Ask Hackaday: Has anyone built a radio telescope?

[Michael] sent in a question regarding the latest advances in software defined radios available for $20 on eBay:

I’ve been looking in to SDR lately, mainly for the possibility of using it for incredibly cheap radio astronomy. So far all I’ve found are whispers. I’m 18 and have very little experience, but I figured you might be able to help me find a little more info.

This really brings me back, [Michael]. I saw Contact in the theater (surprisingly, a rare case when the book and movie are equals), and in my childish exuberance went out and listened to lightning on Jupiter. The financial difficulties of expanding my setup meant the experiments stopped there, but at least I knew amateur radio telescopy was possible.

The latest and greatest advance in software defined radio – namely, a $20 TV tuner dongle – brings something new to the table. Instead of the thousands of dollars in gear that was required in 1997 when I last looked into this, it’s possible to set up a  passable radio telescope for under $100.

I’ll leave it to the Hackaday commentors to fill [Michael] in on the details, but here’s my suggestion:

Optimize your setup for 1420 MHz. There are three reasons for this: firstly, very few things in the universe absorb radio waves at a frequency of 1420MHz; there’s a reason it’s so often used in radio astronomy. Secondly, most government agencies around the world ban (or at least don’t look too kindly upon) transmitting on 1420 MHz. This frequency has been somewhat protected for use by astronomers. Thirdly, most of the Realtek TV tuner dongles have a frequency range of 64 – 1700 MHz, so it’s possible to receive 1420 MHz with this minimal setup.

As far as antennas go, your best bet is probably going to be one of those old C-band parabolic antennas from the 80s. That will make your telescope highly directional and give it a huge amount of gain. There is the problem of having a 20-foot-wide eyesore in your back yard, however. Alternatively, you could use a smaller DirecTV satellite dish, but I’m not making any promises with that. It’ll work, but it’s too small for an optimal setup.

I’ll concede the floor to anyone who has additional information. If you’ve built a radio telescope, send it in and I’ll put it up.

83 thoughts on “Ask Hackaday: Has anyone built a radio telescope?

  1. Secondly, most government agencies around the world ban (or at least don’t look too kindly upon) transmitting on 1420 MHz. As would probably most of aliens’ governing agencies ;)

    1. I don’t mean to be a hater, but i don’t think you know how a radio telescope work. when you use an optical telescope you don’t “broadcast” light into space. With a radio telescope you passively analyze the radio waves from a very very small section of the sky. Most of the time it is useful for scanning for new planets and stars. (e.g. a neutron star spins super fast, the EM profile of this is pretty unique.) So the government really wouldn’t care.

      THAT BEING SAID: most of the time Radio Telescopes are pretty broad band.

      1. My take on his comment was that it may also be rare for aliens to be broadcasting on that frequency and therefore there would be no little green men to listen to. I don’t think he was implying that anyone, alien or otherwise, would get upset at setting up a radio telescope.

        Just my 2 cents. :-)

      1. By the way: now i am using my own SETI-receiving station at home.
        I published all software on my site, so everybody can make his own
        SETI-receiving station now. You can use a TV-disk and a usb-receiver. or
        a short-wave-communication receiver or whatever you have.
        Every frequency without terrestrial signals will be okay.
        An old xp computer does the job.
        Just for the fun

  2. Most C-Band satellite dishes are 10 feet in diameter, although ones as small as 8 feet (central of the U.S.) and large as 15 foot (capable of handling trans ocean satellites or Alaka and Hawaii residents) are available.

    While some folks consider satellite dishes to be eyesores, techies tend to see them as very beautiful tools.

    As a general rule you don’t want to hook up a microwave antenna (including 1420 Mhz.) directly to a receiver, there’s too much loss over any distance. It’s generally better to use a downverter mounted at the antenna which reduces the frequency to something more manageable with reasonable coax cables. Of course one of the key advantages of an SDR is its wide frequency range.

