Help! Very bad corner stars with FSQ85

[HenryNZ]

I would appreciate some help with this please!!
 
I have just taken possession of a new FSQ85. I took it for a test tonight and the result was really really bad with distorted stars in 3 out of 4 corners. At the worst corner the stars are more like lines. I am very very disappointed with it especially considering the premium one have to pay to get a FSQ85 compared to other refractors of similar aperture.
 
With some experimentation, I could eliminate about 80% of the problem by focusing using stars at the worst corner. When I do that stars at the other three corners also get rounder as well and the centre stars are not too bloated. BUT - this is really not ideal especially if one wants to use auto focus routine in SGPro which will take the entire field for focus evaluation.
 
The fact that the problem can be mostly corrected by focusing using stars at one corner suggest to me that the problem is tilt, most likely at focuser or CAA. I wonder if this is something that can be adjusted or whether it is just a design flaw? I read others needing to upgrade their focuser to third party product such as Feathertouch. However to do this will bring the total cost ownership of this already expensive refractor to an even more astronomical level (puns intended).
 
What should I do?
 
PS., I am NOT using any reducer in the set up - I am operating at f/5.3.
 
PPS - I will be visiting the States in the near future. If I need anything to make this problem go away now is the best time to buy it/them to avoid shipping / import duties. So please do let me know soon if I need to buy something to make this problem go away.
 
Thanks
 
Very disappointed,
Henry

[swag72]

To start considering tilt I'd be looking at taking some short subs with your scope pointing directly up at zenith. In that position there will be minimal weight from gravity on the focuser and if that is where the tilt is creeping in, that should give you much better stars in the corners.

[HenryNZ]

Thanks. Test was done near zenith at about 70 degree elevation. I also tried rotating the CAA and the elongation changed shape but not eliminated. The best I could get was to focus using star at the worst corner.  I tried subs from 5-60 seconds with or without guiding with same result.

[Stu]

Henry, I can't help at all, but perhaps posting some sample images would help people advise you.

[Owmuchonomy]

..and to add to Stu's comment, what size chip are you using?

[HenryNZ]

I have tried with both crop sensor dslr (canon 100d) and an atik 460 which has a smaller sensor. Understandably it looks worse with the DSLR as the sensor is lager

 

 

[Owmuchonomy]

So assuming you have used Sara's suggested protocol with the same result then it is down to experimenting with spacing to get the optimal chip distance and reduce or eliminate the problem.  A look at your images may help us a lot.

[swag72]

33 minutes ago, Owmuchonomy said:

So assuming you have used Sara's suggested protocol with the same result then it is down to experimenting with spacing to get the optimal chip distance and reduce or eliminate the problem.  A look at your images may help us a lot.

In the OP Henry said that he wasn't using the reducer, so with the FSQ85 unreduced distance is irrelevant

[steppenwolf]

Spacing isn't an issue with this telescope used at its native focal length so tilt is the most likely cause of this issue.

Sara's test needs to be carried out by pointing at the Zenith so that 70 degree elevation isn't a fair test.

There are several potential areas of tilt that need to be explored:-

1. the focuser tube (likely)

2. components between the focuser and the camera - the CAA has been known to be out of true and can be adjusted (likely)

3. the sensor within the camera (unlikely)

4. misaligned optics (possible but fairly unlikely) - shouldn't happen at this price break but it can

Unfortunately, from my own experience, buying premium is no guarantee of premium performance and I decided in the end that I wasn't going to continue acting as an unpaid R & D tester but this telescope is capable of great results if you can sort out the tilt issue so it is worth persevering for a while yet.

[Owmuchonomy]

2 hours ago, swag72 said:

In the OP Henry said that he wasn't using the reducer, so with the FSQ85 unreduced distance is irrelevant

Oh, I see. Thanks.

[ChrisLX200]

How is the camera attached to the scope? All-screwed connections or are you using a 2" nosepiece? Some pictures would help...

Take an image, then turn the camera (within the focuser) 180deg and take another. If the camera (or sensor within the camera) is tilted with respect to the optical axis then the opposite corners will be affected.

If the scope has a manual rotator (i.e., you can loosen a screw and turn the camera whilst it is still locked into the back of the focuser), then repeat the above operation using that - leave the camera locked in place but use the rotator to turn 180deg. This will tell you if the problem lies within the rotator mechanism.

If it is the _same_ corners that are affected (taking into account you have turned the camera 180deg remember!) using both methods then something within the scope itself is misaligned.

Those are just quick checks, if none of it makes sense then just return your scope to where you bought it and have it checked out.

ChrisH

 

[HenryNZ]

Sara and other

can you please send me the test procedure I will try again and post some results here. Thanks

 

henry

[ollypenrice]

Chris beat me to it. Sadly it seems that Tak have QC issues with this scope. I suspect you will end up sending it back, but try the tests first. Your chip sizes are well within spec for this instrument.

