Odd Star Test Results WO GT71

[Adam J]

All, 

A friend has a GT71 refractor with a 0.8x reducer and images via a ATIK460EX.

He has been having issues with star shape. 

We have ruled out tracking (star shape changes in different parts of the image)

We have ruled out sensor tilt (he has tried a tilt adjuster and after many hours cant get good shape across the whole image) and in any case the distortion does not rotate with the camera. Also its been an issue since he started with the scope, used to think his DSLR sensor was tilted. 

So we though colimation was out, however a star test shows very symmetrical nicely spaced airy rings. 

So the question is what else can we try? One thing to note is that the vignetting is very off centre so we did wonder if the entire lens cell was tilted such that the centre of the light cone is not hitting the centre of the sensor?

WO has agreed to take it back and look at it, but they say as the out of focus airy disc looks good its not likely to be colimation. 

Any suggestions appreciated. 

Adam   

 

[andrew s]

Try using CCD Inspector on the images. It has a free trial period and might give you a better understanding of the issues. 

Regards Andrew 

[Adam J]

Just now, andrew s said:

Try using CCD Inspector on the images. It has a free trial period and might give you a better understanding of the issues. 

Regards Andrew 

Sorry should have said that he has been using that program, I think he is going to post some example images here later. 

[PadrePeace]

As Adam said above, I have used CCD Inspector to try to adjust my way through this problem. Here are a few shots from that research along with the Airy Ring focus images that show the triplet lens cell is well set up as a unit.

CCDI frames as I rotate the camera through each cardinal bearing starting at 360 in the top LH corner (read clockwise so top RH corner is 090 etc).

Airy Shots x 2.

NGC1333 stack showing poor star shape on LHS.

[PadrePeace]

I have a WO 0.8 Flattener Reducer 6A (so scope running F4.7 at a FL of 336mm vice F5.9 420mm unreduced) in the optical train which I would have expected to have given me a much flatter field across such a small sensor as the Atik 460EX (12.5 x10 mm). Back focus has been adjusted from the factory advised 54.8mm through 55.8mm (to allow for the Astrodon Filter correction of 1mm) out to 56.5mm with no significant improvement in star shape in the problem areas. This includes the 13-13.5mm BF built into the 460EX.

[Adam J]

4 hours ago, PadrePeace said:

I have a WO 0.8 Flattener Reducer 6A (so scope running F4.7 at a FL of 336mm vice F5.9 420mm unreduced) in the optical train which I would have expected to have given me a much flatter field across such a small sensor as the Atik 460EX (12.5 x10 mm). Back focus has been adjusted from the factory advised 54.8mm through 55.8mm (to allow for the Astrodon Filter correction of 1mm) out to 56.5mm with no significant improvement in star shape in the problem areas. This includes the 13-13.5mm BF built into the 460EX.

As I remember it when the camera is rotated the distortion moves and does not remain on the left hand side of the sensor. As such I am struggling to see how it can be tilt...although it does look like tilt.

[ollypenrice]

If you have access to a well collimated laser collimator it might be worth cutting out a paper disc to place over the objective. The pinprick from the compass will indicate the centre. Then you could put the collimator in the focuser and see if the beam hit the centre of the paper disk. This will tell you if either the cell or the rear assembly are tilted.

Olly

[newbie alert]

I use a WO 80ed.. been chasing flex, tracking errors etc for quite a while now..more or less went down the oag route until discounted one thing at a time..it's either tilt or what i think is flex in the focuser tube..certainly movement there, are you using yours with the extension, presume you are to achieve focus

[david_taurus83]

I have wrestled with this myself for a while. I have suspected the adjusting ring on the flattener might not be helping so I've wound it all the way in now and used spacers and the thin plastic shims that come with the ASI1600 to get me to 54.8mm. Last image taken with the Flat6A II below. (Been trying a Hotech SCA for a while since) 

 

Still distortion in the corners and the left edge is worse than the right. Stacking (16 x 900s) has made it better than the subs. Have you tested without the flattener to see if the vignetting is still offset?

[PadrePeace]

2 hours ago, ollypenrice said:

If you have access to a well collimated laser collimator it might be worth cutting out a paper disc to place over the objective. The pinprick from the compass will indicate the centre. Then you could put the collimator in the focuser and see if the beam hit the centre of the paper disk. This will tell you if either the cell or the rear assembly are tilted.

Olly

Olly, I have done this using a representatively (simulate the 460EX) weighted collimator firing the laser through both the filter wheel and Flattener/Reducer. I'll post an image, but in essence the red dot was within 1mm of the centre (which was a close to the centre of the OTA as I could reasonably tap it. When I rotated the laser it ascribed a slight circle within 1mm offset from paper circle cross-hairs centre throughout the full travel of the focuser.  I also did this without the optical 'gear' by mounting the laser directly into the focuser tube with very much the same results.  Thinking about that now, I taped the paper disc to the Dew Shield which was fully extended but intentionally seated against its stops. I should retry it with the Dew Shield removed and the circle taped directly over the Lens Cell to eliminate dew shield tilt.  Thanks for making me rethink that one. 

[Adam J]

9 hours ago, ollypenrice said:

If you have access to a well collimated laser collimator it might be worth cutting out a paper disc to place over the objective. The pinprick from the compass will indicate the centre. Then you could put the collimator in the focuser and see if the beam hit the centre of the paper disk. This will tell you if either the cell or the rear assembly are tilted.