    I’ve got an SDR on order and I’m certainly interested in using it for radio astronomy. I am concerned about the gain and signal-to-noise ratio which will probably be pretty bad, but an external preamp and downverter will improve the performance.

      1. People used to do this with wi-fi dongles to increase the range so I would think yes, you could do the same with a radio telescope. You’d have to figure out some way of running such a long USB cable though, unless you had something like a Raspberry Pi sitting nearby to convert the USB into Ethernet.

      2. USB is very short regardless of cable quality. You would have to sit out by the dish with a laptop. “malvineous” has a good idea with the conversion to Ethernet. Would a Raspberry Pi have enough power to do this with sufficient bandwidth though?

        Another thought.. so long as some sort of computing device is being placed near the dish like that.. what if the backend sdr code could just run right there and only use the network for the display and/or playback? Could a Raspberry Pi handle that? I know the Raspberry Pi is very low spec compared to GNU Radio recommendations but I assume the recommendations are including processing power for the display, window manager, etc… If what was running out at the dish was completely cut down, just a minimalist OS with the SDR backend on top might it work?

      3. gnuradio allows you to send the stream from a computer to another, so you can digitize it near the dish with the Rpi, send it through ethernet and on a beefy PC do the rest of the heavy signal analysis. I assume the Rpi is more than enough for this; I guess I’ll know when I receive my Rpi soon and try it with my rtlsdr dongle :)

      4. Use [whatever] to downsample the Rf at the antenna. Then serialise the signal – and light up a LED. Fibre optic cable, photodetector at computer end.

        If you are only listening to one band, you don’t need to send “change channel” signals. You get isolation with optic fibre, which is missing from standard ethernet solutions.

    1. Oops, sorry. Please disregard that “Report Comment” as I hit that instead of the “Reply” button. -_-‘

      Anyhoo, yeah I still tear up at the end. Carl died during film, didn’t live to see the final cut.

  3. This is very interesting. I haven’t built a hardware radio telescope myself, but I’ve used Arecibo and the VLA for pulsar astronomy. Most relevantly, there is MIT’s “Small Radio Telescope” project. We actually have one on the roof of the physics building, and I worked with some undergraduate students to get it working.

    What you can do with such a setup is principally limited by sensitivity. Practical dishes are really small by radio astronomy standards. Nevertheless, with the SRT you can, for example, map the rotation curve of the galaxy (combine this with a map of the visible matter and you can show that there must be dark matter!). Pulsar observations, though dear to my heart, are going to be a real challenge. A little easier if you’re far enough south to see the Vela pulsar (by far the brightest radio pulsar), but you’re still talking about hour-long integration times, folded with a known period for the pulsar, to see anything.

    To get a working SRT-like setup you’d need, approximately:
    * A dish antenna
    * An altitude-azimuth mount for the above
    * A low-noise amplifier for 1420 MHz
    * A hardware band-pass filter
    * A noise diode and maybe some flaps to close the feed, for calibration
    * The RTLSDR gizmo
    * Control electronics
    * Software

    Unfortunately, the biggest limitation you’ll run into is the dish. The SRT uses commercial ~2 m dishes, which are already big and heavy enough to require a fairly expensive alt-az mount. If you could get your hands on one of the bigger (4m?) dishes, you’d gain dramatically in sensitivity, but the mount would be correspondingly heavier and more expensive.

    Once past that hurdle, though, the rest is fairly straightforward (at least by comparison). You can buy LNAs off the shelf; you might want to combine yours with some kind of cooling system, even just a Peltier unit. A band-pass filter you can probably also buy off the shelf; the point of it is that there’s *so* much interference these days that it’s best if you can keep it out of the ADC. The control electronics for the alt-az mount and anything else are easily built using an Arduino.

    For the software, if you have something like GNU Radio, you’re pretty much golden. Almost all radio astronomy measurements amount to simply computing the power in a fixed bandpass over a fixed time. For “continuum” observations (measuring the Sun’s brightness, for example) your bandpass is as wide as you can manage and your time period is usually seconds long. Spectral-line work uses narrower bandpasses but similarly long times. All this is easily implemented within GNU Radio.

