Advice about building dishes for 21cm radio telescope

[Apeirohedron]

As a next step for the 21cm radio telescope I've been working on, I'm planning on building an interferometer out of at least two dishes, and if possible to build the dishes myself. Since I'm still pretty inexperienced with this stuff (the learning curve is pretty steep, as I'm sure you all know), I wanted to just explain my thought process and see if anyone has any corrections, suggestions, or other input.

I know this is a lot to respond to, and I intend on posting some of these as their own questions, but if there's anything in here that someone wants to comment on I would really appreciate it.

So here's my thought process so far:

[Reasons to use homemade dishes]:

1. Going by the rule of thumb that radio waves 'ignore' any details smaller than λ/10, surface imperfections in a homemade dish (most importantly seams between the different conical frustum-shaped segments) would not cause much of an issue so long as the dishes have the same shape.
2. I have thin sheets of aluminum and mesh on hand, which should be easy to work with (can be cut with tin snips and riveted together) even for someone like me without experience with making things out of metal.
3. For smaller dishes, the amount of the dish obscured by the feedhorn is a major issue, and since the offset satellite dishes I've been using have high f/D ratios the horn has to be long and obstruct a lot of collecting area to prevent spillover. If I made the dish myself, I could make it less shallow and shorten the tapering section on the feedhorn, making the dish a bit more efficient.
4. Satellite dishes are relatively expensive: I haven't been able to find anyone local with a reasonably large satellite dish for a reasonable price, and buying two satellite dishes isn't prohibitively expensive but it's certainly more than I want to spend if it's not necessary.
5. It would be a useful experience, and even if it takes longer than buying my own, some of these techniques would be useful for building better feedhorns for future projects (currently, I'm using a steel coffee can with a horn made of posterboard covered in aluminum tape, which does seem to work but also definitely isn't even a little bit waterproof and might not work as well as a horn made of solid aluminum).

[Difficulties/Reasons NOT to]:

1. Radio astronomy is hard and time-consuming enough already without having to build multiple parabolic dishes and fix them if/when things break; if I just use regular satellite dishes, I wouldn't have to think about whether or not the dish is the problem when the telescope isn't working.
   1.  On top of that, interferometry is hard and I've never done it before: with all the difficulties involved in making an interferometer work, any sources of error I can eliminate will make it easier to identify the source of a problem. Even if they're suboptimal, regular satellite dishes are consistent in a way I can't really match.
2. While the feedhorn and receiving electronics don't weigh that much, things add up and the aluminum sheet or mesh definitely isn't rigid enough to support them when the dish is at an angle, so I'm going to need supports to hold everything in position. That could just be PVC or something like that, but it's another step and if the supports aren't strong enough the dish might flex or the feed might move around when at an angle, which would be completely infuriating and make it hard to get any meaningful data.
3. Most importantly, interferometers usually depend on each receiver behaving the same way (having the same radiation pattern, pointing in almost exactly the same direction, etc.), and the imperfections in homemade equipment (especially given my lack of experience) might make it impossible to get the interferometer working properly. I don't know what the margin of error is, or how I could figure that out, and if it's stricter than I can achieve it wouldn't make sense to go in that direction. I want to make a radio telescope, not a bunch of poorly made parabolic dishes that don't work well enough to be useful.

Sorry for the all the rambling, but I hope that given how many different things people on this forum have tried that you all might have some insight into what kind of decision I should make. If I'm wrong about anything I said, which I'm almost certain I am, I would really appreciate any corrections or clarifications.

Thanks,

Aidan

[Henri NICOT]

Hello Aidan,

Your email has caught my attention regarding several issues for which your considerations are also mine. So let me list and comment them which I think is what you wanted:

