156 replies to this topic

[dan_hm]

WO got my Minicat back to me within a week.  Economy shipping to and from Taiwan was amazingly fast.  They didn't give me a formal report of the repairs they did - when I asked Karen Shen she just said they "used precision optical instruments for tilt adjustment and testing" and performed maintenance and cleaning.  But, my initial tests of the scope confirm they did indeed fix the severe decentering I was seeing with my ASI2400MC.  There's still tilt but I'll have to see whether it's coming from the camera and not the optical elements.  Either way, it seems to correct well enough with BXT.  Last night was very cloudy so I didn't get to do much but tonight I'm hoping to try Lambda Orionis before it sets.  

 

While this scope gets a lot of hate, a working copy is really a unique gem.  With a FF camera it offers essentially the same FOV as my old Sigma Art 105 f/1.4 lens paired with an IMX571 camera - a combination that I loved.  However, that lens had to be stepped down to f/2.8 or more to get an acceptable field.  At that f-ratio, the Minicat with my ASI2400MC is substantially faster and actually has a finer image scale.  Also, like all lenses the Sigma lens didn't have a great mounting solution and had very limited options for a filter wheel or filter drawer.  I think the Minicat will finally cure me of my addiction to camera lenses.

[kgb]

While experimenting around with potential indoor detilting methods, I found that my MiniCat 51 has 10µm-ish worth of decentering. …

Chen

Hey Chen,

Which indoor detiltimg methods are you referring to?

Thanks,
Karl

[w7ay]

Which indoor detiltimg methods are you referring to?
 

I have been trying multiple methods.  REEGO, ColliDream, in addition to using a star field on an iPad, etc.

 

First, a comment on using tilt tools like ASTAP with synthetic star fields printed on paper or drawn on an iPad (around 260 dots per inch): the paper/iPad must also be in a plane that is perfectly perpendicular to the optical axis (i.e., parallel to the plane of the lens).  Otherwise, tilt adjustments will simply make the image plane (sensor plane) be parallel to the synthetic star field, and not neccessarily parallel with the optics.  (The interaction of the three planes can be explained by the Scheimpflug principle that is usually used to explain the tilt-and-shift lenses used for architectural photography).  At night, the real stars are automatically perpendicular to the optical axis, so all you need is to get the image plane detilted relative to the object (star) field.

 

I had gone as far as to create an optical rail (like the guys doing Ronchi testing of mirrors) that uses x-y-tilt stages to hold an iPad, but still find it is really difficult to find a way to make the synthetic star field to be perpendicular to the optical axis. So I have mostly given up on that, although the x-y-tilt is still useful for other experiments.

 

So, I went next to the tools that actually measure reflection from the glass itself.  I collected a bunch of tools like the REEGO (an Italian product that you can buy from TS Optiks), and the ColliDream (a Ukrainian product that I know from your postings that you are familiar with).  Both work by throwing some points of light (white LED) at the refractor glass and watching the reflection from the glass surfaces, instead of depending on image and object planes.  The reflections back are focused by the concave surfaces.  You then try to align the different reflections so thay all line up like ducks in a row.   The main difference between the REEGO and the ColliDream is that the later has an array of LEDs that forms a cross, while the former only has 4 LEDs at just the far corner of a cross.

 

Here comes the rub.  Both the REEGO and the ColliDream are made to be used with your eyes.  They both have a small boresight, like the Cheshire tools, to help center your eyes.  

 

Since I have had cataract surgery some years ago, I can no longer refocus my eyes to the different planes that each of the concave surfaces reflect the crosses back.  To solve that, I got hold of the OCAL collimation tool, which consists of a camera whose focus you can adjust by software, and the software also allows you to set the "center" of some crosshairs and circles.  However, the boresight holes are too small for the OCAL to see the entire crosses, so I had to add a transfer lens in between the OCAL and the boresight of the REEGO/ColliDream, and whose focus is right at the boresight hole :-).  

 

OK, you probably see the problem the whole OCAL/transfer lens/ColliDream Rube Goldberg contraption creates --  that subsystem also introduces a tilt of its own.  The collective tilt (the OTA tilt + the OCAL tilt) is now a compound tilt.   Right now, I am in the process of gettiing the OCAL part detilted by adding a Baader tilt plate to the already Rube Goldberg thing, and using the scheme people use to detilt cameras (rotating and see the light spot that is reflected off the sensor) to detilt the OCAL -- but in this case, I can just point the assembly at a wall, rotate the assembly, and adjust tilt of the Baader until there is a static point on the OCAL that does not move through an entire rotation.  That point on the OCAL camera is also the optical center of the OCAL assembly.  

 

That is when I discovered that my MiniCat51 is decentered, or at least the M48 camera adapter on the MiniCat is decentered from the optical assembly that is in the inner tube of the WIFD focuser.  OCAL has a sensor with very fine sensor pitch of 1.4 µm, so you can see even small offsets.  

 

The saga has not yet ended, but it is still lots of fun :-).  Once the OCAL/ColliDream assembly is properly detilted, I can then proceed to detilt the MiniCat itself.  I am also planning to construct an accurate pinhole at the center of entrance pupil of the lens, which is an additional point that should be on the optical axis, for the ColliDream cross to center on. 

 

The ColliDream feels like it should work.  It is sensitive enough that I can see large movements of the crosses with just a tenth of a turn of the tilt adjustment screws. 

 

I am using the MiniCat51 to do this experiment because the short focal length makes it very sensitive to tilt.  If I can detilt that OTA, I should be able to use the same tool successfully on my other scopes.  The rains are not going away any time soon, so I have weeks more of cloudy nights to play :-).  

