0.00
#...
·
·
1
like
|
|---|
|
From my experience of not being still able to properly collimate a newtonian F4 for astrophotography for an APSC sensor I presume you need better tools as said fair point autocollimator or catseye, make sure you have the primary center mark on the center, experiment the backfocus on the range of +-1.5 mm, no flex (may be try to plot eccentricity with height or perform a load test with a camera, make sure your train is threaded I do not like the compression rings) and then try to use your cmos camera as the probe following this guide: https://teleskop-austria.at/information/pdf/FN25010c-new_Photonewton_Collimation_Primer_EN.pdf |
5.57
#...
·
·
1
like
|
|---|
andrea tasselli: Hyperbolic shape also create an aplanatic field which means it's free of coma, one of the most annoying off-axis abberations to correct. |
16.18
#...
·
·
1
like
|
|---|
|
Hi, Thanks to everyone for their help on this one. I bought myself the FarPoint collimation kit and aligned the 1ary and 2ary initially with the laser and then with around 6 iterations of the Cheshire and autocollimator. Was able to get all 4 doughtnuts on top of each other (a 1/128 turns or less on the 2arys screws) through the autocollimator and triangles in the centre of circles on the Cheshire. Did this with a snug fit of the 2inch collimation tools in the focuser in both a vertical position on the bench and in 45deg to the vertical once on the pier/mount in the shed. Unfortunately I still get get decent star images across the field. The out-of-focus star images show a doughnut with an offset elliptical 2ary obtruction. To get the out-of-focus image back into any form of symmetric arrangement, I have to adjust 1ary mirror waa-a-a-a-y more than the interactions I was doing on the bench. I figure I still have to work on getting the 2ary aligned better under the focuser. And I will have to wait for the 2inch Concenter for that. Unless I am doing something else wrong. Do other people find that even with autocollimator, further significant collimation is required on sky? Thanks Brian |
10.79
#...
·
|
|---|
|
Hello Brian, A collimation on the bench should bring you very close so that you might only be left with assessing the airy disk image at focus. At your focal length, I recommend setting up an artificial star at 16x+ FLs distance. With that you can confirm that you’re very close. The procedure for your scope is that of a Newtonian from what I see in the telescopes description. So it’s all backwards: from focuser to primary mirror, i.e. once focuser and secondary are aligned, the only thing left is primary adjustment. A few ideas/tips: Check that your laser is aligned. (You never can be certain). Check if the secondary is centered w.r.t. the tube, i.e. the spider veins have the same length. For secondary alignment, place some white paper between the mirrors to block the light from the primary and also place something behind the secondary on the other side of the tube so you can see the secondary more easily. Best, Björn |
7.56
#...
·
|
|---|
Check if the secondary is centered w.r.t. the tube, i.e. the spider veins have the same length. That you should not. The secondary should appear centered under the Cheshire. The spider legs length (assuming you have one) do not necessarirly have much to do with it. |
|
16.18
#...
·
·
1
like
|
|---|
|
Thanks guys, the 2ary is centred as far as I can determine from the 1inch concenter (and I blocked the light from the 1ary/put coloured card behind the 2ary) and the Cheshire. But I suspect I may not be accurate to within the last 5% or so. The laser is aligned. Farpoint guarantee it with their laser (its built like a battleship), but I did check. Next to try out an artificial star. |
|
10.79
#...
·
|
|---|
andrea tasselli: Can you explain why? There is one degree of freedom (along optical axis of the focuser) where the mirror can be placed without changing alignment visible through the Cheshire. Of course, the spider vein length doesn’t need to be identical AFTER secondary alignment in the Cheshire. Maybe I haven’t explained myself well beforehand. Björn |
|
7.56
#...
·
·
2
likes
|
|---|
|
Björn , In the classical or standard way to place the secondary under the focuser is to have it centered as it appers for the POV of the focuser end (at EP level in fact, to minimize parallax issues). To do so you need to move the secondary away in both the direction of the mirror and from the focuser (as it needs to slide along a 45 degrees inclined plane w.r.t. the optical axis). That is, if you want to keep uniform illumination across the focal plane. This said, there are many different ways to design a spider secondary holder at any assumption about the length of the vanes is frought with risks. Personally, in answer to Brian's plea, I'd add that once the bench collimation is done the fnal registration is through a very high magnification EP with the star in focus. |
3.58
#...
·
·
1
like
|
|---|
| My experience with F5 Newtonians with full frame and APS-C cameras is that it is much harder to reach proper collimation with the full frame camera. |
|
16.18
#...
