Radio Astronomy (on a shoestring)

[morphy_richards]

I recently stumbled across this website http://www.ukaranet.org.uk/uk_amateurs/bobgreef/
It describes the construction of an antenna that can be used to pick up radio signals from the sun and Jupiter and play them on an ordinary short-wave radio receiver (such as a travel radio set) at about 21 MHz.

This is his design for the radio antenna,




He also describes how he manually points the antenna at Jupiter or the sun and picks up readings from them like this...


This strikes me that it could make a particularly wonderful raspberry pi project. The hackability potential is great, starting with just feeding the audio from the radio receiver into a sound card but also the potential for making a directional rig as well as a starting stone for building more elaborate set ups.

I've been interested in the NASA radio Jove project for a while but have been put off by the prospect of having to install something like this on the roof.


Something like this is much more realistically achievable.

[morphy_richards]

And an absurdly simplistic description of how to use a radio telescope to make an image of an object in space is here.
http://www.gb.nrao.edu/epo/image.html
The good thing about this little telescope is that it can be aimed and scanned, so, in theory at least, you could use it to make some sort of an image.

[morphy_richards]

And a very affordable radio receiver has been ordered.http://www.tesco.com/direct/tesco-rad-1 ... 8-0233.prd

[morphy_richards]

In theory, one could achieve the scanning effect needed to build up a picture by keeping the antenna fixed and allowing the Earth's rotational and orbital motion to do the job...
This could then be combined with a fixed visible light telescope pointed in the same direction using a camera module

[morphy_richards]

Something exciting for more tinkering further down the line. Re-purposing TV tuner dongles and analysing radio signals on a Raspberry Pi 2 http://www.rs-online.com/designspark/el ... nu-radio-2
And if you can get your hands on a 1 meter or so satellite dish, a guide to building a telescope that uses this tv dongle based "Software Defined Radio" for astronomy that can provide detail about some good deep sky objects.
http://www.sbrac.org/files/budget_radio_telescope.doc

Still would be a real challenge to build something that can track, pan and scan and so build up a radio or microwave image. Graphs of pulsars and distant galaxies are interesting but they don’t have the wow factor as 'seeing' one.

[morphy_richards]

Musing.
Going back to the original telescope design, what is the impact of having the reflector parallel to the loop?
I've been doing some reading about small loop antennas but I cant find much about using them with reflectors. Would a flat reflector like that just double the power? The circumference of the loop and the distance of loop to reflector seems to have nothing in common in particular to wavelength at 21 MHz as far as I can work out.
Anyway, I'm thinking, trying to automatically angle the whole assembly is quite challenging (given that it's going to be protruding from the roof of my shed and its a meter square of 5mil plywood. But, just altering the angle of the loop in relation to the reflector should be more achievable.

The last time I did anything about radio antennas was in 1996 and frankly I didn’t really understand it even then.

[Burngate]

There are probably many people out there who know a lot more than me; like you I didn't understand most of what I learned.

I can't find anything on the web about that specific type of aerial.
Still, Wikipedia has a page on Yagi-Uda aerials, and you could consider this to be similar, with the folded dipole stretched into a loop and the directors thrown away.

If that's so, then the reflector-loop distance should be important (that page suggests 0.25λ)
Also, since the reflector in a Yagi-Uda is parallel to the dipole in the two-dimensional plane of interest, having extended into a third dimension I would expect to have to stay parallel there also.

On the other hand, a loop aerial has a null response in the direction normal to its plane, with a toroidal response.
Putting the reflector at 0.25λ behind it - does that create an anti-phase image, which should mean that it has a maximum straight ahead?
My maths ain't good enough, but it seems to suggest that tilting the reflector should tilt the response, and having the loop perpendicular to the reflector could make it fire sideways rather than in a toroid.

[GTR2Fan]

Another option would be to buy a cheap satellite dish jack-arm to steer the antenna with and control that via an H-bridge attached to the Pi. I can't remember exactly what the positional feedback signal looks like now (analogue or digital), but they nearly all have end-stop switches, so driving it in one direction for long enough for it to hit an end-stop would at least give you an earth-based positional reference of sorts for starting a mono-directional sweep.