 

Chen

[kgb]

That is interesting. I encountered the same issue with the collidream, that it had introduced some tilt when used for this specific purpose. The bane of my existence is removing tilt from an optical train at night under a clear sky so I too have been searching for a method that could be done on a starless night. Thanks for the info.

[dan_hm]

So the CAA on my Minicat is AWFUL. In fact I was nearly convinced WO didn’t do anything to fix my copy until I tried rotating the CAA and the optical issues drastically improved. The angle I had it at just made everything look terrible. I would advise anyone using the scope to buy a third party CAA and just use that. There is something seriously wrong with whatever WO did here.

[Drothgeb]

I have been trying multiple methods.  REEGO, ColliDream, in addition to using a star field on an iPad, etc.

 

First, a comment on using tilt tools like ASTAP with synthetic star fields printed on paper or drawn on an iPad (around 260 dots per inch): the paper/iPad must also be in a plane that is perfectly perpendicular to the optical axis (i.e., parallel to the plane of the lens).  Otherwise, tilt adjustments will simply make the image plane (sensor plane) be parallel to the synthetic star field, and not neccessarily parallel with the optics.  (The interaction of the three planes can be explained by the Scheimpflug principle that is usually used to explain the tilt-and-shift lenses used for architectural photography).  At night, the real stars are automatically perpendicular to the optical axis, so all you need is to get the image plane detilted relative to the object (star) field.

 

I had gone as far as to create an optical rail (like the guys doing Ronchi testing of mirrors) that uses x-y-tilt stages to hold an iPad, but still find it is really difficult to find a way to make the synthetic star field to be perpendicular to the optical axis. So I have mostly given up on that, although the x-y-tilt is still useful for other experiments.

 

I’m expecting rain today, and was planning to DPAC my MiniCat. Alignment with an IPad is the same as with a flat.

 

Make a paper mask for the front of the scope that has a center point.

Using a collimation laser, adjust the tilt screws on the MiniCat so that the laser is exactly aligned with the center.

Now make hole at the center point so that the laser shines through the hole on IPad or flat. This will allow the laser beam to be reflected back on the paper.

Adjust the IPad so that the reflected beam hits the hole. Then tweak it using the collimation target on the laser.

 

I tried it, except for mounting the iPad and laser setup, it only took a couple of minutes to align them. 
 

Edited by Drothgeb, 10 April 2025 - 05:01 AM.

[w7ay]

Using a collimation laser, adjust the tilt screws on the MiniCat so that the laser is exactly aligned with the center.

Now make hole at the center point so that the laser shines on IPad or flat. This will allow the laser beam to be reflected back on the paper.

Adjust the IPad so that the reflected beam hits the hole. Then tweak it using the collimation target on the laser.

 

I tried it, except for mounting the iPad and laser setup, it only took a couple of minutes to align them. 

What I have done to study the problem in more detail is to get a web machine shop (i don't have a mill) to make me a plate that fits over the MiniCat's optical tube.  Because it is a tube within a tube, I didn't trust that the inner WIFD tube is perfectly centered with the MiniCat's dew shiield.

 

 

This is what the "back" of the disk (but for a Cat91) looks like.  There is a groove that the flange in the WIFD tube slides into, to get as well centered pinholes as I can get.

 

 

Another disk is bolted to the the main disk that have different pinhole patterns.  The one for the Cat91 above has a single pinhole, and this next picture is another variant of pinholes, forming a cross, like the ColliDream:

 

 

With these "tools", I could use an OCAL to look for tilt.  The first thing I noticed is that the OCAL's M42 flange itself can introduce its own tilt (to the tune of about 1 milliRadian).  

 

So the next thing I did was to first tune out the OCALs tilt, since the tilt vector will invariably be different from the vector of the OTA's tilt, and that confuses the tilt adjustments.  So I have monted a tilt plate (actually one of the new Askar tilt adjusters) right next to the OCAL.  To detilt the OCAL, I basically pointed the OTA at a target on the iPad (with the pinhole removed from my disk) on the x-y stage, and rotated the OCAL assembly with the camera angle adjuster of the MiniCat, and tune its tilt until the target's circular motion shrinks to a single point (like how you detilt a camera).

 

The orange and green annulus and black cross are drawn on the iPad, while the yellow cross, and red, green and blue circles are the OCAL's display.  The white disk is simply the defocused entrance pupil of the MiniCat.

 

 

Once the OCAL is detilted, I did indeed find that it is very easy to center a pinhole at the objective end.  But I haven't had anywhere close to clear skies yet to check with an image of real stars.

 

The web machine shop claims to aim for 20 µm precision, and in practice better than 100 µm, albeit, the smallest pinhole that they can drill is 1mm.  The holes appear quite large in the OCAL camera (1.4µm pixels) but probably not too bad:

 

 

One idea that I am currently testing is to check how reliable the measurements are when I place the OCAL at infinity focus, with the iPad target one focal length away from the miniCat -- i.e., the opposite of stars at infinity and camera at focal point. Basically, a collimated beam from the exit pupil of the Minicat to the OCAL.

 

BTW, when I started playing, I had thought that there was some decentering between the WIFD tube and the outer tube of the OTA (where the OTA tilt plate is bolted to is).  Turns out that most of the decentering that I thought I saw was really caused by the tilt of the OCAL.  W.O.'s collimation process must have also included centering the WIFD, because I didn't have to translate the center of the OCAL.

 

Another BTW, to make it easier to adjust the tilt screws on the MiniCat, I had incorporated this hack:

 

 

I placed small silicone o-rings under each of the "pull" tilt screws, with the "push" screws fully retracted.  The o-rings act like springs to push against the pull screw.  That way I don't need to touch the push screws while adjusting, and only tighten them when I am finished. :-)

 

Chen

Edited by w7ay, 10 April 2025 - 06:09 AM.