Guiding issues with Celestron C8/HEQ5 Pro.

[RichardNewbie]

I have an issue with guiding my C8 which I cannot explain.

 

I have just started using the C8 for deep sky, previously used only for solar system, with my ZS73&HEQ5 Pro used for deep sky – and I have had good results with the ZS73 (FL=460mm).

 

Not with the C8. PHD2 (running via Ekos) shows a guiding accuracy of around 1” RA and less on Dec with either scope. That is to say the errors peak around +/- 1” on the PHD2 chart. Oddly, Ekos shows the accuracy as a bit lower on the chart displayed under the guiding tab, but that is another matter.

 

The C8 images (600s, M101) show small star trails (see pic). These become large when the images are stacked (see pic). PHD2 shows no error messages, and seems to be tracking the same star - I have sat and watched over several exposures. The length of the trails can vary a bit between images. What could be causing the consistent tracking failure? If it was just poor tracking surely the stars would be blobs not trails and PHD2 would report much greater error? I have discounted mirror shift as the problem affects all images, roughly the same. Also, M101 is near the NCP and the meridian here.

 

As a control I have shot and stacked 5s images which individually look OK (but for very bad S/N ratio) but show the trailing if 50 stacked.

Any suggestions would be much appreciated.

The setup is as follows:

C8 with FLR (FL=1300mm) on HEQ5 Pro with Rowan belts (correctly tightened), Dec backlash eliminated, RA backlash nearly eliminated – but as good as I can get – any tighter and PHD2 reports unable to move the mount.

Load around 10kgs excl counterweights. RA set slightly ‘east heavy’, Dec balanced.

Clutches set as tight as possible.

Pulse guiding.

Guide scope 50mm/176mm FL mounted forward on ADM-type bar across top of C8 using normal small dovetail, the scope held in rings fore and aft.

No noticeable flexure in guidescope mounting.

Guide camera QHY5L II (1280x960).

Attached to C8 are 0.63 FLR, then Baader DiamondTrack with ZWO EAF, and finally ZWO ASI385.

Guiding via PHD2 on Raspberry Pi via Ekos. Pi also runs INDI server for the mount, cameras, focuser, reporting to Kstars 3.5.8 on W10 remote PC via 1gb Ethernet.

Polar alignment (using Ekos) < 30” RA & Dec.

 

 

[michael8554]

Need to see the GuideLog.

"The C8 images (600s, M101) show small star trails (see pic). These become large when the images are stacked (see pic)."

Which image is the stack ? The first image seems to have the largest trails ?

Neither image shows the increase in quality you'd expect by stacking, are they actually both subs  ?

Stacking software will usually reject any subs with poor star shapes, so the stacking should fail if all the stars are elongated, or the stack image should have round stars.

Which direction, RA or Dec, is the trailing in the 600s image ?

Possibly Differential Flexture if the guiding is good - the guidecam has a large overhang, not threaded connections, and a hanging cable.

Michael

[RichardNewbie]

Hi Michael – tks for yr reply.

 

In answer to your questions:

 

The image with the long tails is a stack of 5 x 600s subs. I thought SGL would put the verbose file names in the item to show this – but it doesn’t.

The single sub is the image with the short tails.

The subs were stacked using Affinity Photo – which does not reject bad subs – DSS would probably have stacked nothing as no sub has good stars.

As for direction – too complicated for my small brain.

Flexure was my first thought – but how could the image shift between each sub? In one session I had image shift for each sub, in the same direction, over 16 subs - that is nearly 2 hrs.

Theoretically, if I took 24hrs worth of subs I would get banana-shaped guidescope supports as the guidescope and OTA images became further detached!. Which is absurd.

As for guidescope overhang, this is the same one as I use for my ZS73. I attach a 4 x 600s stack of NGC6888 taken with the ZS73 on the same mount with the same guidescope – which shows no problem even if magnified.

I am baffled - but thanks for your input.

 

 

[michael8554]

Hi Richard

So your stack is without any registering, showing how the FOV has altered over the period of the session ?

That's just due to a small PA error.

So are the small trails in the sub due to guide error in RA or Dec, or poor Collimation ?

Post the PHD2 guidelog.

Michael

 

[RichardNewbie]

Hi Michael

Affinity does not have registration - but I have attached a stack of 16 subs from DSS which does - it has done its best to align the subs.

Surely the guiding would have compensated for any PA error - which if the Ekos system is good was very low anyway?

I can only presume the trails are RA or DEC guide error - collimation is not perfect but could that would create star trails?

Log for the evening of the 23rd when 16 subs shot of M101 using the C8/HEQ5 - my log for that night indicates no other images were taken but  a few others may have been.

Tks for your interest.

PHD2_GuideLog_2022-03-23_204539.txt

[Padraic M]

Another option to look into is back focus distance from the reducer, and/or camera tilt. The star shapes in the single sub look like coma-inflicted stars.  Here's a sample from my C8/0.63 FF/FR. The back-focus is supposed to be 105mm but I never did manage to get it right. The further out your distance is, the longer the stars get.

[RichardNewbie]

Thank you

Never thought of that. I remember that back focus might be an issue when putting a separate focusser  in the optical train when I bought the DiamondTrack.