The other big advantage to using a jack-arm is that it's a worm-drive mechanism internally, so it's incredibly torquey for its size and fantastic for holding a fixed position even with drive power completely removed, so there'd be zero electrical interference from the motor when holding a fixed position. If you could somehow make use of its positional feedback to tell you where it's pointing, you'd have a very powerful king-size linear servo. After all, that is what it's originally designed to be from the get-go.

If the shoestring is too thin to support such a venture, a vaguely similar thing could probably be achieved with a secondhand car windscreen wiper motor assembly, although I don't know how well one would respond to being PWM-driven very slowly.

[morphy_richards]

Funny you should mention the windscreen wiper thing, I was also wondering about one of those for the torque and weatherproof aspect.
Would be really difficult to get any accuracy from it though. Faster movement and some reduction as well as some sort of encoder might help.
One of the advantages of loop antenna is that they are quite resistant to RF interference from nearby sources.

[GTR2Fan]

One of the advantages of loop antenna is that they are quite resistant to RF interference from nearby sources.

As is one of the old solid (as opposed to perforated) 80cm satellite dishes if you mount the RF-sensitive front-end directly in front of it away from the drive motor, LNB-styly. I'm wondering if it might be easier to just source a secondhand dish with steerer and mount the dipole at the focal point with a few directors in front for good measure. That would also give you a very narrow acceptance-angle improving resolution no end.

Damn! I wish I had the spare room and tolerant enough neighbours to be joining in and taking this approach. I also wish I'd had the room to salvage and store a few of the tatty old dish/steerer setups I've seen sitting in skips outside people's houses when they've moved.

PS If you do take this approach, don't forget to spray the dish matt black or you'll probably melt your dipole if you point it directly at the Sun.

[morphy_richards]

Still, Wikipedia has a page on Yagi-Uda aerials, and you could consider this to be similar, with the folded dipole stretched into a loop and the directors thrown away.

If that's so, then the reflector-loop distance should be important (that page suggests 0.25λ)

Hmmm... That can't be right though because λ at 21MHz in air is a little over 14 meters. It's really hard to understand how this antenna works, Wikipedia just provides general ”length is less than 1/10 wavelength " and points out that it is magnetic in character.

I'm not entirely convinced it is a proper loop antenna. The design shows only one end connected to a receiver with the other end floating. There is no balun. *

Wikipedia suggests a loop antenna is similar to a folded dipole and a folded dipole has a connection between the two ends.

I think the key to understanding this may involve some complex calculus

*edit. I'm pretty much convinced it's not a loop now. So what is it? Is it just a simple wire arial folded in a circle for convenience in front of a reflector?

[Burngate]

... λ at 21MHz in air is a little over 14 meters.

I'm now feeling really stupid.

Engage autopilot ... read "Radio Astronomy" and "21" ... assume hydrogen line at 21cm ... proceed blindly at top speed towards accident

[morphy_richards]

... λ at 21MHz in air is a little over 14 meters.

I'm now feeling really stupid.

Engage autopilot ... read "Radio Astronomy" and "21" ... assume hydrogen line at 21cm ... proceed blindly at top speed towards accident

Not at all, at least you read it!

My worst trick is open thread, go directly to last post, assume I know what the poster is going to say, not read post properly and make totally inappropriate response.

[morphy_richards]

Look what someone found for me ("it's amazing what you can find in skips when you're drunk") it's a bit broken but hopefully fixable.


Will be used to make a school based radio telescope. Just need a means to aim it now.

[morphy_richards]

And now he's found us a much bigger one!
I love the fact that the antennae is literally just a bit of wire poking up inside a tin can that's facing into the dish.
Hopefully we can borrow an oscilloscope from science today and at lunchtime go outside, point the thing up at the sun and see what we can pick up (that should work shouldn't it )

[GTR2Fan]

If you're talking about using the original LNB, you won't get anything out of it without a 12V power supply. You'd then need to split the RF out from the DC. LNBs are very narrow-band receivers, so you won't get much out of it outside the 11 to 13GHz region.