    If you want to do interferometry, you’re going to want at least very stable VFO/clocks, and ideally shared VFOs between multiple receiver modules. Probably doable but it might require some hacking on the dongles.

    On the other hand, since the dongle is *so* cheap, you might have an interesting time building an array of crossed dipoles, with an RTLSDR and a Raspberry Pi on each one, to serve as a LOFAR-style interferometer. Using lower frequencies also gives you higher fractional bandwidths and almost all sources are brighter, so this might be a good way to go.

    TL;DR: Perfectly doable, but I don’t know that you’re going to save much compared to the SRT’s ~$10000.

    1. More on meteors. I bought an ICOM IC-PCR1000 receiver a few years ago and ran into a guy who was using multiple units to track meteors to recover ones that reached earth. Still have my PCR1000, because they go for <US$100 on eBay now and it is not worth my trouble to sell it. Ditto an AOR AR-3000A (the original "DC to daylight" receiver). Maybe I will find a good hacker space to donate these to. Have a suitable dongle on order, to play with lower frequency SDR.

  4. Oops!

    One or more yagi or helical antennae are both inexpensive and visually unobtrusive especially at 1420 MHz.

    One option would be to use a fixed mount and let the earth’s rotation do the scanning.

    Antennas, John Krauss (invented helical) is the best reference I can locate at the moment. I’ll try to remember to look for other references.

  5. How about some radio interferometry???

    Simple, could be done with 2 receivers and 2 phased dipole antenna. Some modification would have to be done on the cheap USB receivers so they share the same LO, or some other way to synchronize the clocks.

  6. i have no clue about this, but why not build a large community driven array to “simulate” a huge antenna? With like GPS as time reference (?), couldn’t you somehow use the data received by multiple cheap SDR + small DirecTV antenna setups that are located all around the globe driven by hobby astronomers? (i would be surprised if “they” don’t already do that)

    1. I know the bigger telescopes(Arecibo and such) are all linked together, basically making the whole planet a rotating radio telescope, albeit with some blind spots. I think an array would be the way to go here. If a small(<4m) dish was used, and the cost kept low, then putting multiple setups in your yard could make a theoretical dish size that would be rather large.
      This is something I've been interested in for a while, and I've done quite a bit of research. All these latest SDR posts never even connected in my mind to radio telescope… Great question Michael!!

      1. I like where your are going. but with an array comes expense. Combiners and phase shifters are not cheap. This would be a monster project to build an array. The smaller dishes mean that you need to have more antennas… also if you want to have a workable array you need to have at least of 100feet between the edges, as well as a 2D array. The VLA has a radius of 13 Miles… this is not ideal for a hobbyist. I attended a presentation at a conference where the presenter was discussing using several antennas from around the planet to make the antenna array. If a large, evenly dispersed group wanted to do such a project , it could be possible. This would require a very accurate clocking system and a central data base to collect the data. By matching the time signatures and manipulating the phase data, one could retroactively re-construct an awesome antenna. The problem would come in the making sure that all antennas are pointed toward the same point in the sky.

        But all in all, It is going to be expensive.

      2. Great idea, great discussion. Why to limit the array setups to our yards only? What about an array distributed over several continents?

        The large array could consist of cheap (< 250 USD) sets with relatively small antennas (i.e. 1m) installed and maintained by enthusiasts around the globe. How could the clock be synchronized? Is the only way to share the LO of the receivers or is there another solution using time reference like GPS?

      3. You absolutely can use GPS time transfer for radio astronomy; it’s pretty common in VLBI. But you need a very stable timebase; a commodity PC isn’t going to be good enough. And the USB that most SDRs use probably confounds the timing enough to make the whole idea unusable.

  7. I have helped build a radio telescope onto the roof of a high school a few years back, but it was in the thousand dollar range and was also about 18-feet in diameter, so probably not what the poster was looking for.

