. I think the old report was easier to follow because it would say loosen the west knob, tighten the east knob. No idea why they changed this, but my best guess is that you would be adjusting the mount from the control box side of things, so that would be the orientation. I'm sure this is described in the updated manual (which I havn't got yet).

Peter
Reply With Quote
  #10  
Old 09-02-2013, 08:02 AM
pvelez's Avatar
pvelez (Pete)
Registered User

pvelez is offline
 
Join Date: Aug 2006
Location: Sydney
Posts: 1,250
Just took delivery of a Lodestar. When the c-mount adapter from QSI arrives I can ditch the guidescope and use the OAG on the camera.

Ran a few images last night - still eggy stars even though I tightened up everything and removed the rotator. Is it collimation? Something else?

Interestingly, I watched the guider relay indicator in Bisque TCS - they flashed when the autoguider was calibrating but not when imaging. I assume this is because the guide signal is so brief. Not sure if this is relevant.

Pete
Reply With Quote
  #11  
Old 09-02-2013, 09:55 AM
pvelez's Avatar
pvelez (Pete)
Registered User

pvelez is offline
 
Join Date: Aug 2006
Location: Sydney
Posts: 1,250
So I bit the bullet and bought CCD Inspector.

Looks like my collimation needs some work - but not too much. The collimation error is between 8 and 13 arc seconds on my RC8.

Also analysing a succession of images from the last few nights, it seems that the aspect ratio shifts between images with a max aspect ratio of 32% on 1 image and only 14% 2 images later. I am assuming that aspect ratio measures the elongation of stars - tell me if I'm wrong here. This says to me that its not a consistent problem ie tilt or collimation which confirms my initial diagnosis that its guiding problems or an inconsistent flexure issue.

Please tell me if you disagree.

I guess this is progress of sorts

Pete
Reply With Quote
  #12  
Old 09-02-2013, 04:01 PM
Joshua Bunn's Avatar
Joshua Bunn (Joshua)
Registered User

Joshua Bunn is offline
 
Join Date: Oct 2011
Location: Albany, Western Australia
Posts: 1,459
Pete,

That collimation error really needs to come down. Try magnifying the cross hairs in ccdins to get a better look at whats happening. Also, Steve67's (username) method, from here, of collimation is really accurite i found aswell. I just got the shadow right in the middle of the defocused star, using the centre star on the chip, and collimation was sorted, even CCDINS agreed.

Yes, the aspect ratio measures the out of roundness of stars, The lower the % the better.

It may not be a tilt of collimation issue, but collimation or focuser/camera flexure could still change depending on the rigidity of yor settup and the location in the sky your pointing to. But if your getting significantly elongated stars, all over the chip its unlikely to be either of these.

HTH,
Josh
Reply With Quote
  #13  
Old 09-02-2013, 04:27 PM
pvelez's Avatar
pvelez (Pete)
Registered User

pvelez is offline
 
Join Date: Aug 2006
Location: Sydney
Posts: 1,250
That helps a lot Joshua

What I find difficult with these tools is applying the results. Its all good to measure the collimation error but what magnitude of error is unacceptable.

Maxim has a tool that measures a star profile and reports flatness - I assume its the same sort of calculation but I have no idea what is or is not an acceptable error.

Out of interest, what FWHM do you set as acceptable? CCDAutopilot allows for auto rejection of images above a set limit but I'm in the dark as to what to set it to.

Need to get started on collimating the RC8 I guess.

Pete
Reply With Quote
  #14  
Old 09-02-2013, 04:50 PM
Joshua Bunn's Avatar
Joshua Bunn (Joshua)
Registered User

Joshua Bunn is offline
 
Join Date: Oct 2011
Location: Albany, Western Australia
Posts: 1,459
Well, I dont know how easy/difficult your RC8 is to collimate, but with my scope, CCDI reports between 0 and 1 arcsec error in collimation and tilt in x and y should be able to be eliminated i.e, .1 to 0".

