Kevin_A I just opened up my Rokinon 135 and it is more difficult if the lens like my Nikon version has the electronic aperture feature.

I hadn't thought of the electronic focus! This is not encouraging :-).

Fun stuff!

Absolutely! This is what the Smart Scope folks don't realize. This hobby offers more than just TikTok snapshots.

By the way, UPS just delivered my bearing holder (this replaces the plastic pillow bearings that I was holding up by bolting it on an Astrodymium ring :-).

It took the web machine shop in Seattle (frontpanelexpress.com) one day to turn it around, and like CloudBreakOptics, also in Seattle, UPS Ground is vitually overnight. It is almost like real time.

Shown here are the raw (top) machined part and the assembled part (bottom). Bearing is a 5mm bearing with a flange. The body is 11mm in diameter, and the flange is 12.5mm. So, I got the shop to drill a 11.03 diameter hole, and concentric with it, a 12.55mm blind hole that is 1mm deep. The bottom one has the bearing pressed in (perfect match, I lucked out in correctly guessing the extra 0.03mm clearance -- just a tap with the T handle of a hex tool got the bearing flushed in place. I added a little metal superglue to hold it there. That 4mm hole is just in case I need to brace it, but the main force will be sideways (on the plane of this photo) from the belt tension, so I don't think I will need a brace. Just an insurance.

This, and the EAF will now be mounted on a common floating plate, so the bearing the the EAF moves as a single unit. The floating plate will then be mounted (with long holes for tensioning the belt) on top of the main plate (you see me using a large plate to hold my lenses and OTA all the time) that also holds the Losmandy dovetail and rack handle.

I have this long weekend to draw up the floating plate and the base plate and send it to the machine shop to fabricate. I wanted to make sure the bearing plate above works out first -- this is the first time I am asking the shop to edge machine the edge holes (the two M2.5 mounting screws above). Stay tuned.

Yep, fun stuff. I am not a mechanical engineer, but an electrical engineer, so this is definitely "fun stuff." I'd worked on both circuit design in my younger days, and developing signal processing, image processing and color science algorithms in grad school and after.

Chen

Kevin_A Not sure if yours has it… contacts on end..

The Rikonon Cine/EOS has no contacts. Appears fully manual.

Chen

Kevin_A Just be aware that these lenses are mostly plastic

Not just plastic, but the "thinnest of plastic" :-)

https://www.cloudynights.com/topic/837977-rokinon-135-lens-repairupgrade/

If I plan on destroying the lens just to educate myself anyway, it won't be disappointing if I can't put it back together afterwards. I have never taken apart any OTA either, for fear of losing collimation.

Thanks for the warning. I shall try not to torque it too much, in case I go through with it.

This guy's lens did not survive :-)

https://www.dpreview.com/forums/post/62158943

Did you actually remove enough to get to the iris? If so, is it sufficient to remove just from the back screws?

Appears to be possible:

https://www.flickr.com/groups/365610@N21/discuss/72157625882212905/

Hmmm, I should at least remove the infinity focus stop of the lenses (including my current Rokinon) so that the EAF don''t hit a hard stop (especially when using ASIAIR "autofocus"). To be safe, I have set the EAF up so that the hard stop corresponds to 0 EAF step size. But the interesting thing is that with the 13.5mm camera angle rotator (instead of a "proper" 15mm spacing), infinity focus is actually at 30 ft instead of infinity, so I am extra safe. (BTW, the camera angle rotator has the Blue Fireball badge, which is the Agena house brand.)

Chen

Aha, the 30 ft focus point on the lens to achieve infinity focus makes a lot of sense...

If we take the simple lens equations (high school Physics), and compute the backfocus of a 135mm lens when it is focused at 10 meters (30 ft, thereabouts), you get 136.85mm. That's 1.85 mm behind the image plane.

If you remember, from the EAF delta curves (post #221 in this thread), the best backfocus for my Rikonon occured with a 1.25mm glass (i.e., pushing the backfocus 0.4mm away from the Rikonon), when using a 13.5mm thick camera angle rotator instead of a proper 15mm worth of spacing. So there is 1.5mm from the spacer, and 0.4mm from the actual backfocus using the zero crossing of my post #221. That's 1.9mm. The simple lens approximation (previous paragraph) says I am 1.85mm off. Yeehaa! Can't get closer than this with back of the envelope calculations :-).

Now, this means that when I use the actual ZWO EOS drawer (that looks like an extra 15mm spacer compared to the ZWO EOS sans drawer + camera angle adjuster), there is no way to get infinity focus that gives proper backfocus!

Adding shims will make matters worse (thank heavens for the 13.5mm camera angle adjuster, otherwise I would be tearing my hair apart looing for the proper backfocus, since you need shims with negative thicknesses, or filter glass of the order of 6mm thick!)

You don't need to use the Blue Fireball camera angle adjuster, of course. The ZWO EOS sans-drawer adapter + 13.5mm + filter thickness/3 would have given the correct result (i.e., a 14.15mm spacer would have worked for a 2mm thick filter).

The camera angle adjuster serves two purposes for me, (1) provides this "negative 0.85mm" shim, and (2) allows me to adjust camera angle with the lens fixed in its tripod ring, and therefore no fiddling with belts.

