I'm completely new to amateur radio, so please forgive me if this is a silly question.

I'm looking at these devices: Assembled HAM RADIO 40M CW HF QRP Pixie Transmitter Receiver 7.023-7.026MHz DIY on eBay

What I would like to know is how feasible is it to use these CW transceivers to create a (low baud) data connection over, say fifty miles, between two Raspberry Pis?

I appreciate that these transceivers are the absolute cheapest that I could find, but this is the point: I would like this project to be dirt cheap.

If you know of any links or guides that you could point me at, I would be most appreciative - if you can think of a better, cheaper, way to do this then I'd love to hear about it!


  • $\begingroup$ CW is probably not a good choice for a reliable data connection. There are programs (like FLDIGI) that can copy CW, but I wouldn't say they do an especially great job. $\endgroup$ – Phil Frost - W8II Feb 29 '16 at 14:09

A Raspberry PI (preferrably a model 2) could run "fldigi", a free open source digital mode software. You might have to rig up some form of CW keying also

Fldigi can run a multitude of digital modes including CW, PSK and Olivia. It can also be programmed to transfer email, formatted and pre-canned messages both in keyboard to keyboard and automatic mode. Each transceiver would require a bidirectional audio interface, such as a low cost USB sound dongle. Pi's only come equipped with audio out.

These non-CW modes rquire a SSB transceiver though.

The low cost "Pixie" transceivers only do CW, but could be "keyed" with an cheap transistor using the circuit included in the fldigi documentation.

Most SSB transceivers are somewhat more expensive, but by using the more advanced digital modes instead of CW, you could greatly improve accuracy and speed in difficult propagation conditions.

You might also look at the low cost (~$30) Baofeng (or similar) 2m/70cm HT's that could probably transmit in MT-63 mode with only a microphone and speaker on the Pi held close the the HT. Some FM repeaters might have sufficient bandwidth to give you 50 mile coverage. Otherwise a yagi directional antenna and tower could provide a strong enough signal without a repeater.


| improve this answer | |

I see no reason why you could not use these devices as some form of low cost CW link - assuming that you can configure fldigi (it's free - and it looks like you are being frugal).

The main issue I would have with this is the VERY LIMITED spectrum they can operate in. Whilst I am UK Ham - I live overseas, and the chance of 2 Khz of the 40M band being free from other users is not realistic.

You may also have propagation issues - in some respects the 60m band could be better with NVIS propagation.

I do not see what you have to loose.... sounds like loads of fun/learning/development for absolutely nothing. Wish I had thought of it

| improve this answer | |
  • $\begingroup$ Although possibly rude, isn't it completely legal to use CW in the non-CW portions of most amateur HF bands? $\endgroup$ – hotpaw2 Apr 28 '16 at 20:52
  • 1
    $\begingroup$ Yes CW does seem to be allowed in all parts of the Band - but I would still try and avoid the RTTY, Digital, DX SSB (endless list) - and stick to the CW allocation. You are less (he said laughing) likely to upset someone. $\endgroup$ – Tim Seed May 9 '16 at 13:58

For one thing you'll need a receiver on each end. One problem with receiving CW mechanically (I.s., not by humans) is that it's very sensitive to noise. You'll probably be unhappy with the result. VHF might be an option in some parts of the country (50 miles is quite a haul for qrp.) You might be ucky to find a couple digipeaters between end points but if they aren't yours, YMMV.
But most of all, have fun!

| improve this answer | |
  • $\begingroup$ Receiving computer-transmitted CW by computer is less of a problem than receiving human-transmitted CW by computer. A lot of the difficulty in accurate computer copying of CW comes from the fact that humans, even the best operators, don't use perfect transmission length and spacing. An experienced human CW operator can compensate for this, but a computer has more difficulty because it can't easily infer things about the transmitted text. $\endgroup$ – user Feb 29 '16 at 10:52
  • $\begingroup$ If the signal can be guaranteed to be either on or off, then yes, it's not horribly difficult to decode CW by computer. The trouble is that CW has no "carrier detect" so you can't always tell if it's noise or a signal, or the end of an element or just a signal drop-out. $\endgroup$ – Duston Mar 6 '16 at 0:10
  • $\begingroup$ With a sufficient S/N ratio within the receiver passband, CW does degenerate to detecting a simple on/off state. With insufficient S/N, copying CW by any automated means becomes fraught with problems, but that's true for every transmission mode. $\endgroup$ – user Mar 6 '16 at 22:36
  • $\begingroup$ Decode might not that difficult. Given a precisely known audio frequency (given the crystal controlled transceivers plus possibly a temperature measurement) and a precisely known WPM (calibrated in software), one can use basic DSP algorithms (such very narrow band filtering and running a matched filter on the envelope of a long enough prefix) to detect and decode the incoming Morse Code messages. A Raspberry has plenty of compute headroom to run this much custom (write you own) DSP code in real-time. $\endgroup$ – hotpaw2 Apr 28 '16 at 21:02

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.