Observatory 2.0 - Robotic Clamshell

[SamAndrew]

Hi everyone,

Thought I'd share another observatory build thread; my last one is lurking in the depths of this forum somewhere.

I originally built a RoR shed back in 2012 in my parents garden which kind of served its purpose. Not long after completing it, I moved out and got a place in London so only occasionally got to use it. After a couple of years however it was taken over by my dad for his expanding business and it became a stock storage shed 😐

It was never meant to be permanant, and I knew I would likely build a better one at a later date. I learnt a few lessons, the key one being the size; It was huge at 2.4m x 3.6m which meant the roof was too big and heavy for the wooden structure and I would never have automated it without a redesign, however the vast size did serve nicely as a mini warehouse.

The other main lesson was the massively over engineered pier, which took a lot of time and effort to remove thanks to using half a ton of concrete.

Fast forward several years and the business has moved out and stock duties are now handled by Amazon, meaning there is space once again for obsy 2.0 with thanks to my understanding mother (it's an extra reason for me to visit )

 

This time it's vastly scaled down, the concept has a footprint only of 1.1m x 0.5m and a height of 1.75m. The exact dimensions will be determined by the final location in the garden so I haven't done a detailed model. I toyed with the idea of parking the scope in different orientations, but the normal home position doesn't add much to the height and is much narrower.

The structure will be steel framed to give it rigidity, with composite cladding and a plywood roof. the base will be concrete (the pier base will be isolated from the rest of the structure)

The clamshell will be opened by linear actuators. It's next to a fence to the north that will shield it from the wind, I may add some sort of concertina fabric arrangement to span the gap between the two halves if wind is an issue.

It will be controlled remotely for imaging so no need for a warm room or space to observe.

Any comments of feedback welcome. Hope to get the ball rolling soon.

Thanks

Sam

[SamAndrew]

Started the build today - modified the design and moved the hinge point up to about 50cm above the ground which will significantly reduce the amount of weight and leaverage for the actuators to deal with.

Pictures show the main side open slightly using 1 actuator (need to bolt it down otherwise it will topple over). Plan is to have 2 actuators on each side; I hope these cheap ebay ones will open at the same rate otherwise that might pose a bit of a headache.

Built a pier as well and tested the remote power up and connection to the mini PC I'll be using.

Think I need another half day to finish off the frame, brace all the corners and then tighten it all up. Half a day for the ground works and another half day for cladding.

Got one day booked off next week to work on it and then will use the Christmas hols to complete

[Alan White]

Interesting use of Dexion type material.
Watching with interest.
I take it this is for imaging.

[SamAndrew]

On 03/12/2019 at 13:04, Alan White said:

Interesting use of Dexion type material.
Watching with interest.
I take it this is for imaging.

Yes imaging only. Dexion is useful stuff for prototyping  built a robot for Robot Wars out of it too. 

[SamAndrew]

Made some more progress today; photo taken after I added the second door and all the actuators. I then added in some bracing to stiffen it up (not pictured yet).

I then dug and poured the foundations (was dark when I finished so no photo). Hopefully they're up to the job; only 250kg of balast used in total including 50kg for the Peir; the last one I did was 0.5 tons for the pier alone.

I've set a dozen 150mm bolts into the concrete for the frame and pier.

Set the scope up on a tripod inside just to confirm it still fits - some slight doubt in my mind now as to weather I'll have a complete range of movemnt when the doors are fully open after changing the design. Theres about 10cm of clearance at the top so I can shorted the legs to give the scope more clearance if necessary.

Also written an app to control the doors and capture weather data; I'll start wiring the electronics up this week.

Hope to do the cladding this weekend: plan is to bolt wooden battons verticaly all around and then screw PVC cladding to that.

Edited December 9, 2019 by SamAndrew

[sloz1664]

Hi Sam,

Very interesting project & looking forward to your progress. I understand you are still in the developing/build phase of your project and is far from complete. But, may I add some pointers you may consider:-

Your actuator clamping points need to be strengthened. I advise adding fillet plates to the vertical & horizontal framework on both vertical sides of your Obsy. To strengthen the framework & a means to securing the actuator hinge points, I understand the actuators are taking the brunt of the work, but, your hinge points need ample support also. In fact, any moving part mounting point needs to be robustly supported.