Olly

[gnomus]

I had problems with my FSQ 85 and had to send it back in the end.  If you want to read my sorry tale, then ... https://stargazerslounge.com/topic/255860-ovoids-with-mesu-and-tak-fsq-85/

I would strongly suggest trying Sara's idea of shooting a short exposure (30 secs or so) straight up at the zenith so that the possibility of focuser tilt is reduced or possibly even eliminated.  If you still get the issue then that will tell you a lot.  I see you have tried two cameras.  Don't let your dealer try to claim that this is tilt in your camera set up.

Good luck

[HenryNZ]

Hello

 

I have done the Zenith test with the camera rotated through 0, 90 and 180 degrees taken at 5, 10, 15 and 20 second exposures. With each rotation the focus changes so much that I need to refocus, so I don't know if it is still a valid test. I found that at 0 and 180 degree there are oblong stars, whereas at 90 degree rotation the stars are almost good enough that I can accept. However I cannot reproduce it trying to go back to it from 180 degree. I have also done some test removing the CAA, and the obligation is still there so it is not the CAA.

 

I have uploaded the images as a ZIP file on dropbox.

https://www.dropbox.com/s/mg1kuix8bw6kj8o/FSQ85Test.zip?dl=0

 

I have also emailed the seller but he is overseas, so not sure how it will be received  So in the mean time any help will be appreciated.

 

Henry

[gnomus]

Hi Henry

I had a look at some of the images you posted - in particular your 20 second tests at 0 and 180 degrees (which you said were the worst).  

Firstly, I agree that your corner stars are not round.  I would suggest that you look at a number of pictures taken with the FSQ85 on a site like Astrobin (you can search out images by scope make and model there).  These images will not usually be presented at 100%.  I think you will find that some degree of elongation (in one or other direction - yours appear to be radial) is not uncommon for this scope - I certainly saw it in a number of these less-than-100% images.   I am no expert, but I have also seen a number of full-size images from these FSQs and I would say that you have a reasonably good example of the marque - it is certainly much better than mine was.

Many folks will say that these stars will 'round' out during the stacking and processing (although presumably they will all be slightly larger than they needed to be as a result of the elongation).   You might also be able to improve the corner shapes by focussing on a star that is towards a corner (around a third of a way in from the edge) rather than focussing on a central star (I wonder if this is why your issue varies as you have been refocussing between shots).  I received this advice from Takahashi.  However, I was using automatic focussing and I considered that if I needed to resort to this then my central stars would not be in precise focus.

I do have to say that Takahashi promise a flat field and that is what I expected.  I was extremely disappointed in my scope and returned it (after a short battle with the supplier).  But as I say it was much worse than yours.  I see that you got it from 'overseas' and I wonder, from what you write, if it was a private sale.  This may make things a bit more 'awkward' for you.  However, as I said above, I think many folks would be happy with the the results you are getting and if you decide that you cannot live with them, you would probably be able to sell this on without too much difficulty.  

I hope this is helpful.

Steve     

[ollypenrice]

I rather agree with Steve. I thought the stars were distinctly better in the longer exposures and I think they will probably round out in stacking. On the face of it I'd just get cracking and take some images.

What's the camera here? And have you tried it in other optics? It just seems odd that stars should have a pixelated shape on one side, as they sometimes do. I don't recall seeing that before.

Olly

[HenryNZ]

Thanks Steve and Olly,

The pixelation is likely a JPEG compression artefact.

I bought this telescope while overseas in Hong Kong. The price was good but after sale service as expected is next to nil.

Therefore, returning it is not really an option.

I am faced with three choices:

1. Accept the defect as is. However I am a bit disappointed. If I accept that I might as well get myself something 1/3 the cost. The other problem is, to use the scope as is, I will almost certainly need to throw more money in in the form of a motorised focus. This is because it takes me a long time to achieve that "best focus" where the aberrations in the corners are minimised. Without a motor focus and with these defects I do not think I will be able to use it practically.

2. Sell it as is. As I got it quite cheap my loss should not be too much, but it is a loss nonetheless.

3. Try to fix it. I am not talking about collimating it myself, but more along the line of changing the focuser, perhaps to something that can be collimated e.g. a Moonlite. If I can be certain that a replacement focuser can fix the problem, it may not be so bad and I would be happy to spend the extra money since if I am to keep it i will need a motor-drive anyway. However it would money down the drain if the aftermarket focuser does not fix the problem.

Therefore, I would like to seek help in isolating the source of miscollimation / tilt.

I have performed more tests tonight. This time I locked the focuser tight once the best compromise focus point is achieved, so that I can rotate the CAA without having to refocus. I have attached the images below. One set of test is done at Zenith but the stars are a bit sparse and moonlight get in the way; the other set is around 60-70 degree elevation and the stars are not so obscured by the moon. For uniformity they are all 20 second subs unguided.

Can you please look at these latest test results, and help determine where exactly the tilt is please? It will really help me decide which way I am going to go.

