Carl Johan N's DIY observatory

[Carl Johan N]

Dear Stargazers, here's a short short text on my remote observatory build. I hope to extend it in the future and maybe post some design advice if people are interested. I'll be very happy to answer any questions if you're pondering about building something yourself  

The observatory is controlled by a Raspberry Pi running Astroberry.  The construction is a plywood box with aluminium rails. A worm drive motor raises the roof. Further details:

Telescope: Sky-Watcher 130PDS
Camera: Nikon D5300. Consumer grade DSLR, nothing fancy.
Mount: Sky-Watcher HEQ5 PRO. 
Guiding telescope: Orion Mini 50mm.
Guiding camera: ASI120MM.
Focuser: DIY Autofocus solution, Arduino.
Control: Raspberry Pi 3B+ running astroberry. Controlled remotely by VNC connection (looking to set up something better though, maybe reverse tunnel ssh?)

Included image of the observatory and first light

Kind regards,

Carl Johan N

PS: if it's frowned upon I will remove this, otherwise more pictures and some more info at https://carljohan.space/

 

Edit part two:

As promised, here's some more photos of the interior and some general design advice. First of all, do make a (remote) observatory!  It's not the most difficult project in the world (although rather difficult if you want it remote, I do admit.) Astrophotography is about exposure time and a dark location. The cost of this project was probably less than half of a monochrome ccd and in my opinion much more important than a better sensor. A better sensor might halve the time it takes you to get a good image, while a (remote) observatory increases the time you can be imaging manyfold. In my case something like 10 or 20-fold since the observatory is in a location I would only have the time to visit a couple times a year. So, first off, my design in a few more images and description.

You can see the two aluminium rails (simple builders merchant (byggemarked in danish :D?)) with two wheels on each side (under the roof, not shown). The green raspberry pi computer controlling the relay on the left, this allows shut off lights, telescope and forward/reverse of the motor (reverse/forwad control is possible with just 2 relays). On the right under the wooden rail is the limit switch which is connected to the raspberry pi. In the first image you can see an aluminium arm going from the roof and underneath the wooden arms (a bit hard to see, sorry for lack of images, im working with what I have here ). This prevents the roof from flying off in strong winds, and is what hits the limit switch under the right wooden arm(not shown, but you can see the wire going there on right)

Design advice (I will update this list as I think of new items to add, and I will also add suggestions from comments if you have one )
Mechanics:
M1. Think about robustness of the design, that means designing for exact constraints and not overconstraining. Note that my design above is severely overconstrained, which actually has led to some issues. What do I mean by overconstrained? Any object in 3D has 6 degrees of freedom, it can move in the x, y and z directions and rotate around x, y and z. In an "ideal" design, there is only one constraint that unambigiously constrains one of these freedoms. An example configuration is a 3-2-1 scheme.  For instance, if we call the direction of the roof movement x, and the transverse direction y and finally up z, then an exact constraint design could be 3 wheels touching the rails constraining in the z direction. Of these 3 wheels only 2 of them should restrict movement in y, and finally the wire of the worm drive motor should constrict the last freedom (x-direction). So why did I not do this? it is clear that the design above is severely overconstrained with 4 z-direction constraints and also 4-y direction constraints. Well it was simple and I thought I could get away with it . The issue that this lead to in my case is that the roof can twist slightly which will lead to the motor pulling the roof more and more twisted. This problem was remedied by the use of oiled felt pads on the sides of the wheel blocks. This reduces the friction significantly and it thus doesnt get stuck, but still, good design is better than sympton-fixing! if you have any questions about robust design I'm happy to answer. Note that I am definitely not saying it's easy to come up with an exact constrained design that you can do with the parts you have available, while also providing a sensible pulling location for the motor. If you're interested terms such as "exact constraint" or "robust design" will lead to more internet sources on this kind of thing. On a final note, 4 constraints in the z-direction is of course perfectly fine, but this is because of the flexibility of the roof in this direction. Same reason why a long table can be seem perfectly stable with 4 legs, where exact constraint design states it should technically only have 3 .
M2: Think about failure modes. Notice that the roof is carried by a rather limsy string. This is on purpose, as the motor is quite powerful. If the roof gets stuck for any reason, the string will break and the roof will slide down safely to an end stop. The alternative is the motor pulling a more important part of the observatory to shreds. Yes this safety feature has been inadvertently tested :D. 