I will measure the approximate back focus and make changes - such as removing the FLR or the DiamondTrack.

I will update post with results - but it might be some days with the weather forecast! Not quite sure how this would affect the image shift, but your image looks familiar so it's a place to start.

 

 

 

[Same old newbie alert]

I work with the same sort of system, but with a few changes.

I use a oag with a qhy 290( couldn't find a guidestar with the qhy5 11)

I use a ioptron cem60

One thing that I've found out is @ long focal lengths everything is highlighted and amplified... Everything needs to be much more critical and work together.. so collimation, spacing and balance play a huge roll...

if I was to make an assumption, if the image isn't rotated  I'd say your Dec balance is out, poss abit of pa alignment ...

As a suggestion I'd say run the guiding assistant in PhD,that should tell you what the mount is doing

I use a mono cam/ filterwheel and to get the spec'd 105mm spacing required lots of messing about with different adapters but I finally achieved it only to find my stars where awful and apt was showing my focal length was out on  the platesolve software( showing 1320mm) only once I'd achieved the 1280mm spec that it should be my stars looked like stars

So I'd say it's a balance of getting things right and working together.. very different from using a widefied frac

[RichardNewbie]

Based on the advice given and gratefully received - thank you:

 

Back focus from rear FLR element was about 105mm when the Steeltrack was at mid-position (15mm from min/max position). However, focuser was removed and replaced with extension tubes and back focus to sensor set very close to 105mm from the FLR rear element. Ekos now shows FL as 1250mm after plate-solving. C8 focuser about 17 turns in from extreme a/clockwise position. This will affect the focal elngth?

EAF now attached directly to C8 focuser.

 

Collimation screws replaced with home-made thumbscrews, requiring a full re-collimation. Now collimation is slightly better than before (and much easier) and as close to perfect as I am likely to get it (at the image centre). Thumbscrews much better - even if you have to buy them.

 

Home-modified finder scope (50mm Obj, 175mm FL ) used as guide scope replaced with new Altair Astro 60mm/225mm guide scope.

 

I dismantled the HEQ5 Pro RA axis, cleaned and re-greased with the recommended (very expensive) PTFE grease, and reset backlash.  RA Axis had become progressively stiffer (‘stiction’) in the past few months– it now swings very freely allowing for much more accurate balancing. Backlash seems a little less than previously. Regret forgot to do a PHD2 mount analysis.

 

Guiding peak error however remains about +/- 1”, with RMS: RA 0.20, Dec 0.16, total 0.26, using PHD2 with the new guide scope FL.

 

Image shift between exposures and star elongation is now much reduced. However, stars still show a little elongation – and these are only 300s exposures.

Attached image is M106, 5x300s subs stacked in Affinity Photo, C8 plus 0.63 FLR, ASI183C binned x2.

 Work in progress:

PA. This is done via Ekos as I have no sight of the Pole star – I shall check its accuracy using PHD2.

PHD2 mount analysis.

I have yet to identify the direction of the elongation. Astap won’t plate solve the C8 images as the FOV is too small.

Variation in star elongation at difference exposure times.

 

What made the difference to imaging performance? Your guess is as good a mine! But not there yet.....

 Comment re SCT from Same Old..... very true. The magnification and light grab of the C8 make it worthwhile trying...

[Same old newbie alert]

I have yet to identify the direction of the elongation. Astap won’t plate solve the C8 images as the FOV is too small.

Astap will platesolve at 1280mm as I use it myself, it's fussy on star shapes thou, so focus need s to be fairly good

That glow in the center tells me your spacing is wrong, and if I remember correctly you will need to reduce it to get your focal length to increase... I'm talking of only a few mm

[RichardNewbie]

OK. I will adjust the spacing such that when focused I get 1280mm FL when plate solving (Ekos does this). I obviously do not have the correct index file for Astap - I only have H18. I will add it to my 'Work in progress'.

All I need is a clear night!

[RichardNewbie]

Further to the suggestion that back focus was an issue, as promised the following was undertaken last night, despite the poor conditions here (Northants).

 

PA to < 1’ each axis.

Steeltrack used as variable extension tube.

Back focus (using Steeltrack and then C8 focuser to regain focus) altered to get a plate solve FL of 1280.6. See attached.

Guiding by PHD2. Max error was 1.5’ each way both RA and Dec, with RMS 0.26 RA 0.23 Dec.

FITS header after Astap platesolve attached – it would solve these subs if made mono first, but not otherwise. Shows rotation of 82’, as does the Ekos platesolve.

This image of M106 (15x300s subs Altair 183C stacked using Affinity Photo with darks but no flats) shows elongated stars, but to a lesser extent than the previous C8 image of M106. However, the stars are certainly not as good as Same Old Newbie’s image.

 In the light of the work over the past month at least I can now get half-decent images (if not enlarged). But that is not good enough so I am going to buy a 700mm F7 4” APO triplet – but do not know which one yet. Any suggestions gratefully received! I normally use mono/NB (as with NGC6888 above) and I think a ‘frac would be much easier.

 

The C8 can take pics of the planets.

 

Thanks to all for your input.

 

2022.04.14_1280FL_Test.txt