Ad Astra Observatory

I hadn’t posted in awhile, and I just realized that I’ve been fixing some stuff in the observatory that probably warrant mentioning.

First, I had a couple problems crop up during Snowpocalypse 2019 (dumped 16″ of snow on the deck, over a week or so!); one of the shutter’s home position brushes (the copper loops that transfer power to the shutter motor) snapped due to metal fatigue. Also, while trying to open the iced-over dome, I over-tensioned the shutter cables, and they started slipping, requiring me to manually assist the motor in closing the dome.

I also had a few things on the longer-term todo list that needed doing. More on this in a moment.

The copper loop was one of those 5min fixes that takes a couple hours and two people to set up. It’s a piece of copper flashing, hand-cut to 1.5″ wide or so (one of them is actually not rectangular, but I’m getting ahead of myself), held into a loop by two machine screws. The screws throughbolt the plastic L-bracket that attaches to the dome wall. The L-bracket has slots cut into it for adjusting the loop, but it was already all the way retracted, and the loop was cut so long that it dragged as the dome rotated. This was a source of bending fatigue, and was going to fail eventually. I feel like Technical Innovations (the dome manufacturer) often sends bits slightly oversized, because that could be a source of spare parts in what could be a pretty remote environment. So, thanks to them, there was plenty of “meat” left on the part, to drill a couple of new mounting holes.

I had to get the stupid thing off the bracket first, though. This would normally not be a problem, but I’d done an aftermarket modification to the mounting brackets; they hold the limit-switch circuit that I use to sense when the shutter is open or closed. I had to partially dismantle my add-ons in order to reach the mounting bolts for the L-bracket. Sigh. I also hopped outside the dome for part of the process, to make sure that I wasn’t spinning the mounting bolts. Needless to say, it was very helpful to have a friend assisting with the job (thanks, D_!).

Once the copper strip was off, it was pretty easy to drill some new mounting holes (lots of leeway, here, as the strip was cut long in the first place– thanks, TI!), and reinstall it. While I was messing around, I decided to shorten the other brush, too; I’ve never been happy with how they scrape the dome base while it’s rotating in azimuth, and this was definitely going to fail from fatigue at some point, too, so I figured “might as well fix it”. With the brushes now floating above the dome base all the way around, I was also finally able to install the “brush cleaner”, a little chunk of green scrubbie and double-sided tape, that the brushes wipe across as they go, which should keep them nice and coppery for a good long time.

It feels good to have the shutter working again, and it’s definitely an upgrade that it doesn’t make yucky scraping sounds all the time. Yay!

Buoyed by the success, I decided to try to get the “motto” buttons working on the desk again; they’d gone buggy while I was installing the buttons for dome control into the desk. I didn’t note the problem on the blog, but I had a couple of todo items sitting on the whiteboard in the observatory for what turns out to be a bit over two years. Have I mentioned that I dislike leaving projects in a broken state? *eyeroll*

I got the left-hand motto button working pretty easily; it just required re-seating the cable ends on the button-to-Arduino cable run. I’ve been dissatisfied with the press-fit connectors that I use for standard header connections; there must be some kind of tool that gets used to make these, but I usually end up stabbing myself in the hand several times, and still end up with sketchy, failure-prone results. This is what had happened to this cable.

I tweaked the cable end a bit, and got it all working again. The LHS button now works again! woot.

The RHS button had a more annoying failure — the buttons light up because they have an integral LED. I’d spent some time testing these out, and like all LEDs, the final installation required a current-limiting resistor. I chose 220Ω because it made a pleasantly-dim light and was easy to find resistors for, and then direct-soldered the resistor to the button pin (“dead bug” style). Yeah, the resistor lead snapped off. *unamused* So I pulled apart the RHS circuit to work on it, only to later realize that I’d set aside all the necessary parts in a baggie, when I had first messed it up. So now in addition to soldering in the new resistor, I had to go back and figure out how the thing was all wired up, so I didn’t hook something up backwards by mistake. Lucky for me, I had a fully working example on hand (in the form of the LHS button) 😉

Messing up the project even further made me realize I’d found a good stopping point for the evening. As things go, it took another week or so before I got around to working on the shed again.

When I picked things back up, I got the button all soldered in place (I burned my finger in the process — I think that’s maybe only the second soldering iron burn I’ve ever sustained; I’m usually quite careful), got it reinstalled, and hooked up all the wires, only to find that it was having some kind of power short. After a bunch of frustrating troubleshooting, I came up empty again; the bits were all together, but I needed to get back to basics and figure out what was wrong. Too much for my energy level that night, and I again walked away in disgust.

About a week later, I took another crack at things, with two main problems left on my top-priority list. I needed to fix the shutter cables, and I needed to fix the RHS button.

Re-tensioning the shutter cables went first, as it’s nice to be able to open and close the dome without a helper (definitely on the critical path to a successful night under the stars). After spending some real quality time with the shutter cables (I do sometimes wish for coated cables, although I avoided getting any cuts on my hand from frayed shutter cable this time), they were properly tensioned, but still slipping and requiring manual intervention in spots. hmm. I checked outside the dome, and nothing obvious sprung out. Then I had my helper move the motor back and forth while I was looking carefully at the gearbox, and I figured out that one of the drive sprockets was slipping 😮 I tightened up the setscrews for that sprocket, and lo and behold! the shutter works like a charm now. Thanks again, D_, for being my super helpful assistant-and-sounding-board during these repairs.

The dome is now working really well mechanically, and it feels really good to have the shutter concerns off the todo list.

Getting the button working was a clinic in hardware troubleshooting. After determining that the 5v cable was faulty, I re-cut it and put a new end on it, which showed that it was *still* faulty (oh, and I have lost my DSO Nano — I used the homebuilt oscilloscope that I built from a kit, which was super helpful for seeing the flaky kind of connection that was broken in the power cable, which didn’t show up on a DMM). I sacrificed a known-good USB cable that I had lying around, put a new end on it, and got that all working great… only to find that the faulty cable had apparently fried (of all things) the wall wart itself. Are you kidding me?! Who checks the wall wart for problems!? I stole another wall wart I had lying around (did I mention that I need to re-IP the webcams under the dome?), and within seconds, the RHS motto button was working again.

All the broken junk got thrown away, I chopped the faulty power cable into pieces, and bent the power pins on the broken wall wart, just so I don’t come back 6 months from now and say “oh *here’s* an extra…”.

And I cleaned up the todo list on the whiteboard (you’re welcome, Jr. Astronomer 😉 ). That felt good.

In all, it was a successful set of maintenance in the shed. I got a bunch of stuff knocked off the naughty list, and I feel like the whole system has been left in a good place, and is ready to both be used, but also to continue working through the “getting older every day” todo list, which had been stymied by being left in a bad situation. Here’s to crushing what’s next (which is dome controls at the desk… and getting the webcams back up… and…) without it taking 2 years to get to. 🙂