Electrics:
E1. Keep a good order on your cables. Mine look like ~ but it's actually not too bad. Only a single usb cable goes to the OTA assembly (hub on telescope splits to focuser, two cameras etc). A single power cable and usb cable goes to the HEQ5, the power cable then continues to the OTA assembly and splits to camera, secondary mirror heating, cooling fan etc.
E2. Make electrics easy to dismantle and test. Notice how the electronics sit on a wooden board for themselves, and the motor is connected by soldered quick-disconnect connections rather than something screwed in or even directly soldered. This means the power/raspberrypi/relay board can be removed and tested in a minute.

Edited April 30, 2020 by Carl Johan N
added M2

[JamesF]

That's a very neat little micro-observatory.

James

[CraigT82]

That's great! Is there a weather sensor and will it close automatically if cloud or raid is detected? 

[Carl Johan N]

9 minutes ago, JamesF said:

That's a very neat little micro-observatory.

James

Thank you kindly!

5 minutes ago, CraigT82 said:

That's great! Is there a weather sensor and will it close automatically if cloud or raid is detected? 

Not currently no, I keep an eye on forecasts and open up remotely only if its looks good. Further there's a camera at the bottom of the obs to keep an eye on the weather. I'd love to work on a more automatic solution, but I too worried about mistakenly closing the roof onto the scope if it doesn't fold down properly. So i'd need to make automatic checks for the scope position before the roof could be closed automatically anyway. Curious what othe people have done with this if anyone wants tochime in

[scitmon]

Very Nice, I keep tooing and throwing between something like this, and a rollout setup.  Would love to see more pictures of the setup!

[teoria_del_big_bang]

Very neat indeed, and I agree with @scitmon some more pictures from above would be great.

Steve

[Carl Johan N]

3 hours ago, scitmon said:

Very Nice, I keep tooing and throwing between something like this, and a rollout setup.  Would love to see more pictures of the setup!

Definitely something like this, but then I'm biased  

3 hours ago, teoria_del_big_bang said:

Very neat indeed, and I agree with @scitmon some more pictures from above would be great.

Steve

Thanks both, I have updated with more images and description. Although it might be a bit hard to read as it was a bit fast and I don't have many images on hard .

-Carl Johan

[old_eyes]

A very neat micro-observatory.

I must confess I solved the scope position problem on my full-size ROR by following the @ollypenrice dictum to always make sure your roof can close without collision no matter what the position of the scope. If there a possibility of collision it will happen!

I like the weak link idea for the roof. Neat; although it does mean you need to go out to fix the problem. I think @Gina monitors motor current to look for the roof jamming. 

[Carl Johan N]

On 05/05/2020 at 11:32, old_eyes said:

A very neat micro-observatory.

I must confess I solved the scope position problem on my full-size ROR by following the @ollypenrice dictum to always make sure your roof can close without collision no matter what the position of the scope. If there a possibility of collision it will happen!

I like the weak link idea for the roof. Neat; although it does mean you need to go out to fix the problem. I think @Gina monitors motor current to look for the roof jamming. 

I'd love for the roof to be able to close in all positions, but that'd sadly make the observatory so much larger  Motor current monitoring is a good idea, and I actually considerd it. Considered as in I bough a sensor and then remembered the rpi doesnt have an ADC 😛.. so I'd need moore devices, and I think there's quite enough already! So flimsy strings to the rescue.. 

[TomBee]

Dejligt observatorium!
I am thinking about building something similar in my backyard (Nyköping about 100 km south och Stockholm) for my 150PDS och HEQ5.
 

[thierry34]

Dear Carl,

 

I'm interresting about your observatory but I can't see your photos . The link doesn't right.

 

Can you post them?

 

Tnaks a lot.

 

Thierry