• I am located in Europe, with no past experience in electronics nor astronomy, just a taste for it which I can finally develop now that I have retired.
• Radio astronomy, I have discovered, is not widely spread and calls upon so many other sciences (electronics, physics, statistics, geometry and above all mathematics), that it is easy to make detours while stepping into it. Being also a beginner in amateur radio astronomy I could not guide you much but having embarked on a big detour into the design of parabolic antennas, I can share my experience in that field.
• The λ/10 rule of thumb allows the use of mesh instead of solid plate and that makes it easy for the reflective surface of the paraboloid. But that is true only up to around 12 GHz where the tolerance shrinks to 2 mm. So you can do the 1-10 GHz range easily. Below 1 GHz, the antenna source becomes a bit cumbersome, above 10 GHz the reflective surface is too demanding for a wide diameter classic mechanical approach.
• The λ/10 rule of thumb imposes severe constraints not only on the surface of the reflective area (seams, bolts, rivets and retaining rods) but also on the overall structure of the antenna. It may not twist, bend or even shake more than the λ/10 tolerance and that includes the source support as well as the paraboloid. Obviously, the bigger the diameter the easier it becomes to flex the structure through gravity, rotation or wind. One centimeter is quickly achieved. The result is not disastrous but you get a dent in your transit curve which starts to look like barbed wire, a dissymmetry in that transit curve and an overall downward descent of that curve. If you look at the sun or the moon, no problem but if you are trying to detect a weak source, you will never know if you did or if you hit the default of your antenna. 
• Spillover has been a major issue for me. You will read that what counts is to match the angle of reception of the source with the f/D of the paraboloid. It is true but when you start flirting with the border, sooner or later you will get a spillover. It could be the case when you change your source as not all of them have the same angle. The revealing acid test for me is the following: Put the antenna to the zenith and read the signal as you turn the azimuth from 0 to 360°, keeping the elevation at 90°. You should have a constant signal corresponding to the temperature of the cold sky, whatever the azimuth. I used to have an intensity varying widely with the azimuth because spillover from various origins (wall of the house next door, neighboring tree, unleveled horizon was being mixed with the cold sky temperature and giving me a false value of that temperature. This is complicated with the polarization of the receptor if it is not circular. I reached the conclusion that if I wanted to have a clean and pure signal, I had to "pay" for it. I have decided to build an antenna with f/D of 0.25, being sure that no signal can get directly inside the receiving source after a chicane down inside the cup and then back up in the source. Of course I lose a bit of the antenna reflective area but there are sources that can capture close to 180°. If the goal is emission or reception of strong communication signals, terrestrial or satellite, or even the sun, then yes f/D at 0.30 is the optimum but if you look at receiving very weak signal, buying a "no man's land" between 0.25 and 0.30 is a wise investment. That is the price to pay for a pure signal. Indeed if no spillover is polluting the cold sky signal then you can appreciate a source hardly stronger. It relies on the sensitivity of your downstream amplifying system. If it is polluted right from the beginning, you will never know. Sorry if it sounds too basic and boring for experienced radio astronomers, it is just something that I experienced myself and yet need to confirm after I build the 0.25-f/D antenna.
• It is possible to limit the spillover by lowering the f/D of the antenna. Changing the shape of the receiving source, the tapering section of the feedhorn and the are actions that require deep knowledge of antenna theory.
• I designed and built a 0.50-f/D geodesic parabolic antenna of 1.90 m diameter, associated with a satellite LNB receiver. The association antenna f/D and LNB receiver was obviously a big mistake from which I learned a lot. I have now finalized the design of a 4.70 m diameter geodesic parabolic antenna, with a f/D of 0.255 and I am currently doing the production drawings.
• I am interested in it, read a lot about it but have no experience with interferometry. I would consider it after I have passed my teething problems with one antenna, reception of full power signal at 10 GHz and frequency range reception on the hydrogen line.

Attached is a picture of my prototype. If you are interested, another time we could discuss the differences with a more classical parabolic antenna.

I hope they are not too many mistakes in the above, that it will help you and that the more experienced colleagues who were patient enough to read it all, would correct or complete my words where necessary.

Thank you

Henri Nicot

--
--
You received this message because you are subscribed to the Google
Groups "Society of Amateur Radio Astronomers" group.
To post to this group, send email to sara...@googlegroups.com
To unsubscribe from this group, send email to
sara-list-...@googlegroups.com
For more options, visit this group at
http://groups.google.com/group/sara-list?hl=en
---
You received this message because you are subscribed to the Google Groups "Society of Amateur Radio Astronomers" group.
To unsubscribe from this group and stop receiving emails from it, send an email to sara-list+...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/sara-list/ce8cf4fe-1ff7-4209-a306-0ba13c7b7f7cn%40googlegroups.com.

[mike....@gmail.com]

Henri,

Your comment "curve which starts to look like barbed wire"  struck home with me last night and prompted me to go out into the cold (25F) dark to check the tighness of the bolts holding the feedhorn frame to the dish.    I have been seeing "spikes" more so during the day(?wind related).  They are not RFI related because I live on a farm far from town.  I have been continuously drift scanning for a couple months gathering data. But when the altitude is the same for 3 or 4 days the plots do resemble each other in a couple major features but otherwise don't look a like in drift of "baseline" and roughness.  Usually in science repeatability is important and I'm not there yet.

Thanks for the insight

mike  W9YS

My dish is a 3M Birdview sat dish made from steel with a segmented gear elevation mount.

[Apeirohedron]

Thank you both for your responses, they really helped clear up some parts of the process that I wasn't sure about. The fact that deformations in the dishes can look like and even obscure real signals is quite disturbing, and while the dish I had in mind was considerably smaller than either of yours (I was planning on making mine in the 1-2m range, probably close to 1.5m, with the main benefit being the ability to design the dish to suit the feed instead of spending a lot of money on commercial dishes that might be less useful) I hadn't appreciated how little deformation is required to see effects in the signal. My goal with this project is to get data that's consistent enough to be analyzed, so if building the dishes myself gets in the way of that then I won't go for it. It's going to take me some time to understand some of the information here, so I might need to ask some more questions later to clarify.

Thanks,

Aidan

P.S.: Good luck to both of you with your projects!