Antenna Rotator... and Satellite Dishes
I previously wrote about my antenna rotator, which is actually a Pan/Tilt mount intended for large security cameras. In that post, I mentioned there were a few improvements I wanted to make. I've managed to do most of them, but ran into various issues along the way. Let's talk about that 😎
Mounting (the mount)
There's a video of the antenna rotator swinging a large yagi around my garden. The antenna rotator was attached to my old dining table using four screws inserted at an angle. Without nuts and bolts, the antenna rotator makes it impossible to screw into something, at least without one of those fancy right angled screw drivers. The rotating part gets in the way. Anyway, screwing it to the table also sucked because the tail of the yagi would smack into it if I wanted an elevation of more than around 70 degrees. So, I started thinking about what I could build or buy to support the rotator, some kind of podium or speaker stand and even asked some hackspace members about advice on structural stability of stuff built from aluminium extrusion. Then I realised, I have a perfectly good bike stand which I rarely use.
The bike stand has four legs (yes it does wobble a bit sometimes) and extends such that the highest point is roughly 2 metres high. You'd have to be quite a tall bike mechanic to raise it that high. Getting the antenna high up helps a tiny bit with reducing my minimum effective horizon (because of stupid buildings getting in the way, namely a car sales place behind my house). The only problem was how I'd mount the rotator to the end of a bike stand. The bike clamp fortunately comes off very easily, but then you've essentially just got a stick with legs, unless of course you put something on the end of the stick. Which is what I did, thanks to a recent purchase: A Bambu A1 Mini - he's got googly eyes now and does an amazing job. I designed a quick mount in OpenSCAD. Here's a rendering of it, followed by a picture of the printed mount (from completely different angles).
The yagi isn't particularly well balanced. I think they made an attempt at this but couldn't due to the positioning of the active antenna. Anyway, I was a bit concerned about the weight so I went for PETG and used a fairly high infill of 35% ish. It's surprisingly stable. I suppose most of the weight sits on the bike stand directly, and this just handles the pulling from which ever side has the most weight. Okay, next improvement…
Mounting (the antenna)
Uhh, my quick solution (which was in the previous post) was to buy an antenna mount from Screwfix, which to be fair was only £7 or something. It was 30cm long and was intended for mounting a dish to a wall, but predictably the mounting holes didn't line up with what was on top of the rotator so I mounted it using only one bolt. This would slip around and wobble a lot. I decided that I to put the 3D printer to good use again.
My CAD skills do not result in stylish things, but they at least work. And this worked amazingly. The angle of the hole was 27 degrees, meaning I'd be able to point the yagi all the way to 90 degrees elevation and cover overhead passes. It was also printed out of PETG, for the same reasons I guess.
Pointing it at the right things
As mentioned last time, the built in position encoders are not great, but the main annoyance is during setup. I have to get a compass out and figure out that it's pointing in the direction shown on the computer. What I wanted was an electronic compass to get the azimuth and some kind of gyro to get the elevation. Knowing nothing about these things I ordered a three axis gyro, which didn't do what I want at all. Moaning about it on IRC, Toby helped me pick out a sensor which should do the job. I ordered a "Tilt Compensated Magnetic Compass (CMPS12)" from The Pi Hut.
I got the thing working pretty quickly using an ESP32-S3 and tweaked the rotator code to use azimuth and elevation from MQTT. The ESP32 was programmed to push this every 100ms. Rotating it on my bench and moving it around seemed to work well, so I moved it to the rotator and found myself getting annoyed.
The compass was showing the incorrect heading - I'm not sure if this is due to its "constant calibration" being bad at startup, or if the magnetics of the rotator were causing issue, but it was far worse than using the rotator's built in encoders. I've temporarily given up / put this issue aside for the time being and might return to it eventually.
Satellite Dishes are Cool
Here's a screen grab of Dr Ellie Arroway sitting on the boot of her car with headphones the wrong way around, steering dishes of the Very Large Array moments before hearing signs of intelligent life from outer space.
While my plan isn't to find aliens, I did / do want to receive high resolution images of earth from weather satellites. The yagi is fine for receiving APT (Automatic Picture Transmission) because the frequency used was pretty much in the 2-metre band - 137MHz. The same satellites also send images in the L-band, which is between 1 and 2 GHz, usually around 1.7GHz. Attenuation through air, clouds, coax, connectors, etc is much higher at these frequencies and so a parabolic dish is typically used. Thanks to JonW picking up a 60cm Sky dish from the side of the road, I now have a small dish.
The dish is off-axis, as opposed to a prime focus dish where the antenna is directly in the middle. This has the advantage of not casting a shadow, but also means you have to point the dish at an offset angle. Since the dish was unknown, I stuck a laser pointer through some pieces of cardboard taped around the LNB mount. This allowed me to find the centre of the dish. I marked a dot and stuck a small mirror there, which gave me an angle of about 30 degrees. Actually, if you live in the UK (or at the same latitude) you'll notice that most domestic satellite dishes of this kind are all mounted almost vertically. The elevation from Nottingham to satellites in geostationary orbit above the equator is about 25 degrees, so this all makes sense. 🙂
Once I'd figured out the dish angle, I built a helical feed antenna using the scaffolding by derekcz. The feed requires a ground plane, so I mounted it to a square of aluminium plate. With some tuning, I ended up with something seemingly resonant at 1.7GHz.
I had run out of PETG, so I bought some Bambu PETG-CF (Carbon Fibre reinforced!). Printing with carbon fibre filaments requires a hardened nozzle, so I had to buy one of those too. It came out so nicely, it looks far better than anything I'd printed out of PLA. I decided to print some soap dishes out of it too because of how good it looks.
With that all setup, I was finally able to receive my first HRPT image from NOAA-19. It definitely missed some frames and didn't do the composite images right, but it's still cool and I'm very happy with it. There will be some gradual improvement overtime I'm sure. I will leave you with my bad image (showing the NO layer (I don't know what that is exactly)), and a cute selfie with the dish.
Related posts:
Wanting to leave a comment?
Comments and feedback are welcome by email (aaron@nospam-aaronsplace.co.uk).