The Mk1 PiCamCollimator RPi 2W

[paul]

 

After much gnashing of teeth and sacrificing hair I've managed to spew out a 3D print for the PiCamCollimator. there is nothing like diving head first into new-to-me tech. The 3D printer has been excellent, I just need to learn what makes a good print rather than just a functional one. But still I'm pleased with my first project. After several prototype I now have something rugged if a bit brutal in aesthetics.

 

Freecad takes no prisoners - straight forward to use  it suckers you in. All goes well until you hit a bug or you want to tweak something - that's even using the recommended branch.  What a nightmare, I had to scrap several projects because I broke the model doing something that seemed obvious. I need to find a better free parametric modeller.

 

I still have to design the lid (it will be an L-shape secured to the back of the camera module). The collimator only needs power (I VNC to the deskptop and run the app from there) but I've kept the ports open for now so it can do other things if I need it to. I have to wire in the LED illuminator.

 

With the zoom functioning it is now as capable as the commercial system. Once finished I'll post more details.

 

Edited April 7, 2022 by paul

[len2376]

On 4/7/2022 at 5:06 PM, paul said:

I need to find a better free parametric modeller.

Or not. I just use Tinkercad, yet to find anything I haven't been able to model but then my astro needs are not that complicated. Bit like Lego, add a bit here, subtract a bit there. What I see above would be easy. Maybe there's something in there that I can't see that needs parametric modelling?

[TerryMcK]

It is looking good Paul.

[paul]

There is now a lid which has to clip and fit in place.

 

The camera sensor is offset from the centre of its pcb. I measured all the dimensions to ensure its dead centre to the eyepiece. That's pretty critical to the accuracy. Its taken me a few iterations to allow for the tolerances of the printing process. Keeping all that in a spreadsheet simplifies the rework. It also means that three pieces adjust their dimensions if needed automatically. 

 

I also have an LED pass through that goes through all three parts which is tricky to align. The LED illuminates around the sensor which eases the final stage of alignment. 

 

I've learned a lot about the tools and materials. 

 

If there is ever a Mk2 I will probably reduce the number of fasteners and  have the camera mounted in the eyepiece section to reduce tolerance build up and simplify the design a bit. 

 

The goal was to make an camera collimator, I've achieved that. And it works well and is very useful to get the whole optical train aligned. The vignetting on my subs are nicely symmetrical which is something I could never achieve by eyeballing it with a Cheshire. 

[paul]

Over the weekend I managed to do a lot of testing with the Mk1 collimator and comparison with my Baader laser collimator which I take as my reference for dialing in the secondary and primary tilts.

 

The focusing is working remarkably well, I can adjust from about 5cm (the end of the focuser), to the secondary, to the primary and back to the focuser reflection. Combined with the software zoom it makes the alignment process precise.

However, I have a collimation issue with the collimator itself.

 

The Mk1 works exceptionally well for aligning the secondary central (coaxial) to the focuser tube but is only good for rough secondary and primary mirror tilt adjustment. Checking with the laser collimator showed the laser pointer to be just on the edge the primary's central doughnut which of course throws the primary tilt off also.

 

I tweaked collimation with the laser then replaced the Mk1 back in the focuser to see a small but obvious offset.  The roundtrip light-path for my big newt is 2.4 metres - it doesn't take much for the image to be off by 5mm. No big surprises - I need to build-in collimation for the collimator!

 

I'm currently designing the Mk2:

1. Stiffer plastic design needed. The camera already seems rigid but there is enough flex to shift the image if you press on the body lightly. This is easy to sort out.
2. Collimation adjustment needed for eyepiece relative to the sensor, despite careful measurement it is obviously not coaxial. That offset can compound errors.
3. Tilt adjustment for the camera board. This is to ensure the sensor axis is aligned with the focuser mechanism and not the case housing.
4. Build in a triangle of three LEDs around the sensor for the final stages. It's very dark looking back down the focuser tube. They will be the final precision collimation stage.
5. Click to zoom. Currently only the central portion is zoomed.
6. Provide a switch for hotspot or WiFi AP mode.
7. Provide a power down initiating button

 

[TerryMcK]

You are certainly moving on a pace with the collimator. I'm interested in the 3D printer you got. Which one was it?

[paul]

39 minutes ago, TerryMcK said:

You are certainly moving on a pace with the collimator. I'm interested in the 3D printer you got. Which one was it?

Yes, the project is moving along quickly. I wasn't prepared for all the prototyping I would need but the printer is working very well out of the box. 

 

It's a Voxelab Aquila X2 which is a Creality Ender 3 clone. Very easy to put together and not too noisy if you can live with the fan noise. Out of the box it produced excellent results. I have it on a table next to me and you can hear the fans but neither I or my wife find it objectionable (we work in the same home office). 

 

I haven't used any material but the provided red PLA filament and a reel of black PLA so far. 

 

I have designed and printed the collimator, a device for centering my secondary and confirming the focuser is square to the OTA and a utility belt for the asi1600 to anchor cables and provide a rotation index.