FWHM will depend on factors in your control such as your equipment and factors out of your control e.g, atmospherics. So if you havnt overexposed the image, havnt got bloated stars, have good focus, good collimation, no ccd tilt and not viewing over thermals (pavement, concrete, houses), then your doing mostly what you can to reduce your FWHM. I realy dont know what you should expect at your location, but 1.5 - 2.5" is a pretty good result. Others may chime in here .

Josh
Reply With Quote
  #15  
Old 10-02-2013, 09:58 PM
PRejto's Avatar
PRejto (Peter)
Registered User

PRejto is offline
 
Join Date: Jan 2011
Location: Rylstone, NSW, Australia
Posts: 1,397
I am imaging with a refractor so do not know if my comments are valid for your scope. I too use CCDIns to check every image. I have noticed a few things that lead me to believe that you cannot depend on a single result from CCDIns to draw much of a conclusion, rather a general conclusion can be made from a group of images. For one thing the various measurements do not seem to be independent of one another. Strangely, I have noticed that when FWHM is lower it is often accompanied by somewhat higher aspect distortion. I think it might have something to do with distortion having a possible larger effect on on a smaller diameter object...that is, distortion possibly shows a higher percentage the smaller the circumference. This is just my theory....no idea if I am correct! However, I have also noticed that aspect map changes when I change filters, even after careful focusing and that results are very difficult to achieve that are in close agreement. At best, I think the tool is very useful to give general trends and clearly a conclusion can be drawn only if a lot of images are taken and the results averaged out. There is discussion about this in the manual for CCDIns along with specific instructions on just what sort of image can be used to evaluate field curvature, etc.

Pete, re guiding. What do your images look like if you take really short exposures...say 10 sec. Do you see much distortion? What you see in a short exposure is probably optical...with longer exposures it guiding or flexture.

Peter
Reply With Quote
  #16  
Old 11-02-2013, 08:37 AM
pvelez's Avatar
pvelez (Pete)
Registered User

pvelez is offline
 
Join Date: Aug 2006
Location: Sydney
Posts: 1,250
Quote:
Originally Posted by PRejto View Post
Pete, re guiding. What do your images look like if you take really short exposures...say 10 sec. Do you see much distortion? What you see in a short exposure is probably optical...with longer exposures it guiding or flexture.

Peter
Peter

thanks for your response.

The shortest expsoure I tested on was 120 seconds and found some eggy stars still. That said, I also managed a few 300 second images and found nice round stars hence I thought it was likely guding.

I have yet to try very short frames eg 10 - 20 seconds. Will give that a go.

I've done some more thinking/poking/prodding at the weekend (without any actual imaging as it was cloudy) and reached the following tentative conclusions:

1. My scope wasn't properly balanced - I removed the rotator but didn't compensate by rebalancing. Elementary error of course. As a result, there were some shifts which threw the guiding out.

2. My PA isn't perfect. I couldn't get TSX to play nicely but have since determined that this was because I hadn't synced in TSX before trying a plate solve. I can address this when I next have clear skies.

3. My collimation is not as good as it could be. At around 13 arcseconds, this is also contributing to eggy stars. I can look to settling this when I next image as well.

4. There may be some flexure as well though this is less likely. I've tightened everything up and am waiting on adapters for teh QSI so I can use the Lodestar on the OAG and then ditch the guidescope. This makes balancing a bit easier.

I'm not sure whether there is anything else at play here - but if I get these issues under control, I can then start focussing on the mount (excuse the pun).PEC and Protrack are the next steps.

Now for some more fine weather!

Pete
Reply With Quote
  #17  
Old 11-02-2013, 09:17 AM
PRejto's Avatar
PRejto (Peter)
Registered User

PRejto is offline
 
Join Date: Jan 2011
Location: Rylstone, NSW, Australia
Posts: 1,397
Hi Pete,

Yes, have a look at short exposures for sure because that will pretty much take tracking and guiding off the table.

