Understanding my guide results

[AndyKeogh]

I have only recently stared guiding seriously  so I am still at the experimental stage of finding out what works best.

Currently I am using a Megrez 90 as an imaging scope with a Canon 50D (1.94 "/pixel) in a side by side arrangement with a Zenithstar 70 and a Lodestar v1 for guiding, all on an EQ6 and using EQMOD  and pulse guiding.

I used PhotoPolaralign for polar alignment and it said I was within 0.5 arc minutes so settled on that and started PHD calibration.

Despite the bright moon there were plenty of stars around M33 so settled on that for the nights target.

The phd output from PHDlab is below and seems to show that guiding in both axis was working  although it suggest that my polar alignment is out by 2.31 arc mins

I stacked nearly 5 hours of 3 min subs in DeepSkyStacker and plotted the resultant drift in excel which is reproduced below.

My question to the group is why is there so much drift in RA  when I thought the guide graphs were reasonably flat.

TIA

Andy

[Carina Lass]

From where I am looking, your pa is slightly off. This will cause that sort of drift over several hours.

Guiding eliminates that so yourpercieved track is fine.

[ollypenrice]

I've seen no end of threads where people say that this or that software says they are polar aligned to some very good value when the evidence of a night's imaging says otherwise. I tend to believe the evidence!

Provided your guiding is good enough to keep drift sub pixel for the duration of each individual sub the stack will be fine, though the borders will have drifted and rotated (it won't be by much) and will need cropping.

I would do a drift alignment if the setup is permanent. This is a direct and software-free method of perceving misalgnment. On a temporary setup I would just be happy to get quickly on with it provided each sub were not trailed.

Olly

[AndyKeogh]

Thanks Carina Lass and Olly.

I've seen no end of threads where people say that this or that software says they are polar aligned to some very good value when the evidence of a night's imaging says otherwise. I tend to believe the evidence!

Provided your guiding is good enough to keep drift sub pixel for the duration of each individual sub the stack will be fine, though the borders will have drifted and rotated (it won't be by much) and will need cropping.

I would do a drift alignment if the setup is permanent. This is a direct and software-free method of perceving misalgnment. On a temporary setup I would just be happy to get quickly on with it provided each sub were not trailed.

Olly

I was trying to avoid drift alignment as it usually takes me hours.    This particular session was to get some quantitive evidence as to whether PhotoPolarAlign had merit.

As an engineer I am having some trouble visualising the RA drift as down to PA error.

I have looked at the PHD debug log from that session and there were 5942 guide subs in which the guide star was found and it was always within +- 1.3 pixels on each axis.  So guiding has worked.

Would I be right in thinking that the 17 pixel drift in the dec axis is down to PA error?

That would leave me hunting for why the RA axis of the 2 cameras drifted apart by 171 pixels on the 50D ; equivalent  to 5.6 arc minutes.

The guide scope is attached via a Skywatcher Guidescope Mount. I wonder if that could be a factor?

I guess another experiment is required once I get a clear night again.

I think:-

1) I will check Polar alignment with PhotoPolarAlign  to establish a baseline.

2) I will drift align.

3) I will check Polar alignment with PhotoPolarAlign  to establish changes in its baseline.

4) I will attach an SBIG St-i guidescope directly to the rings of the Megrez 90 and guide from that and monitor what happens with the Zenithstar.

5) Collect another 5 hours of data and analyse again.

Meanwhile if anyone has any more thoughts on the issue I would be pleased to hear them.

Andy

[Carina Lass]

The graph line has a slight curve to it which shows how field rotation has more effect as it gets further from its origin.

If this is not pa, then I can only suggest that the guide scope is flexing, and the curve is caused by more flex as it moves further from its axis.This usually isn't the case as the mount gets sturdier as it reaches a max.

[themos]

A declination drift rate of 17 pixels in 4hrs and 20min is (17 * 4 arc-sec/pixel = 68 arcsec) / 260 minutes = 0.26 arcsec/minute. At a declination of 30 degrees, http://celestialwonders.com/tools/polarErrorCalc.html gives a PA error of 1.17 arcminutes.  I don't know why PHDLab reports twice that. Next time, perhaps you can repeat the PhotoPolarAlign process at the end of the imaging run and see if it calculates a similar value. The much stronger RA drift must have a different explanation, I suspect flexure (my chief worry is the mating of the camera to the OTA). I would like to see what the drift in DSS' s idea of rotation angle is. 

