Why am I seeing field rotation?

[Peje]

I have always noticed some field rotation with my images, some targets seem worse than others. I have really investigated too much about this but I suspect I see it more with high altitude targets.

Googling suggests that I have a polar alignment issue, I have recently used PHD drift align to get my alignment to under 0.5arcsecs in alt & azi. I would have thought this would have been ample. Another potential issue would be the tube rings being loose, I checked this and I couldnt move the tube no matter how hard I tried.

I have attached a screenshot of a rather poor stack that was taken over two consequtive nights, this shows the dramtic rotation. The camera was not removed in between sessions. My FoV is quite small so this means the rotation is pretty extreme.

[Davey-T]

Maybe the scope isn't lined up with the mount RA axis, what they call orthogonal or cone error I think.

Dave

[carastro]

This was the same side of the meridian?

Carole 

[carastro]

Beat me to it Dave, that's what I was thinking.

Carole 

[Peje]

Just now, Davey-T said:

Maybe the scope isn't lined up with the mount RA axis, what they call orthogonal or cone error I think.

Dave

Hmmn... how would I end up in that condition?

Just now, carastro said:

This was the same side of the meridian?

Carole 

There was probably a meridian flip on the second night.

[carastro]

I have a similar problem when I do a Meridian flip, and I have been reading up only today about Cone error, and I think I might have the same problem.

Watch this video it is quite enlightening.

Try stacking the subs on only one side of the meridian and see what happens.

Carole 

 

[Davey-T]

13 minutes ago, Peje said:

Hmmn... how would I end up in that condition?

There was probably a meridian flip on the second night.

Normal manufacturing tolerance in the components.

Dave

[DaveS]

I got a lot of this until I made sure my camera was orthoganal to RA and Dec.

Trail exposures in either NS or EW and adjust the camera angle until the trails run exactly in line with the software crosshairs, or at a pinch the edges of the frame.

[Peje]

42 minutes ago, carastro said:

I have a similar problem when I do a Meridian flip, and I have been reading up only today about Cone error, and I think I might have the same problem.

Watch this video it is quite enlightening.

Try stacking the subs on only one side of the meridian and see what happens.

Carole 

 

That certainly makes sense, typically my dovetail and rings don't have the adjustment bolts so I will have to go down the shims route. I will diagnose by stacking exposures before and after the meridian flip.

Looking at the spacing I actually think I can see a difference, that could just be me convincing myself though.

I do want to find a different method of resolving the issue as I really don't want to touch my Azimuth bolts and [removed word] up my polar alignment

[Davey-T]

Soft drink cans make good shims when cut up with decent tin snips.

Dave

[Peje]

2 minutes ago, Davey-T said:

Soft drink cans make good shims when cut up with decent tin snips.

Dave

I was planning to order a couple of sets of feeler guages and use those, that way I could actually be a bit more methodic / accurate.

[Davey-T]

10 minutes ago, Peje said:

I was planning to order a couple of sets of feeler guages and use those, that way I could actually be a bit more methodic / accurate.

Aluminium cans are easier to cut and make bigger "U" shaped shims out of and stack on top of each other, if you get to need a lot piled on top of each other you can use thicker aluminium sheet.

The only problem with shims is they're only giving you vertical movement, if you need side to side adjustment you end up filing out one of the dovetail bolt holes, been there, done that, very time consuming and can waste clear skies, best done when the Moon is out.

Dave

[Peje]

5 minutes ago, Davey-T said:

Aluminium cans are easier to cut and make bigger "U" shaped shims out of and stack on top of each other, if you get to need a lot piled on top of each other you can use thicker aluminium sheet.

The only problem with shims is they're only giving you vertical movement, if you need side to side adjustment you end up filing out one of the dovetail bolt holes, been there, done that, very time consuming and can waste clear skies, best done when the Moon is out.

Dave

Perhaps I'm being overly optimistic but I think with cast rings and dovetails I should really only be looking at tilt. I've never considered that my bolts could leave room for movements left to right though...good call

[tooth_dr]

5 minutes ago, Peje said:

Perhaps I'm being overly optimistic but I think with cast rings and dovetails I should really only be looking at tilt. I've never considered that my bolts could leave room for movements left to right though...good call

Crude....but effective. 

[Merlin66]

Hmmmm

I think side to side movement only impacts on the Dec settings, not the cone issue.

Shimming the front or rear attachment points (rings/ dovetail) is the way to fix it.

 

[ollypenrice]

This is pretty odd. What happens just on one side of the meridian? The easiest way to test is to take a first and a last sub pre-flip and combine the images based on star alignment. Then look at the frames: how aligned are they? If they are not well aligned then either your PA is not as good as you think it is or you have something rotating. The camera angle doesn't matter. (I always run aligned with RA and Dec when I can but if I need to rotate for a framing then I do and nothing changes.)

On the other hand if the misalignment only appears after the meridian then, yes, I'd go for cone error.

Olly

[Davey-T]

Snippet from my 10Micron instructions.

Guess it's only the star in the centre that's orthogonal  though.