Because it is an Agena house brand, it won't be easy to find this camera angle adjuster elsewhere. But any spacer wil work if you don't care about camera angle. On the other hand, since it is a house brand, it is usually in stock at Agena, and don't become out of stock for long.

How others can achieve correct backfocus using the ZWO EOS-with filter drawer adapter is beyond me. Do hobbyists really accept a backfocus that is 0.85mm off? Or are they pushing the focus ring way past the infinity focus mark?

Chen

Kevin_A I could not get to the iris as the electronics cable was just too short to fully remove the mechanisms and end plate in front. Just the way it was trying to flip it out n over was too short a ribbon cable.

The Canon version of the lens may not even have the cables. OK, I will definitely get the Samyang non-Cine EF mount lens to play with it.

Other than the cable, can you give me the "least amount of unscrewing" to get to the iris? Can I do it compeletely from the back and not have to remove the front glass elements?

BTW, the infinity lock of lenses usually hides under the rubber focuser grip. The Cine version (at least of the Rokinon) does not have a rubber grip:-). I suspect that the Cine version (follow-focus focuser ring) does not have an infinity lock. The Rokinon is all belted up right now, and I am too lazy to find out, but will do the next time I remove the belt.

The reason to place a circular iris at the right place is that the front filter mask can give you aperture vignetting (notches in the diffraction pattern) unless you use small apertures. An internal iris will not have that problem even at f/2.4. And if it is circular, there won't be spikes.

To get an idea what a circular iris might cost, I did a fake project at Front Panel Express for a plate that is 60mm in diameter, 1.5mm thick, and a 25mm hole, and it comes to $22. No other holes, yet. Not bad at all if it gets rid of spikes.

The 60mm outer diameter costs $4.63 to mill, and a 25mm inner hole costs $1.17 :-) :-).

Chen

Kevin_A the aperture rings are removed from the back end. No need to touch the front elements.

Ah, very good. Thank you!

If we can change the iris, it could be a game changer. Right now, we are stuck between a bladed iris' diffraction spikes (everywhere in the FOV), or a aperture mask's notches from aperture vignetting at the corners and edges of the FOV.

If the f number is slow enough, there would be no aperture vignetting, but we are getting in the f/4 region with this lens, which is a bummer when the lens is so good.

A circular aperture at the right place would have neither problem.

One of the first things I need to do (assuming I don't destroy the lens even before getting to the iris) is to record the iris diameter for each of the aperture settings. And then measure how much thickness is used by the iris -- if it is less than 1.5mm, then I would not be able to use the Seattle web machine shop to fabricate them.

BTW, after opening up the lens, do you see the possibility of removing the bayonet mount, and replacing it with either a 48mm threaded spacer, or 54mm threaded spacer? Heck, even the good old 42mm/ 1mm pitch Practika lens mount if you forgo full frame FOV. Or glue the camera angle rotator to the back of a bayonet-less lens, ha ha.

Chen

Kevin_A

Back to the FSQ -- remember the little glitch in the retaining ring of the IDAS NBZ-II causing two large spikes? I substituted a retaining ring from a different filter, and the spikes are now gone.

Prefilter to reduce halo? Bunk!

An Optolong UV/IR cut filter placed between the FSQ flattener and the NBZ made the halo problem way worse! :-)

This is a stack 10 frames of Arcturus (Mag +0.15) at 2 minutes per frame, and gain 0 on an ASI2600MC. The image is first cropped to 2048x2048, and scaled down by 2 to a 1024x1024 JPEG:

You can see a weak halo around the star, but a huge halo, about half the size of the cropped frame! So, confirmed that a prefilter does not reduce halo, but adds a gigantic (and strong) halo.

This is with the Optolong removed, leaving just the IDAS NBZ-II:

Again, you can see a weak halo (slightly magenta -- so probably coming from the OIII bandpass, and not the H-alpha bandpass, which would have been red). But no more gigantic halo.

This is what Alioth (Mag about +2) looks like, so about 10 dB less brightness:

The halo is definitely there (I saw it in the original NBZ too).

Here is something interesting, but should not be a surprise. When I removed the Optolong, I did not have to refocus the FSQ -- because focus determines the distance from the Petzval to the flattener, and I did not change anything there. I simply reduced the backfocus adjuster to compensate for removal of 1.85mm of glass, and was back in business without having to refocus the EAF :-). Everything done by the books.

Anyhow, it is good to know that the spike was from the retaining ring, and not from the backfocus adjuster, at least for APS-C sized cameras. This means I can just leave the backfocus adjuster on the FSQ, instead of getting an equivalent spacer fabricated by PreciseParts.com. This way, when I change filters, I don't have to fumble with shimming washers -- simply pop in a filter into the filter drawer, and dial in the 1/3-glass thickness amount into the backfocus adjuster.

OK, done with the prefilter experiment. And the FSQ is back to normal without the weird spikes -- but now with a permanent backfocus adjuster.

I might try for NGC7000 to see if Xi Cygni has a halo -- some narrowband filters often shows a halo on that star when exposing to get enought light from the North America nebula. The nebula should come over the trees in an hour or two.

Chen