 

60 deg elevation, CAA knob at 12 o'clock:

60 deg elevation, CAA knob at 3 o'clock:

60 deg elevation, CAA knob at 6 o'clock:

60 deg elevation, CAA knob at 9 o'clock:

Zenith, CAA knob at 12 o'clock:

Zenith, CAA knob at 3 o'clock:

Zenith, CAA knob at 6 o'clock:

Zenith, CAA knob at 9 o'clock:

 

 

 

 

 

 

 

 

 

[HenryNZ]

Here are the eccentricity plots for the tests done at 60-70 degree. It is quite telling:

From CAA knob at 12 o'clock:

To: 3 o'clock:

To: 6 o'clock

To: 9 o'clock

 

However, at Zenith, the plots look different:

12 o'clock:

3 o'clock

6 o'clock:

9 o'clock:

Does it mean that the tilt is behind the CAA (camera), at the CAA or in front of the CAA (i.e. focuser and OTA)?

Does it mean that there is focuser sag, since being at Zenith eliminates the tilting axis?

 

 

[gnomus]

Have a look at your first image.  In the top corners you will see that the elongation is in different directions.  To my eye, in the Top Right it is NW to SE, in the Top Left it is NE to SW.  That is why I say this is radial elongation and not Tilt.  I don't think this is as 'simple' as a tilt issue - and I don't think tilt is a simple issue.  I very much doubt that getting a Moonlite focuser would help.  In fact, I think it would be a step backwards from the good quality R&P focuser on the Tak.

Focusing in the corner:  get a Bahtinov mask.  Place a bright star (like a typical 'alignment star') in one of the corners of your image.  It is best to choose a star close to your final target.  Get the best focus you can.  Lock it down and slew to the target.  You will have to slew back and forth to the 'focusing' star every time you want to focus (like following a filter change or if the temperature drops significantly).  The other alternative is to use your capture software's focus routine (such as the FWHM method) and just pick a star in one of the 'thirds'.  I would hazard a guess that this is what most folks do.

I would follow Olly's advice - take a number of images and try stacking them.  You might find you are happy with the results.  If not, then move it on and get something else. 

   

[ollypenrice]

Whatver you do, don't buy a Moonlite for imaging. It's a fundamentally flawed design with insufficient grip between the roller and the drawtube. It is entirely unsuited to motorized focus whch works by counting steps, so slippage is a royal pain. At best we can say that we can get ours to work but it would never ever be a first choice. Go for a Feathertouch. (I use one as well as my pair of Tak focusers and the Moonlite.)

Olly

[Owmuchonomy]

Those test shots are miles better than my 4" AA APO was and I could significantly reduce the effect by stacking images. Have you tried stacking your subs to see what result you get. You may be pleasantly surprised.

[opticalpath]

4 hours ago, ollypenrice said:

Whatver you do, don't buy a Moonlite for imaging. It's a fundamentally flawed design with insufficient grip between the roller and the drawtube. It is entirely unsuited to motorized focus whch works by counting steps, so slippage is a royal pain. At best we can say that we can get ours to work but it would never ever be a first choice. Go for a Feathertouch. (I use one as well as my pair of Tak focusers and the Moonlite.)

Olly

I'm sorry you had such a bad experience with your Moonlite, Olly.  I must say I'm pretty happy with mine. It's carrying quite a heavy QSI camera package and gives repeatable, slip-free focusing with my DIY stepper motor belt drive. 

When I set it up at the beginning, it did need some adjustment of the tension screw - enough to lift the imaging train (plus a little extra weight for contingency) vertically without any slip, but not so much as to stall the motor drive.  That was quite a long time ago and I haven't had to touch it since; it just seems to work every time and I don't have to think about it.  The digital focus position moves a little as expected with temperature changes but otherwise seems rock solid.

In fairness, with my (Mak-) Newtonian focus position, in practice the focuser is not often required do anything approaching a vertical  lift, so it's not tested so severely in that respect as it would be in a refractor.

I don't think anyone would ever go wrong by choosing a Feathertouch focuser, but I thought it was only fair to report one positive experience with a Moonlite.

Adrian

[HenryNZ]

The way I see it is, if the tilt or mal alignment is after the correcting element, It should be remediable with a better focuser or some shimming of the focuser. If the misalignment is ahead of the corrector towards the objective then it is no good and will need factory adjustment. My question is, how can one tell? Correct?

[HenryNZ]

I have eventually sent the telescope back to Japan factory for recollimation. They determined that a spacer has come loose. I have received the recollimated scope back and ran some more tests, but it seems there is ongoing problem with elongated stars at the corners. Unlike before though the elongation is now symmmetrical at all four corners.  I have emailed the factory again and they replied that it is performing to factory specification and merely suggested focusing between centre and edge. Needless to say I am not very happy with that. 

 

Also so I should add that the test images I posted above are from a small chip size CCD (atik 460).  I have since get myself a DSLR and the problem is distinctly worse with its larger chip size. 

 

Not it sure what to do now.