Polar alignment not consistant

[philhilo]

I have a SW 200PDS newt on an HEQ5 Pro. The guiding is using the SW9x50 finder. I polar align using Sharpcap (rotate through 90 degrees) and can get it down to under 10 arcseconds. Alas when I swing the rig through 90 degrees the other way it tends to be a minute or more out. Back again to the centre and it not as accurate, and back the original way and it is again not as accurate as when it started. This suggests that the mount isnt properly aligned and I get rotation in my images which again suggests a lack of alignment.

The OTA feels solid on the mount and there is no slack in the mount itself. The finder/guider doesn't move. Is this problem caused by the OTA 'sagging' as it is rotated from upright to being on its side relative to the dovetail? Would the problem be resolved by a bigger Losmandy style dovetail and a matching replacement saddle for the mount? Or would I be better off getting a Polemaster (especially as one has come up in the used section!) or would I still have a problem? Guiding is fine. Or is there another reason behind the unrepeatable PA and rotation?

[Robindonne]

I think when you rotate the scope from a perfect aligned position, and see misalignment when returned to the first position, something in one of the axes is not tight.

I was going to suggest the problem is caused by our way of leveling the mount, because while we use almost nano technology to polaralign, our initial leveling of the mount is way less exactly done.  The finetuning used for polaralignment is basically useless when the leveling of the mount itself is done less accurate. But in your situation, returning to home position to find out you suddenly off the target, seems like some parts are moving during that rotation.  Solid ground, tripod legs, altitude bolts, mirrors etc.

[scotty38]

I get this but have assumed it's the mount not being level to the accuracy of polar aligning as already mentioned. If the mount is not 100% level then I'm assuming when it's rotated back to the "parked" position the arc it scribes will not be in the same plane as it would have been if level hence the second plate solve is "different". I usually just tweak the polar aligning again and carry on.

[Synchronicity]

If the alignment is changed when you go back to the same position then something in your setup must be moving.

There's no way that mount leveling can change the polar aligned position if moving back to the same place.

Could it be backlash somewhere?

Michael

[philhilo]

OK it looks like I need to check the level as a starter. The rig isn't on a pier but it has been in position for at least 6 months (maybe 12) sat on gravel. Cheaper than a new saddle and dovetail! I got the Polemaster so can double check and it's kit that can be moved on if I needs be (Sharpcap does a good job for £10 a year and lots of other stuff so keen to keep it).  Thank you folks. 

[michael8554]

13 hours ago, philhilo said:

Guiding is fine.

So why worry about PA ?

PA error is extremely difficult to measure, so getting repeatable readings is unlikely.

Having said that, if your SW9x50 finder is in the wobbly holder and finder shoe, that won't help.

PHD2 only nags if the PA error is larger than 5arcMINS in a Guide Assistant run, the PHD2 gurus consider that to be easily guided out.

So stick with your 10arcSECS and get guiding !

Michael

[ollypenrice]

13 hours ago, Robindonne said:

I think when you rotate the scope from a perfect aligned position, and see misalignment when returned to the first position, something in one of the axes is not tight.

I was going to suggest the problem is caused by our way of leveling the mount, because while we use almost nano technology to polaralign, our initial leveling of the mount is way less exactly done.  The finetuning used for polaralignment is basically useless when the leveling of the mount itself is done less accurate. But in your situation, returning to home position to find out you suddenly off the target, seems like some parts are moving during that rotation.  Solid ground, tripod legs, altitude bolts, mirrors etc.

This is incorrect.  A mount can be perfectly polar aligned while being as far from level as is mechanically possible. The best fast-polar-alignment routine on any mount is Takahashi's and, in their system, there is no provision for levelling the mount. It sits at whatever angle the tripod places it.

Try this thought experiment:  Take a mount and polar align it perfectly. Next, remove the finder and insert a steel shaft through the polar axis housing and weld that shaft to supports from the observatory floor. The mount is now permanently locked in perfect alignment. Now you can loosen the locknuts used for levelling and point the pier in any direction you like and the polar alignment will remain perfect. So levelling is, ultimately, immaterial.