Note the bit that says it's not necessary

Dave

 

[Peje]

A little progress,

I decided to order a new dovetail with adjustment bolts and actually realised the Skywatcher one that came with my scope is used for my guide scope. Rather than mess about swapping them over I have just ordered amother one. I know I could have adjusted using the top one but it would mean removing the guide scope and for 20quid I'd rather just wait.

For the adjustment method, I found an application called Cone Sharp which means no fiddling with my azimuth bolts or polar alignment. Ok it means using clear sky time but I'd loose that either way.

[RayD]

I'm not so sure on this personally.  I think Dion's explanation is wrong, and that "cone error" has no effect at all here.  He is doing no more than you could achieve by moving the az bolts.  All he is doing is mechanically adjusting for something that the mount would do by way of software offset (I should be pointing at X but I'm pointing at Y and this is how far apart they are).

By way of example, if I align my mount to polaris, as is shown in the video, and I then point my scope in the same general location and I then rotate my mount through 360 degrees, I will scribe the same circle at every point from the centre of the mount right out to the centre of the scope, not an egg shape as it appears to be shown.  I then shim the front of my scope by 2mm and do the same.  The circle is identical in its roundness, just bigger.

The only way to get an oval shape would be to have something moving in relation to the rotation point, which if everything is fixed solid isn't possible, so something is loose/moving.

[Peje]

2 minutes ago, RayD said:

I'm not so sure on this personally.  I think Dion's explanation is wrong, and that "cone error" has no effect at all here.  He is doing no more than you could achieve by moving the az bolts.  All he is doing is mechanically adjusting for something that the mount would do by way of software offset (I should be pointing at X but I'm pointing at Y and this is how far apart they are).

By way of example, if I align my mount to polaris, as is shown in the video, and I then point my scope in the same general location and I then rotate my mount through 360 degrees, I will scribe the same circle at every point from the centre of the mount right out to the centre of the scope, not an egg shape as it appears to be shown.  I then shim the front of my scope by 2mm and do the same.  The circle is identical in its roundness, just bigger.

The only way to get an oval shape would be to have something moving in relation to the rotation point, which if everything is fixed solid isn't possible, so something is loose/moving.

I spotted that too but assumed this is what he meant when he said about the drawing not being totally correct.

If you think of the name 'Cone Error', it is basically what you have said. We were making a circle of diameter x but the actual diameter was different. Like moving through the height of a cone.

[RayD]

5 minutes ago, Peje said:

I spotted that too but assumed this is what he meant when he said about the drawing not being totally correct.

If you think of the name 'Cone Error', it is basically what you have said. We were making a circle of diameter x but the actual diameter was different. Like moving through the height of a cone.

Yes but with imaging the cone error is pretty much irrelevant as once you have done a synch with you planetarium software, this cone error has been accounted for.

If you spin your mount through 360 degrees it makes no difference if the front of your telescope is up or down a bit, it will still spin in a circle, and as long as your PA is spot on, that circle will be just that and not an oval unless something is loose. 

[Peje]

I'm not sure I entirely agree with that, I haven't seen any setup where it corrects for the cone.

It would not be hugely difficult to do but I don't know of software that does it. It would probably need to be incorporated at the mount control level, for me this is EQMOD .

I do agree with the non-oval statement, perhaps the cone comes from being not well polar aligned.

[RayD]

When I say it has no affect here, I mean in respects to imaging and tracking.  I understand and agree that it affects goto's, but surely this is why you would star align, so the mount knows what this error is and applies an offset?

My head is hurting 

[Peje]

3 minutes ago, RayD said:

When I say it has no affect here, I mean in respects to imaging and tracking.  I understand and agree that it affects goto's, but surely this is why you would star align, so the mount knows what this error is and applies an offset?

My head is hurting 

LOL I know the feeling. Yes, i get that some mounts hand controller does an alignment which may correct but if you don't use the hand controller then I wonder what does it? I've never seen stellarium or sg pro run an alignment routine.

It would affect tracking though as the circle your scope is moving around would be taking it away from the point you were imaging. This would be corrected by guiding but I assume this would lead the the image rotating over time. 

[RayD]

2 minutes ago, Peje said:

I'm not sure I entirely agree with that, I haven't seen any setup where it corrects for the cone.

It would not be hugely difficult to do but I don't know of software that does it. It would probably need to be incorporated at the mount control level, for me this is EQMOD .

I do agree with the non-oval statement, perhaps the cone comes from being not well polar aligned.

Yes you will have a cone, but it is only relative to where you think you are pointing, and where you actually are. 

For example, forget targets, just bolt everything together and have a perfect PA, now don't do any star alignments, just turn your mount on, start imaging and tracking at whatever your telescope is pointing at, which of course could be no target at all as your mount is aligned with polaris, but your telescope isn't.  So long as your PA is perfect you should be able to image that same target all the way through 360 degrees without it moving from the centre of the camera, which according to the video it would do.

I need to do some experiments with this as I reckon I'm missing something and wrongly simplifying it.