[morphy_richards]

That should be doable with a standard school science department oscilloscope* and a 12 V supply both of which I can get ...
*assuming it can do those frequencies that is.
edit - it's 1.7 Ghz

[GTR2Fan]

Well, the output of the LNB is more in the 1 to 2GHz range as it contains an IF amplifier. I'd be very surprised if the science lab has a 'scope that goes even that high. Realistically, you won't be able to successfully separate the DC and IF without replicating the input stages of a satellite receiver. An old analogue satellite receiver would be ideal if you're happy to limit yourself to the very narrow frequency band an LNB covers.

I don't think an LNB is particularly useful or suitable for what you're trying to do, but the dish itself (including the LNB mounting arm) is as it allows you to construct your own antenna and mount the dipole at the dish's focal point.

[morphy_richards]

For today it would be great if we could just point the thing at the sun and see something go "wheee" ... sadly I didn't bring in my RTL2832U based USB radio receiver dongle (although I had meant to I was halfway to work before I realised I'd forgotten it)

I haven't got a satellite finder here, the only thing I have got to hand are things like oscilloscopes, multimeters and the like.

For the long term the plan is to see what we can cobble together in terms of an auto-aiming mechanism using steppers from old laminators, some sort of lift mechanism from an old circular saw and a kind of lazy Susan thing that isn't really but might be as good....

It's worth pointing out that this cam about as a a result of a random conversation with someone handy at school about satellite dishes.... the "other" more simple radio telescope for the roof of my shed is still going on, albeit in parallel with this one which is now a school lunch club project.

[morphy_richards]

Realistically, you won't be able to successfully separate the DC and IF without replicating the input stages of a satellite receiver.

Wont I just see (in theory) a signal, except raised up on the Y axis by 12 V instead of being centred around 0V?

[GTR2Fan]

It's worth pointing out that this cam about as a a result of a random conversation with someone handy at school about satellite dishes.... the "other" more simple radio telescope for the roof of my shed is still going on, albeit in parallel with this one which is now a school lunch club project.

Aha. Two similar but separate projects. Now I understand.

[GTR2Fan]

Realistically, you won't be able to successfully separate the DC and IF without replicating the input stages of a satellite receiver.

Wont I just see (in theory) a signal, except raised up on the Y axis by 12 V instead of being centred around 0V?

I doubt it's that simple, but it may be. I'd have thought that you'd at least need an inductor in series with the supply to raise the load impedance for RF signals or it's going to be sunk by the low impedance power supply rail. It's going to be very hit-and-miss without recreating the exact load an LNB is designed to run in to. I'll be interested in your findings.

[stevend]

I doubt it's that simple, but it may be. I'd have thought that you'd at least need an inductor in series with the supply to raise the load impedance for RF signals or it's going to be sunk by the low impedance power supply rail. It's going to be very hit-and-miss without recreating the exact load an LNB is designed to run in to. I'll be interested in your findings.

Not far off. The input of a satellite receiver has a "bias tee" - basically an inductor to feed in the d.c. power, and a blocking capacitor to remove the d.c. from the signal. You can buy these ready made from people such as Minicircuits, or you can make your own - for the inductor something like 5-10 turns of 0.5mm dia enamelled wire about 2.5mm in diameter (use the shaft of a 1.5mm drill bit). For the capacitor something less than 10pF should do; ideally one intended for high frequencies.
LNB output frequencies typically start around 900MHz, and can go up to 2GHz or more. If you were decoding "commercial" satellite signals, a standard DVB receiver board could probably be located. Otherwise you could try and find a cheap "down converter", which can be tuned to pick out part of the LNB output spectrum and translate it to a lower frequency - typically centred around 70MHz or 140MHz. Maybe one of the software-defined radio sticks would do that.