  8. If [Michael] happens to live anywhere near me, I’d be happy to donate a recently decommissioned 10 ft. C-Band dish. :)

    I just can’t seem to force myself to scrap all of those aluminum parts yet.

  9. Dear Rigid,
    About your suggestion on interferometry with a
    gps controlled timing.
    Gps is not anough, as far as i know.
    Each interferometry station needs more time precision.You can only do that with a rubidium/ceasium clock, which is gps controlled.
    I do phasing interferometry on 31 MHz now for 2 years, i can make maps of the sky with it.
    If a worldwide network of amateur interferometry stations will start, i would like to participate!
    My results:
    Look at the third experiment: phasing interferometry at 31 MHz

    1. LOFAR uses lots of antennas. I do use only 2 antennas ( and 2 coherent receivers), but put one antenna every day at an other position, following a regulair grid.
      A 2D-FFT gives for every measuring moment ( recalculated to MST time) a map of the whole sky…..
      It’s the other way around, but it gives real maps!

      1. In my maps i see Cassiopeia-A, Cygnus-A, The Milkyway and Taurus-A

        a disadvantage is: it took one year of measuring and gave about 2 TerraByte of .wav files.

        a .wav file contains about 1 000 000 numbers for
        one measurement of 35 seconds per receiver. I do every minute a measurement.
        So i can make 24 * 60 maps of the sky….

  10. So has anyone any suggestions on how to connect the USB dongle to the LNB on a sat dish? The LNB requires a DC power supply over the coax (13/18VDC) so without some method of injecting the power, and filtering the signal out so the dongle does not get fried people are going to have a less than ideal experience.

    1. To inject power into the LNB you just need a filter that separates DC from RF. Making one is simple, though at those frequencies components quality and placement is crucial. You also must keep connections really short. Forget about using breadboards or general purpose PCBs; the circuit is so simple you can solder it directly on its connectors, making it both faster to assemble and better performing.

      Take a look at this schematic for example.

      The inductor value should be very low, well under 1 uH, and the 1nF capacitor must be a good quality ceramic one. I would add a small 10 pF one in parallel to it.
      The potentiometers are not necessary and can be omitted, just connect the output directly to the capacitor.

    2. Maybe you could also use the power injector from a TV masthead amplifier, since those have an aerial connector on them you can plug into the USB dongle (instead of your TV.)

    1. YES, we just all need rubidium clocks calibrated with a GPS DO, all of which can be found on ebay.
      Open-LOFAR, using the internet to relay data, maybe with a distributed computing client similar to SETI.

      I have a rubidium waiting for this experiment, finally time to get a GPSDO (gps disciplined oscillator).

      BTW, if you want to learn about atomic clock and meet some time freaks, check out the time-nuts webgroup, some really amazing people.

  11. I’ve used a standard satellite dish to detect the sun at 12 GHz or so. You can just use a satellite TV LNB connected to a “satellite finder” (box which emits a noise or has a dial when there is a strong signal). They’re very cheap. You can detect the sun, ground, people and maybe the moon (reflected radiation from the sun), plus satellites of course.

    I did try connecting a Funcube Dongle SDR to the setup via a power injector, but the solar signal was just too wide band to see anything. I did see some line emission from satellites, so I think the setup was working.

  12. This is one of the intended projects with my rtl-sdr stick.
    In fact, the machine I will be using for radio is named “ellie” (after Eleanor Arroway, the character portrayed by Jodie Foster above).

    Bookmarking all this so I can reference it once I leave the office. Thank you all!

  13. Another option might be to use hardware from the HAM world – a 2M Softrock Ensemble SDR receiver paired with a microwave transverter based on the designs from W1GHZ –

    Given the range of frequencies that have been used with the W1GHZ transverter, designing one for 1420MHz shouldn’t be too bad, he explains the calculations and theory in his writeup. Down East Microwave (DEMI) has plenty of parts for transverters and other microwave devices.

    For open source SDR software QtRadio is pretty awesome (and works with the TV dongles), you could share your radio telescope over the internet with other interested users! Many of us are using QtRadio with various SDR hardware covering HF, VHF & UHF.

