Sunday, April 3, 2011

Simple direct conversion receiver - MRX-80

I have built my first analog receiver! This is maybe one of the simplest ones you can build ( a crystal radio might be simpler depending on how you count). The direct conversion receivers basically mixes the incoming signal with the local oscillator and the result is within frequency for our ears. The classic basic combo of ICs, NE602 and LM386 are used. The circuit is from the book "More QRP Power" from ARRL and is called the MRX-80. As you might guess it is for 80 m. 

The build is done inside an old box used for an Ethernet network hub. The BNC connection hole was already there but I had to to use a side of PCB material so I could mount the pots for gain and tuning. Yes, talking about the tuning, it is somewhat limited. I can change crystals and a diode is used as a varactor so a small swing of frequency is allowed, but I have not yet measured how big this is, maybe 1 kHz or so. The build is a mostly with leaded parts built ugly/Manhattan but all resistors are tiny 0805 smd, just because I wanted to try and see if it works. Maybe it does or probably not completely.  The build is not the prettiest, but if you see any bad choices of building other than aesthetic, please comment or if you have any other comments.

When I turned it on, the first thing I did was to measure the voltage out from the regulator. OK! I then turned my small Degen 1103 radio on and listened for the local oscillator. It was loud and clear so OK on that too! Then I connected my signal generator, but nothing heard, apart from noise. I started tapping on some connection and I heard them. So I guess the audio amp was working too.  Hmm, I started double checking other connections and soon found that I had connected earth connections between antenna and gain pot but not to the real earth on board. Then my test signal could be heard. When put in S9 mode it was OK, but not in S1. I wonder about what to expect from this simple thing. The real test will of course be when I try it with a decent antenna and see if I can hear any real on the air signals. 

I might have heard some today, or maybe not, because I never checked if they were present on another receiver. Next weekend I will go out to a 'quiet' place and listen. 

Below is a audio recording. Some signal can be heard, but QRM level is high so much noise. It does not sound as what I am used to what a radio should sound like. Either this is what a direct conversion receiver sounds like or I have some construction errors. I will bet on the last one. 

The sum up, a good build I think. Fun to do and easy. First small steps and more experience in radio homebrewing!

Update 2011-07-15:
The receiver has worked fine and the signal on the recording is SL0FRO by the way SM5-1252 Ullmar told me. 


  1. Well done Eric! Yes, I mean it!

    (Grrrr, I had a nice long comment written, and when I pressed post, an error occurred... I try to write up my thoughts again)

    Just grabbed my copy of "More QRP Power". Looking at the diagram, your receiver just does what could be expected from the schematics.
    You wonder about the amount of white noise that can be heard in your recording. There is nothing wrong with that. The schematics show that the NE602 front-end/L.O./mixer is wire up to the LM386 audio amplifier by a capacitive voltage divider only. No audio filtering is present. A simple RC (or LC) low pass filter just in front of the AF amplifier will solve that issue. You will find online calculator on the inet.

    The other issue could be lack of sensitivity. That is due to the simplicity of the design. The Gilbert cell (that's the NE602's mixing circuit) can easily overload in direct conversion when the L.O. level is too high. This is a common effect if the NE602's internal oscillator is used. To overcome this, more complex designs make use of an external oscillator with properly adjusted level. In the book you mentioned, you will find design by Dave Benson (K1SWL, ex NN1G) which do just that.
    The other remedy would be a big antenna creating lots of signal.

    Your construction looks great! Dead bug style on a solid ground plane... best practice.

    73, Joachim pa1gsj

  2. ps:
    You could, for wider band coverage, replace the crystal by a 3.58MHz ceramic resonator. The ceramic resonator will be less stable, should not be a problem for regular reception.

  3. Joachim, thanks for the input!