So why do some manufacturers use it? Two reasons. 1) it sets the polar scope reticle to orthogonality with the sky (but it would have to be long way out to be discerniblly tilted anyway.) 2) It means that iterations of altitude and azimuth when using drift alignment, or variations thereof, do not interact with each other. ie An adjustment in Alt will not affect an adjustment in Az. Again this intereaction, in reality, is minor. BTW, there is precisely no reason whatever to level the mount N/S. Why would the system care about where the tilt to the equatorial angle came from? Some manufactures make equatorially tilted piers.

Whatever the OP's problem is, it is does not arise from mount levelling.

I agree that something in the mount must be moving. It could be primary mirror flop but, if individual subs are good, this is not quite so prime a suspect.

Olly

Edited March 19, 2021 by ollypenrice
Typo

[philhilo]

3 hours ago, michael8554 said:

So why worry about PA ?

PA error is extremely difficult to measure, so getting repeatable readings is unlikely.

Having said that, if your SW9x50 finder is in the wobbly holder and finder shoe, that won't help.

PHD2 only nags if the PA error is larger than 5arcMINS in a Guide Assistant run, the PHD2 gurus consider that to be easily guided out.

So stick with your 10arcSECS and get guiding !

Michael

The guiding isn't the issue, its the field rotation thats a pain! First up I have to crop more than I would like around the edges, and secondly DSS produces tartan images if you feed it steadily rotating data. This means buying something expensive like APP or learning something new like ASTAP. However better guiding is never a bad thing, especially with a heavily laden mount.

[michael8554]

3 hours ago, philhilo said:

The guiding isn't the issue, its the field rotation thats a pain!

Good point. 

I've always Dithered every frame, so movement of the subs and cropping are a norm for me.

Shouldn't you Dither too, if you're getting tartan images ?

Michael

[philhilo]

I do dither, but I started to get this tartan pattern appear  after stacking that would be visible in the stretched image.  Consultation  with the learned folks of Stargazer Lounge showed i had a steady rotation in my images. When DSS stacked these the pattern appeared, bilinear interpolation I think. It was suggested I try different stackers and sure enough the pattern vanished.

Thus I was trying to understand why I wasn't getting a consistent pa and needed to get one despite the decent guiding. Long story but fixing problems in astrophoto usually is!

[ian_bird]

On 19/03/2021 at 00:11, philhilo said:

Would the problem be resolved by a bigger Losmandy style dovetail and a matching replacement saddle for the mount?

I would say yes. You need everything to be as rigid as possible. And the Vixen rails wobble - badly.

I stuck a Losmandy saddle on my old CG-5 mount and was churning out 30 minute subs quite happily. Though that was using a 3nm Ha filter - which is a bit forgiving.

Cheers.

Ian

[philhilo]

Thanks everyone for the feedback on this, life all got a bit hectic and I had to step away for a while. My mount has gone off for a tune and belt mod so hopefully it will come back much improved (the guiding went right off in April, don't know why, but the summer hiatus seemed like a good moment for modding etc). I may well get an EQ6R Pro if they ever arrive in the country again!

 

[iapa]

On 19/03/2021 at 01:11, Robindonne said:

I think when you rotate the scope from a perfect aligned position, and see misalignment when returned to the first position, something in one of the axes is not tight.

I was going to suggest the problem is caused by our way of leveling the mount, because while we use almost nano technology to polaralign, our initial leveling of the mount is way less exactly done.  The finetuning used for polaralignment is basically useless when the leveling of the mount itself is done less accurate. But in your situation, returning to home position to find out you suddenly off the target, seems like some parts are moving during that rotation.  Solid ground, tripod legs, altitude bolts, mirrors etc.

Levelling isn’t overly important once you have an accurate PA.

I’m wondering if the mount is not in a firm surface and is sinking, or perhaps an extended leg is slipping ?

Sitting on gravel is not particularly solid as a surface.

[Robindonne]

37 minutes ago, iapa said:

Levelling isn’t overly important once you have an accurate PA.

I’m wondering if the mount is not in a firm surface and is sinking, or perhaps an extended leg is slipping ?

Sitting on gravel is not particularly solid as a surface.

No correct. I might have been thinking wrong all the time.  I assumed when the mount would hang a bit to the left or right side of the polar scope but still polaraligned, it would have to correct much more during guiding.  But i realise the 3 staralignment or platesolving does the initial correction thats needed for smooth guiding