As you are going to using this in a remote capacity I would certainly add an anemometer to check wind speed. Unfortunately, and I can understand why, your design is not aerodynamic and the main body would act as a sail in windy conditions. I use the AAG Cloudwatcher, with the additional anemometer, as my weather sensor and set-up in safe mode within SGPro.

Steve 

Edited December 10, 2019 by sloz1664

[SamAndrew]

Thanks for the suggestions; I hadn't thought about a wind speed sensor; I was planning to rely on the weather forecast for wind information, but a sensor is a good idea. It's fairly well sheltered on two sides, which will reduce the wind load, we'll see how it goes.

Hinge points will be strengthend up once I'm happy witht the positioning - haven't been able to fully open it in the garage.

Abandoned the idea of building my own H-Bridge motor driver and just bought a couple of BTS7960's which has made the Arduino project very simple.

Made a bit of progress today - drilled some holes in the house to get power and network connections outside and setup an all sky camera and IP camera - currently watching the rain fall on the ASC now

The ASC is an ASI 120mc-s in a cheap CCTV dome enclosure with some added waterproofing.

[sloz1664]

Can't wait to see this up and running & it's a great idea to have a compact robotic enclosure to show you don't need an enormous amount of room to house your equipment. I'm particularly interested in the arduino project driving the system.

Steve

[SamAndrew]

This is the diagram for the motors; just a simple high low signal to the driver for each direction. I'm using two so I can control each door separately. Will post the code up later.

[old_eyes]

Very neat design.

I am planning to use a linear actuator to operate the South flap on my ROR observatory. It looks very similar to the ones in your pictures and has a 150mm throw. Maximum load is >100kg in pull and about 150kg in push. Since I can raise and lower the flap with one hand I was hoping one would do the job. I could always use some counterweights to reduce the load if necessary.

I think @sloz1664 is right about wind loading. That partially open clamshell design  could experience some serious pressure if the wind gets up. Is there room to open the clamshell completely? That would dramatically reduce the loading.

[SamAndrew]

There will be enough space either side to open it all the way to the ground if necessary, I could raise the lower actuator pivots up so they're in a more horizontal position when closed, as there is plenty of stroke to play with, we'll see how the design does initially. It's unlikely I'll be using it if it's that windy.

Tested it for the first time yesterday; I wanted to double check the scope had enough space to move in all directions before reinforcing the pivot points. Thankfully it did, and I have about 15cm of space above the scope so could lengthen the pier or reduce the overall height; not sure a long refractor with camera hanging off the back would fit mind. 

Pic below shows the actuators at about 60% extension.

I have also added a plywood roof and bolted some vertical wooden strips to the sides to attach the cladding to.

Now just need a free dry day to get it outside and bolted down so I can fit the cladding. 

[SamAndrew]

Quick update. Fitted the cladding to the upper clamshell and then moved it outside, the below pics were taken in progress; the top section is now completely clad and has essentially been on trial for the last week to confirm weather proofing (not that there has been any real rain yet). The rear side against the fence has a plywood panel that can be removed for access in the event that it doesn't open for some reason as there's no way to open it from outside. No insulation or membranes fitted; I'll see how it goes with dew forming inside and then possibly add a heater or fan as necessary.

Planning to get it operational this Saturday; jobs still to do:

• Move all sky camera to the fence post behind the observatory, add dew control heater.
• Install PC inside the obsy and connect up to the electronics for the actuators - test and adjust things as necessary.
• Complete the lower section - triangulate the corners, add batons, clad
• Adjust the pier and install the scope (polar align if the weather allows!)

Already got ideas for improvements for version 2.1 and 3.0 as I now realise there was no need to have the narrow sides as part of the clamshell; these could have been fixed and then just have two flat sides that butt up against. Also instead of hinges the sides could slide to expand like a caravan/motorhome but that's a much bigger change. Will start sketching up some new designs. The good thing with this build is it's all bolted together so can easily be adapted in the future.

[SamAndrew]

Well it's operational! It has taken longer than expected, but then that's the problem with building something at a remote location and only getting a few hours a week to work on it. In hindsight I could have built it at home and borrowed a van to transport it, which is what I will do for the next version.