    There is also an open source sound card optimized for SDR work – the sdr-widget

    A number of cheap options are popping up in the SDR and microwave areas, I can’t wait to see where things go in the next few years. I’m currently working on a 35W all band (160-10m + 2m) SDR based on the Softrock 6.3ng, MOBOKit, sdr-widget, Pandaboard ES, and a few other components – most of which are open source –

  14. Hey guys,

    Thank you so much for all the new information. I didn’t think I get such a good response. I’m getting a job pretty soon so hopefully I’ll have the money to do this.

    I’d probably be taking the DirecTV satellite dish route. Due to all the foreclosure around here (Florida) there are a lot of them available, so I could probably get an array put together later on in the future.

    One of the main question I have is, would there be a way to create a mapping software that converts the radio data into images, sort of like weather stations do with clouds? It would be cool to actually SEE what I’m picking up rather than just hearing it.

    It’d be great if we could get a decent amount of people around the world to put one of these cheap RTL-SDRs together so we could start listening to the skies ourselves rather than relying on major organizations to do it for us.

    1. Standard single-dish telescopes are just big single-pixel cameras, so you can build an image just by scanning them. You want to be kind of smart about how you do it because they drift and you want to save calibration time, but it’s basically not too complicated.

      Interferometry is more of a black art, but on the plus side the same software we use at the big interferometers is open-source and no worse to use for a small hobby interferometer.

  15. Dear Michael,
    I think a simple way to make a map of the sky with
    a disk antenna at GHz frequencies is:

    Put the antenna pointing to the south with an
    elevation of 20 degrees.
    Do measure 24 hours.
    You get now a line with maxima and minima.
    This is one line of your image ( map)

    make the elevation 23 degrees and do measure again for a whole day.
    Now hou have a Television picture of two lines.

    And so on…. until you have a map… do choose
    the elevations etc after a lot of thinking about the properties of your disk..
    But this is the principle of making a map
    with a disk antenna ( i think.. i do not have
    experience with it…..)

    Good luck

    1. 25-10-2012
      Well, i did the experiment now myself with a
      1.5 m TV disk and an LNB for 12.5 GHz.
      After that a AR5000 receiver of AOR.
      The sound of the receiver was connected to
      the soundcard of an XP computer.
      I wrote a program to use the soundcard as
      digitiser and so a could do the experiment
      i mentioned above.
      The only thing i could see was the sun…..
      So this setup is not sensitive anough.
      See my site for more information:

      look at experiment 4.

      Wim Apon

  16. Michael,

    First thing I do is work on your amateur radio license — technician then to general then to advanced — find a local ham club — find a mentor in that ham club — go on a fox hunt — all of these things mentioned above will help you in your adventures — I have the se passion about space — telescopes — radio — and building things
    Read up study and have fun with this !!

  17. OK, so I’m a grad student in astrophysics. I specialize in building radio telescopes, particularly high frequency ones (100-3000GHz). Since you’re only receiving, I don’t believe you need a HAM license (although having the knowledge would be helpful, I’m sure).

    eBay is your friend. My friend, too (scientists don’t have too much money, either!). If you know what you’re looking for, you can get enough hardware for a receiver pretty cheap. I ran a simple interferometer (looking at GPS satellites) using off-the-shelf parts and such that probably cost about $500-1000, total (and could have cost less).

    Low frequency dish antennas need only be as accurate as one tenth/twentieth of the wavelength. So if you’re looking at the 21cm hydrogen emission line, your dish can have surface inaccuracies of up to 2 cm! So you can build your own dish out of plywood and screen door mesh, and it’ll work great.

  18. The statement that “very few things in the universe absorb radio waves at 1420 MHz” is the exact opposite of the truth; since the 1420 MHz line is emitted (and, by reciprocity, absorbed) by hydrogen, the majority of the mass of the universe absorbs radio waves at 1420 MHz. But by extension, most of the universe emits there too, and since space is pretty empty, that makes it a good place to observe.