[AndyKeogh]

A declination drift rate of 17 pixels in 4hrs and 20min is (17 * 4 arc-sec/pixel = 68 arcsec) / 260 minutes = 0.26 arcsec/minute. At a declination of 30 degrees, http://celestialwonders.com/tools/polarErrorCalc.html gives a PA error of 1.17 arcminutes.  I don't know why PHDLab reports twice that. Next time, perhaps you can repeat the PhotoPolarAlign process at the end of the imaging run and see if it calculates a similar value. The much stronger RA drift must have a different explanation, I suspect flexure (my chief worry is the mating of the camera to the OTA). I would like to see what the drift in DSS' s idea of rotation angle is.

Actually it is 1.94 arc seconds/pixel according to astrometry.net so the error is closer to what PPA suggested .

Since the weather tonight is so bad I have brought the Megrez indoors and will check for movement as I rotate it through 90 degrees tomorrow.

Here is an image of the focussing end of my rig.

I may have to look at the guide camera as well although it is light compared to the Canon and the Borg DG-L.

Andy

[AndyKeogh]

As you can see I had lots of problems with this post... I blame it on the Chardonnay.

[themos]

I always worry about focusers that have to be racked out that much. 

[AndyKeogh]

Well the Megrez seemed solid enough so I went looking elsewhere for flex and quickly found a plausible source in the Skywatcher guidescope mount

When I applied lateral pressure to the dovetail bar that sticks out beyond the scope there was quite a bit of flex. I had to tighten up the bolts above and below the altitude adjuster and where the top plate is mounted on the rear pivot. 6 bolts in all and each required at least a quarter turn . So it appears my mount was not as rigid as the utube video suggests .

So I will try again when the skies next clear.

Thanks Themos for pointing me in the right direction.

Andy

[themos]

Hope you get more data soon!

[AndyKeogh]

The much stronger RA drift must have a different explanation, I suspect flexure (my chief worry is the mating of the camera to the OTA). I would like to see what the drift in DSS' s idea of rotation angle is.

There is very little rotation - a maximum of 0.05 degrees. I have no idea why excel duplicates the y axis values in the following graph.

Andy

[themos]

I think that proves the Polar Alignment is good, that's remarkably little rotation for 4+ hours

[AndyKeogh]

I guess another experiment is required once I get a clear night again.

I think:-

1) I will check Polar alignment with PhotoPolarAlign  to establish a baseline.

2) I will drift align.

3) I will check Polar alignment with PhotoPolarAlign  to establish changes in its baseline.

4) I will attach an SBIG St-i guidescope directly to the rings of the Megrez 90 and guide from that and monitor what happens with the Zenithstar.

5) Collect another 5 hours of data and analyse again.

Well I had another attempt on Friday night and there was some improvement.

The test rig now looks like this:-

Only change on the Megrez is the addition of the SBIG ST-i guider. On the original guidescope I replaced the 2" push fit adapter and the canon EOS interface with the Hotech SCA and T2 extension tubes; a much more rigid solution. The focuser is now racked out less than a cm

I skipped steps 2 and 3 as it was 11pm by the time I had completed step 1 but Polar alignment was reported as 0.23 arc minutes :-

I used the Pleiades as a test target and I was really surprised at how well the ST-i actually guided for 5 hours 20mins before my neighbours pencil conifer got in the way.

Again there was very little rotation reported by DSS.

and the drift from DSS converted into arc seconds was

.

So Polar alignment is still not perfect but has improved and the flex is considerably improved but looks like it is only good enough for 10min subs.

I have ordered another Hotech SCA from FLO and that should be here on Monday. I will try again and see what effect it has on the result.

I did monitor the images from the Lodestar but my analysis is incomplete. I will try a test without the Skywatcher guidescope mount. Currently the flex appears to be

horrendous :(

Andy