·
|
|---|
andrea tasselli: Andrea, That is really a very helpful piece of information. Just so that I haven't got this wrong, you recommend the final tweaks based on centring the artificial star in the diffraction pattern i.e. rather than the out-of-focus doughnut. This actually makes more sense, although the out-of-focus technique is more prevalent on the internet ;-) This lead to two further issues. 1) Can I get a diffraction pattern from an artificial star? Assuming I can achieve a separation of 20m [before I trespass on the neighbours, run out of garden or get blocked by trees] then the hole would have to be around 0.1mm for a 1arcsec angular size, or about 2 x diffraction limit for my 200mm telescope. Not sure if I can make a 0.1mm hole in a piece of tin foil. Would you recommend buying one of those artificial star things? The only advantage I can see is that their holes much smaller than I could produce, but I do worry about the power of the torch bean with the $2 torches they put in these things.... 2) Is a high powered eyepiece and human eyeball better than a sensor at the native focal length (800mm). Given the pixel size, the sensor pixels would only pick up the outer diffraction rings. So I suspect it may not work for this method. I wonder if my old eye are up to the task. CS Brian |
|
3.58
#...
·
|
|---|
| I have tried a pinhole in a piece of tin foil on the gate post about 20M to investigate round corner stars. I didn't manage any diffraction patterns but perhaps the holes were too big? |
|
7.56
#...
·
|
|---|
Brian Boyle: The answer is yes you can, as long as the distance is the right one. For this sort of checks one can err on the short side and assume acceptable an angular size twice that of the diffraction limit of the scope at hand. With a 0.05 mm diameter of the artificial star the distance should be for an 8" 35 meters. If the star size is 0.01 mm then the distance 7 meters and so on. I warmly recommed buying one of the aformentioned artifical star gizmos (disclosure: I have one). You have no idea how fiendishly difficult is to pierce a very small hole through tinfoil. I found the trickiest part of this business is to firmly attach the artifical star to something that can be adjusted and securely too. Hangining it from a tree branch is out of question. Eye as detector of tiny differences is still king of the hill, as far as I am concerned. Not young anymore as well but the trick is using really high power or as high as you need in order to comfortably see the diffraction rings. Normally I go 300x. Long pupil extraction is a must. |
|
10.79
#...
·
·
1
like
|
|---|
|
Yes, you can get a diffraction pattern with an artificial star. If you are working carefully, you can easily achieve a hole of 200 or 100 micron in a tin foil with a small needle. If you’re hole is too large, larger than the size of the diffraction pattern, you‘ll have problems to determine what’s from diffraction and what’s from the extended light source. If you‘re in possession of a microscope, you can shine the artificial star into the objective of the microscope to reduce the size of the star. Then you can achieve 20 or 10 micron effective size. Personally, I prefer the eyepiece (+barlow). I can assess the diffraction pattern much better visually than with a camera. Björn |
|
10.79
#...
·
·
1
like
|
|---|
andrea tasselli: I think it’s worth to add that the closer you go, the farther the image plane moves away from the scope which can introduce problems: with all extensions you might be off the optical axis. The amount of additional „back focus“ can be estimated with d = FL/(n-1) , where FL is the focal length and n is the distance of the artificial star as a multiple of the focal length. Björn |
|
7.56
#...
·
·
1
like
|
|---|
andrea tasselli: Errata Corrige: somehow a factor of about 2 got stuck somewhere in my calculation so the actual distance is 28 meter. Better, but not a lot. For a 1 arcsec angular apparent size. That should do for an 8". |
|
16.18
#...
·
|
|---|
|
Andrea, and Bjorn, Thanks for the tip on the value of those artificial star gizmos and the issues around focus. TS-Express sells an artificial star with a 25micron aperture, including a 1/4inch female screw thread for mounting on a camera tripod. I have also realised that, with the front door of my shed open, I can get a 64m line-of-sight. [If the light is bright enough]. That should give me a 0.1arcsec star about 7mm or so back from normal. It may also be useful for a star test on my RC8. I will let you know how it goes... Brian |
|
10.79
#...
·
·
1
like
|
|---|
Brian Boyle: Absolutely. That’s how I do it. I don’t have the 65 meters but only 28. However, it does suffice for the price of about 100mm of extension tubes. Since my star is located along North-South direction, I can turn the OTA to both sides an check if flexure is an issue (which turned out not to be). At this distance, you can already work with a self made star (at 20m, a 0.2mm star is 0.2 arcsec. The Airy disk of your Newtonian is about 1.4arcsec). ERRATA: it’s 2 arcsec and not 0.2. Björn |
|
7.56
#...