You mentioned synching before plate solving. Not sure why. If you have a model and you have not touched the mount as far as polar alignment, you ought to get reasonable pointing right after you home the mount. If it is not plate solving you might have too many, or not enough stars. You also might need to increase the number of fields, and of course, you need to have the resolution set exactly, though there is also a setting that allows you to set some latitude in that parameter. I have found that I can plate solve pretty easily with 5 to 30 sec exposure, but that an 8 min sub of exactly the same field will not solve.

One of the very best ways to synch, if you need to, is to take a picture anywhere, plate solve the picture, and then synch to the photo. By synching to the photo you are synching to the middle of the photo, not to a star. To do this, left click on the photo anywhere except on a star or celestial object. The drop down menu should have a synch to photo option. I might be wrong about this, but, I think if you synch every time you start up, and you don't follow the complete instructions on synching back into a model, you will seriously degrade your model and subsequent pointing. I never synch, but then I am permanently set up. Contrary to what is said in the manual, I have successfully removed and replaced the scope without seeming to hurt my model...at least not very much!

Peter
Reply With Quote
  #18  
Old 11-02-2013, 11:54 AM
pvelez's Avatar
pvelez (Pete)
Registered User

pvelez is offline
 
Join Date: Aug 2006
Location: Sydney
Posts: 1,250
Thanks Peter

I was fiddling with my PA and so doing a modest TPoint run, adjusting PA and then starting again.

My plate solve problem was trying to plate solve without a model and without first syncing. TSX wouldn't plate solve as it had no RA and Dec to work with in the FITS header. What complicated things was that if I connected Maxim to TSX and took a photo - without first syncing - Maxim would plate solve as it would take the RA and Dec from TSX anyway. Then when I had plate solved in Maxim, Maxim would add the RA and Dec to the FITS header and TSX would then plate solve. It drove me round the bend working out why some images woudl solve and others didn't.

Pete
Reply With Quote
  #19  
Old 12-02-2013, 09:32 AM
PRejto's Avatar
PRejto (Peter)
Registered User

PRejto is offline
 
Join Date: Jan 2011
Location: Rylstone, NSW, Australia
Posts: 1,397
Hi Pete,

You can drive yourself completely batty doing modest T-Point runs and then adjusting for PA. At least I sure did.... I would say, if you are fairly close, do a modest run of say 50+ points on both sides of the meridian including to the lowest altitudes you might image at, run the super model and do your best to move the mount exactly as instructed. I found it pretty easy to get well adjusted in azi but not altitude. I would run the model and it would say I'm low, then the next time say I'm too high. This would happen even following others recommendation to move only 50% of the value. This makes you completely nuts because each time you must do a complete run. Finally I got out of that cycle by using the camera method I spoke about earlier. You might be able to do it for both azi and alt but I never tried that. Once you are adjusted well enough go back and do a 200+ point run so that you can take advantage of protrack. The polar axis report should be very similar to your earlier run. If it's close enough don't touch it or you will need to start over. Once you get fast this fiddly stage you will really start enjoying this mount!

Peter

PS...you are probably doing this already, but I would always slew to my first target before getting rid of the model about to be discarded. That way the mount knows were it is a lot better than if you wipe the model and then slew. And, if you use the "synch to linked photo" method things go quite fast.
Reply With Quote
  #20  
Old 12-02-2013, 11:49 AM
pvelez's Avatar
pvelez (Pete)
Registered User

pvelez is offline
 
Join Date: Aug 2006
Location: Sydney
Posts: 1,250
Thanks Peter - that is all very helpful.

Now if only the clouds would part so I could test all this!

Pete
Reply With Quote
Reply

Bookmarks

Thread Tools
Rate This Thread
Rate This Thread:

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off

Forum Jump


All times are GMT +10. The time is now 09:26 PM.

Powered by vBulletin Version 3.8.7 | Copyright ©2000 - 2024, Jelsoft Enterprises Ltd.
Advertisement
Testar
Advertisement
Bintel
Advertisement