I got the scope setup on the 3rd of Jan, and then fettled it today and have taken some trial images. Had the double whamy of everything being completly new (scope, mount, camera, aluto focuser, PC), but it's all connected up and working relatively painlessly. Everything seems to have survived a week of damp weather. Overall finish is a bit untidy but I have had to rush things during the limited hours I was working on it

Main outstanding issue is I've experienced the PC restart on a couple of occasions, so there is something not 100% with the power supply (has been fine for 5 hours this evening); I'll look to add an uninterruptible power supply. I'm currently turning everything on and off using a wifi controlled 3 gang extension lead, but I want more control over each item so I'm going to add some arduino controled relays.

I want to move my IP camera so that it's directly in front of the scope facing south so I can more easily recover to the home position if I do get a power disconnect (Camera is currenty mounted on the house facing NE). I'm also going to upgrade the actuators to ones with built in hall sensors, and then replace the arduino with an ESP32 to have better control of the door positioning, but neither of these things will stop me using in in the meantime)

It has been an interesting project; not one that I spent much time planning, but I'm now much better aware of the challenges, and I'm working on an improved version. Total cost has been about £550 not including the PC or all sky camera.

First cope of pics are from Friday the 3rd when it had it's first partial opening and setup.

Other pics are from today where I was adjusting the cladding and sorting out the wiring, and added a plastic membrane to the hinge joint to seal the gaps. 

[SamAndrew]

Well, quite amazingly, it works! . First night in action, only issue so far is the guide scope is too out of focus to guide with 😐 manually focused on a distant object when i set it up and then added a bit, but haven't quite got it right. Planning to add an off axis guider, so might bring this upgrade forward and sort it out in the next few weeks.

Currently just running some 1m unguided subs on the bubble nebula and it looks like they're coming out ok. The heater on the secondary is protruding into the light path and causing some odd star shapes, so that will need addressing too. Other than that, so far so good.

[SamAndrew]

Didn't push it too hard last night, packed up when some clouds rolled in around 10pm. Had a heart in mouth moment when one side didn't appear to want to close! but it was just the IP camera buffering thankfully!

Shot a few different targets, here is the best 30 x 1 mins on the Horse and Flame nebula. Just a stretch of the raw stack, no darks (didn't think to shoot any at 1 min when I was originally setting up the camera) and no flats yet. Testament to the EQ6-R how well this has come out without guding at 1 arc second resolution (quite a high sub rejection being the obvious disadvantage of not guiding)

In action again tonight - 1 hour on IC63 so far whie I waited for M42 to pass the meridian. 

Edited January 20, 2020 by SamAndrew

[SamAndrew]

Here is a shot from the IP camera; scope facing south. As you can see the scope is fairly obscured so if I lost position it would be fairly tricky to repark; think I'd have to get on the phone to my mum if it was an emergency; I'll give her an emergency tarpaulin at some point

House obsures the view to the west so the door doesn't need to be opened any futher than its current position.

Wonder if any of the IR light from the IP camera is having an impact on the SNR, will have to do an experiment at some point but I don't want to turn it off at the moment.

[SamAndrew]

Well it has survived the second storm of the year already, worst in 7 years, so the construction is sound!

Planning to fit the OAG this weekend and make a few adjustments to the power situation and actuator positions.

Have been working on the next version. More detailed planning this time; still deciding on the best way to position the actuators.

I've purchased some actuators with hall sensors so I can syncronise and accurately control the positions. Have also built a new control box based on and ESP32 this time which should handle inturrupts from 4 inputs better than the arduino and give me some more options.

Edited February 11, 2020 by SamAndrew

[Xplode]

Interesting project!
Guidecameras usually don't have IR filters so if anything that's the camera that will get any effect from the IR light.

[SamAndrew]

Thanks!. Hadn't thought about that, yes that will be a good test to see how much light the IP camera puts out using the guide cam.

Want to build an accurate model of the mount and scope to position in the renders so they have a bit more perspective, but tricky to do now that I no longer have the scope at home 😐

Edited February 11, 2020 by SamAndrew