  19. Anyone looking for some movie-based inspiration on building a long baseline array should check out the “The Arrival” starring Charlie Sheen.

    I won’t put in any plot spoilers, but Charlie’s character is a radio astronomer who he gets a job with a satellite TV company at one point. He uses his job as cover to offer some subscribers free “upgrades” that allow him to remotely take over their TV antennas during the early hours of the morning and use as a long baseline array.

  20. If you check the back issues of Sky and Telescope, starting in about May of 1978, you’ll find a series of articles on how to build An Amateur Radio Telescope. This design was a six yagi phase switching interferometer at 73.8 MHz.

    I worked on this project as an undergrad and had a ball. It turned into a multi-year project. I did get it to work, but there are many errors in “the back end” that were never corrected with errata published by Sky and Telescope.

    Good Luck.

  21. You can use any of around ten different low cost GPS chips that are available today to do very accurate positioning using RTKlib. If you tap off their 1PPS output from the GPS module (its a very short pulse – the rising time of which signifies the beginning of each new second) you can also use that directly to discipline your NTP clock – you can also split and distribute that 1PPS around your lab and use it with your NTP in numerous ways to determine the exact time of arrival of signals. Having very accurate time measurement (and also knowing *exactly* what your position is) is a hugely useful thing.

  22. Its exciting to see so much discussion going on about amateur science. With financial aid and parents wages falling and the cost of graduate and postgraduate education being what it is, huge, a large and growing number of potential scientists in the US are being prevented from ever entering their chosen career path – poisoning the economic future of our nation.. a nation that really needs them – by their lack of formal credentials. (One needs a PhD to get a stable job in the sciences and it seems as if at least an MS is needed now to get unpaid or internship type positions helping scientists do research, supporting research organizations computing needs, etc.)

    But, amateurs can do science, and every once in a while, amateurs make real discoveries. Before the current system came into being, all scientists entered the field by self-teaching themselves.

    Newton, even, became interested in physics through of all things, alchemy!

    As the yields on many investments is so low that many parents can’t afford sending kids to school any more, wouldn’t it make sense to abandon our current system of funding education with saved wealth, and borrowing, and instead, funnel some of that huge “black” and “defense” budgets (more than the entire rest of the world combined) into educating our young people in science and engineering instead of hard knocks?

  23. Rather than spend money on your own hardware (antennas, feed lines, filters, amplifiers, receivers, computers), you could just send the cash to The SETI Institute and help fund some real peer-reviewed research. Or, if you have a seriously good idea, describe it to them and receive a grant. The Internet makes this kind of Big Science accessible to anyone.

    By the way, I’ve successfully listened for meteor ionization trails reflecting distant TV stations, but be careful of what you read. That technique quit working when our TV broadcasts converted to digital format. The NTSC analog signal had a strong purely sinusoidal carrier in the video luminance component, which could be heard as a peep or a ping, or sometimes even a howl, on a VHF SSB receiver when a meteor reflected it. But no more, alas.

  24. There is a way to use a radiotelescope without any costs!
    A fantastic initiative of Onsala Space Observatory in Sweden.
    Log in to there computer, direct the 2.3 m. disc ( on 21 cm wavelength) and
    do measurements. Download your data and make maps.
    All needed programs available on there site:

    There is a webcam so you can see the antenna move when you
    give the commands.
    Its great!!!

  25. Im sorry the Radio Telescope I built is to big to send any ware. It rotates 360 degrees and you can alter the pitch angle of the dish. It is all remote controlled from my laptop. But the thing takes up my entire garden. I can get signals from Andromeda though which is nice. Some signals I have received with my dish have even suggested that Contact (the film) was not all science fiction and SETI gave up to soon.

  26. This may be an old thread, but I have a telescope in the works for my senior project. The prominent option right now is a dipole-based interferometer. Anyone ever built an interferometer who can give me any advice? I’d appreciate hearing from you, I’ll watch the thread.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s