Variation in star shapes across one imaging run - focuser flexure?

[AMcD]

Yesterday evening I did a four-hour imaging run with my new QHY268M.   The initial frames were not successful due to a cable snag, which I eradicated.  Guiding was also initially an issue due to DEC backlash, but I addressed that also.  However, the images below represent the beginning, middle and end of the run.  I am almost certain there were no more cable snags and guiding was around RMS 0.60 for the session.  Notwithstanding that, the stars seemed to move from round(ish), to 'doubled' to 'kidney bean' shaped over the imaging run.  Does anyone have any ideas on a possible cause.  I am wondering whether this is flexure at the focuser as my QHY268M and filter wheel is considerably heavier that my QHY8.  However, the frame with the doubled stars perhaps does suggest a cable snag that came free during the exposure.

[Stuart1971]

Hello,

‘I would say that’s deffo cable snag or something holding the axis back from moving as it should, but not flexure, that would be far more subtle…the other thing would be well out of balance as I have experienced this years ago and it was down to balance….or not having some east side bias.

 

[michael8554]

I straight-away thought PE, but as you said, the middle image has stars that are separated, not linked.

Diff Flexture, the focuser sagging or guidescope flexing seems a possibility - is it me or does focus look off ?

"guiding was around RMS 0.60 for the session."

The GuideLog would show if guiding was good for every exposure.

Michael

 

[alacant]

Hi

At that focal length (guessing around 1000mm?) any wind, people (and cats) walking near the mount and mechanical imperfection no matter how slight, will cause odd shaped stars. That you have differing star shapes on different frames suggests a combination of factors, so it's unlikely to be only the one you suggest; The mount and associated mechanics -e.g. camera-focuser, dovetail and guide telescope connections have to be better than perfect to give your guiding software a fair chance.

The good news is that imperfections such as these will be minimised if you have sufficient frames and use a good clipping algorithm to stack them. It's also a good idea remove any outliers before stacking.. Be ruthless! To help with the latter, Siril has a plot making it easy to de-select frames.

Finally of course, stars shape (star shapes?) can be corrected using the psf algorithms present in modern astro-processing apps, some better than others. Just be careful not to overdo it.

Cheers and HTH

 

 

Edited September 22, 2022 by alacant

[ollypenrice]

I'd have a check for backlash, which would certainly explain the middle image. (The scope may have been sitting on one side of mesh for a majority of the time and on the other for a minority of the time, producing different sizes of doubled stars.)  If you do find backlash you can fix it mechanically or follow these tips:

For RA backlash, run the mount out of balance in RA so the east side is heavy. That will keep the drive pushing and won't encourage the payload to oscillate across the backlash. Reverse the imbalance after a meridian flip.

For Dec, run slightly polar misaligned and disable one Dec guiding direction so it only corrects for the misalignment (identified by experiment and reversed after a meridian flip.)

Fixing it is, of course, best.

Olly

Edited September 22, 2022 by ollypenrice
typo

[Stuart1971]

20 minutes ago, ollypenrice said:

I'd have a check for backlash, which would certainly explain the middle image. (The scope may have been sitting on one side of mesh for a majority of the time and on the other for a minority of the time, producing different sizes of doubled stars.)  If you do find backlash you can fix it mechanically or follow these tips:

For RA backlash, run the mount out of balance in RA so the east side is heavy. That will keep the drive pushing and won't encourage the payload to oscillate across the backlash. Reverse the imbalance after a meridian flip.

For Dec, run slightly polar misaligned and disable one Dec guiding direction so it only corrects for the misalignment (identified by experiment and reversed after a meridian flip.)

Fixing it is, of course, best.

Olly

I do this guiding in one direction only, for my EQ8 in DEC, and it works very well….👍🏻

[AMcD]

1 hour ago, ollypenrice said:

I'd have a check for backlash, which would certainly explain the middle image. (The scope may have been sitting on one side of mesh for a majority of the time and on the other for a minority of the time, producing different sizes of doubled stars.)  If you do find backlash you can fix it mechanically

Many thanks to everyone for your advice and suggestions.   I really am most grateful.

Olly, the Guiding Assistant in PHD2 indicated last night that I had a very significant problem with DEC backlash and having checked the play in the RA axis, it feels greater than it did when I last adjusted it.   I have recently installed Losmandy one piece worm blocks on my G11, which make adjusting the worms much easier.  I will look to see whether I can better seat the worms as a first port of call and then will try your tips.  

Whilst I dealt last night with an obvious cable snag before taking the images above, I think I am going to have to consider the cabling again.   My set up is remotely operated and I run all the cabling on the scope to a Pegasus PBA.  From that I run a pretty heavy USB3.0 cable and a 12 volt power cable.  The cables from the park sensors on the mount also drop to the floor.  At present, they are fixed at one point to the scope ring and then drop straight to the floor.  The wires are however, quite heavy between them and might be pulling on the scope.  I suspect it will be better for me to have two drops, the first to the top of the pier, with sufficient play to allow the scope to move fully in RA and DEC, and only then a drop to the floor. 

I suspect my old QHY8 allowed any number of quiet issues with respect to backlash, flexure and cabling issues to pass unnoticed.  It may be the QHY268M is revealing a multitude of sins 😀

Many thanks again to all.

[AMcD]

8 hours ago, michael8554 said:

Diff Flexture, the focuser sagging or guidescope flexing seems a possibility - is it me or does focus look off ?

Many thanks.  I think the focus is OK on the full frame subs.  I was using autofocus in SGPro, which is generally pretty accurate.  In addition to the steps I set out above, I will do a thorough check of the guide scope fixings and focuser with respect to possible flexure.

 

[ollypenrice]

36 minutes ago, AMcD said:

Many thanks to everyone for your advice and suggestions.   I really am most grateful.

Olly, the Guiding Assistant in PHD2 indicated last night that I had a very significant problem with DEC backlash and having checked the play in the RA axis, it feels greater than it did when I last adjusted it.   I have recently installed Losmandy one piece worm blocks on my G11, which make adjusting the worms much easier.  I will look to see whether I can better seat the worms as a first port of call and then will try your tips.  

Whilst I dealt last night with an obvious cable snag before taking the images above, I think I am going to have to consider the cabling again.   My set up is remotely operated and I run all the cabling on the scope to a Pegasus PBA.  From that I run a pretty heavy USB3.0 cable and a 12 volt power cable.  The cables from the park sensors on the mount also drop to the floor.  At present, they are fixed at one point to the scope ring and then drop straight to the floor.  The wires are however, quite heavy between them and might be pulling on the scope.  I suspect it will be better for me to have two drops, the first to the top of the pier, with sufficient play to allow the scope to move fully in RA and DEC, and only then a drop to the floor. 

I suspect my old QHY8 allowed any number of quiet issues with respect to backlash, flexure and cabling issues to pass unnoticed.  It may be the QHY268M is revealing a multitude of sins 😀

Many thanks again to all.

Al, the favoured cable routing among my robotic clients is to gather all the cables and run them to the top end of the counterweight bar and from there to the ground. This minimizes the variation in cable position and the lifting moment.

Olly

[AMcD]

30 minutes ago, ollypenrice said:

Al, the favoured cable routing among my robotic clients is to gather all the cables and run them to the top end of the counterweight bar and from there to the ground. This minimizes the variation in cable position and the lifting moment.

Olly

Many thanks Olly, excellent tip.  I feel a re-cabling day coming on for Saturday...

[david_taurus83]

Don't accept a decent RMS value in PHD as everything is going good. All it takes is one big 2 or 3 arc second spike in either axis to ruin a sub. Check your peak to peak values as well.