·
·
1
like
|
|---|
|
This got me bugged, so I got paper and pen and redid all the calcs rather than in the back of my mind. For an apparent angle of 0.7 arcsec the distance needs to be 14.73 meters for a 0.05mm pinhole, given that: distance (in meters) = diameter of pinhole/(angle in radians) = (0.05/1000)/(0.7*1/3600*pi/180) = 14.733 For an Airy disk of 1.4 arscsc that distance is halved, thus L=7.36 At 20 meters a 0.2 pinhole would subtend an angle of 2.06 arcsec, so a bit too short, as per the formula below: d/2 =L*tan(angle/2) |
|
10.79
#...
·
|
|---|
andrea tasselli: Ah, yes. Apparently I must have misplaced the decimal point. It’s indeed 2 arcsecs. Björn |
|
16.18
#...
·
·
1
like
|
|---|
|
At the risk of flooding this forum with my personal focus problems, I thought I would post my progress, in it is instructive to those facing similar problems - and to keep those who have helped so much so far appraised of how I am going. Finally managed to get some more on sky time after extended session with FarPoint autocollimator (2ary tweaks) and Cheshire {1ary tweaks). I am very impressed with the FarPoint kit (including the single point laser), solid, accurate and fits very snugly in the 2inch focusser barrel. I settled on the Running Chicken nebula, to test a few test exposures before the cloud rolled in. I attach an extra-focal, intra-focal and "in-focus" image all taken in the Ha band. Image integration times 30sec, guided with 0.5arcsec rms. Seeing poor 4-5arcsec.    The intra/extra focal images are binned x4 and the in-focus image is binned x 2, just to minimise file size. To my inexperienced eye, it looks like I am not far away from collimation. The out-of-focus image of Lamdba Can looks quite spherically symmetric to me - more so on the extra-focal image. However, what does strike me is that both out-of-focus images display a radial effect where the doughnut appears increasingly skewed (or cut?) towards the edge of the field. And it is in the opposite direction for the extra- (asymmetry skewed towards centre) and intra- focal (asymmetry skewed towards outside) images. If this was my RC I would say that it was a sign of a non-flat focal plane. If I extrapolate to my current system it might suggest that I have the wrong back focal distance to my corrector. TS advised me that the back focal distance was 61.8mm and I have 62mm worth of spacers. It might be worth investigating this. Certainly I need to do better. The "infocus" image looks horrible towards the field edge. Confusingly when I run FWHMeccentricty script in PI or run CCDInspector, it gives a result that suggests my FWHM is flat across the width of the sensor but rapidly goes off on the short axis, or a tilt of the imaging train along the short axis of my sensor. Looking at the infocus image I am not sure the fitting is actually giving me an accurate picture. But I could be wrong. A little worryingly I get a simliar FWHM plot when I use the same camera/filter wheel/OAG assembly with my Esprit 100mm refractor (where I have unsuccessfully to circularise the FWHM plot by changing sensor tilt.  My next plan is to experiment with back focal distance, and put the weird FWHM plot down to fitting issues. Grateful for any advice which might suggest that tilt and not back focal distance is my biggest issue. Thanks again Brian |
|
16.18
#...
·
·
1
like
|
|---|
As a postscript to the above, I also ran the WavefrontEstimator on the intra- and extra-focal images. It sums together the out-of-focus star images across each frame giving:   for intra and extra focal images respectively. This suggests I might have a little way to go with collimation yet. Brian |
|
10.79
#...
·
|
|---|
|
Hi Brian, The star in the center of the field in focus looks pretty good to me. A single short exposure of a bright star in focus helps to see how your collimation is. The final judgement on the collimation must be done in focus on the Airy pattern. I admit, I always had issues doing collimation with a camera so I gave up on it. I only use my eye. Then an image of a bright star as mentioned above gives the verification that the alignment is done well. It seems to me that you have some backspacing and tilt issues left. Try to add 0.5 or 1mm to see how the image looks. The cut doughnuts are simply mechanical vignetting from the entrance pupil. I haven't been using the WavefrontEstimator but I'd be concerned if the vignetting and the field aberrations (from tilt and backspacing) are giving reliable results? CS, Björn |
|
7.56
#...
·
|
|---|
|
Brian, The picture is too small on my screen to give judgement based on in focus stars. As Björn suggested the clipping of the paraxial rays will lead to the field Fresnel rings appearing cut (the cat's eye). I strongly suspect you have a tilt issue, possibly within the imaging kit. Again, I don't think this is a backfocus issue. At all. |