Olly

How far out is too far out? Are there any maths to determine this?

[Adam J]

8 hours ago, PadrePeace said:

Olly, I have done this using a representatively (simulate the 460EX) weighted collimator firing the laser through both the filter wheel and Flattener/Reducer. I'll post an image, but in essence the red dot was within 1mm of the centre (which was a close to the centre of the OTA as I could reasonably tap it. When I rotated the laser it ascribed a slight circle within 1mm offset from paper circle cross-hairs centre throughout the full travel of the focuser.  I also did this without the optical 'gear' by mounting the laser directly into the focuser tube with very much the same results.  Thinking about that now, I taped the paper disc to the Dew Shield which was fully extended but intentionally seated against its stops. I should retry it with the Dew Shield removed and the circle taped directly over the Lens Cell to eliminate dew shield tilt.  Thanks for making me rethink that one. 

I just spent some time working out the maths for Depth of focus considering your sensor size / f-ratio and pixel size. I will post the Maths on here for future reference later but the answer is that you need the laser spot to be within 2.5mm of the centre of the lens to avoid any issues with tilt. However you may get away with more than that as there will be a difference between detectable tilt and significant tilt. 

[PadrePeace]

3 hours ago, Adam J said:

I just spent some time working out the maths for Depth of focus considering your sensor size / f-ratio and pixel size. I will post the Maths on here for future reference later but the answer is that you need the laser spot to be within 2.5mm of the centre of the lens to avoid any issues with tilt. However you may get away with more than that as there will be a difference between detectable tilt and significant tilt. 

Thanks Adam. Good timing as here are tonight’s laser collimator tests with the cross hair paper disk fixed directly to the front of the lens cell (dew shield removed).   The first image is my rig if anyone is interested. Comprises TS Optics rotator, Baader clicklock, WO 0.8 flattener/reducer FLT 6A. TS Optics Tilt adjuster, ZWO FW. 460EX. 

Next is the laser test rig mounted in the clicklock. I rotated the laser  using the TS Optics rotator during the video. 

I also tried other combos of compression ring adaptors I have including the original WO 2” screwed directly into the focus tube but the results were variable and generally worse than the clicklock set up. 

So, is this centring error  excessive? Given Adam’s calcs I would not expect it is, which clears the focus tube and rotator clicklock rig alignment as a suspect. Thoughts anyone?

[PadrePeace]

12 hours ago, david_taurus83 said:

I have wrestled with this myself for a while. I have suspected the adjusting ring on the flattener might not be helping so I've wound it all the way in now and used spacers and the thin plastic shims that come with the ASI1600 to get me to 54.8mm. Last image taken with the Flat6A II below. (Been trying a Hotech SCA for a while since) 

 

Still distortion in the corners and the left edge is worse than the right. Stacking (16 x 900s) has made it better than the subs. Have you tested without the flattener to see if the vignetting is still offset?

Dave. I’ve not been able to test without the flattener/reducer in play. The FLT 6A (previous version to yours)  is pitched as a doublet but is in fact a TRIPLET (see pic with a green laser revealing each lens face) Would this be relevant?

[david_taurus83]

Ah I see. Mine is also a triplet so its probably the same lens in a different housing. I also have a tilt adapter on there but I've not had the chance to try and iron it out, if it is tilt. I seem to get ok stars on one side and not as ok in an opposite corner and then one corner even worse again. One thing I have thought of, is it could be slight play in the drawtube and this may change angles as the scope moves. But even with slight play in the drawtube, it's a tiny lateral movement. How much could this effect the plane of focus on the sensor?

[PadrePeace]

David

Adam and I have been mulling over my issues for several weeks and tried many things. You seem to have similar star issues. Try taking a single sub (long enough to get a solid star field) and then rotate your camera through 90 degs. take another shot and rotate again. Do this to get four images and then compare where the bad stars are on each image. If the bad stars rotate (so start in BLH corner and next appear TLH corner if rotating clockwise) then your tilt is in the scope/up-scope of your focus tube/rotator. If the bad stars remain in the same corner of each frame (so in my example the BLH corner) then the tilt is rear of your focus tube/rotator. Knowing this will help you work out where your tilt adjuster needs to be in the optical train; either up-scope of the rotator/end of the focus tube or....on the camera side of the rotator/end of the focus tube. 

Hope that makes sense.

[andrew s]

It might just be the optics are not capable of "perfect" images full field.  Has anyone got this setup and achieved the required precision?

Regards Andrew 

[alan4908]

If you put the jpeg images through CCD Inspector then they clearly point to the issue being some form of camera tilt. The first image indicates a very large tilt whilst in the second one the tilt is much reduced.  However, the problem with using CCDInspector on stacked jpeg images is that it can give very misleading results. 

So, in order to eliminate focuser droop and minimize the influence of tracking/guiding errors, I'd suggest you take 6 single subs with a very short exposure with the scope pointing vertically upwards and then run CCD inspector on the 16 bit subs in averaging mode.  

I'd then repeat the test with the scope pointing towards the horizontal (this will stress out all the components). 

Using this methodology you should be able to work out if anything is creating additional camera tilt when the scope in pointed in different directions. 

I've found that to minimize camera tilt you really need to have an all screw connection set up.

Alan