15 Aug 2012

Simple 28MHz Polyakov Mixer WSPR RX

As the first part of my WSPR transceiver project for 28MHz, I breadboarded a Polyakov harmonic mixer based, crystal controlled, direct conversion receiver this afternoon. It was a great success!

28MHz Polyakov mixer based WSPR RX
The circuit shows the idea implemented. A low cost GQRP club 14.060MHz crystal is pulled to 14.0623MHz in the oscillator/buffer stage. This is then injected into the back to back diode "Polyakov" harmonic mixer which switches at 2x injection frequency i.e. 28.1246MHz, the WSPR USB dial frequency. The 28MHz input from the antenna is filtered and directly applied (no RF amp) to the diode harmonic mixer with the audio output amplified and filtered in the 2 stages of audio gain. The output then goes to the PC sound card. I may improve the audio filtering and add a 600ohm line transformer between the audio output and the PC mic or line input.

Minimum discernable signal (MDS) is around -124dBm, possibly a shade better. AM rejection was tested by injecting a 100% AM modulated carrier at 28MHz. An input of -53dBm was needed before the AM signal was detected, so 70dB AM rejection, which I think is very good.

More bench work to do, then I'll try looking for 10m WSPR signals on-air with the RX, but for about an hour of work, a very fine result. Once the RX is air-test proven, the TX part will be added to make a complete WSPR transceiver.


DD5CF said...

Very nice idea Roger thanks for putting this on the web, I look forward to your next Blogs on your WSPR RXTX.
Vy 73 de Colin DD5CF / G1ZOS

Anonymous said...

Good work. I like those KISS designs. New for me is the tuned circuit (CCT) prior to the mixer.
Amazing that no initial 28MHz amplification is required.
Thanks Roger
73 Ron

Anonymous said...

Just a quick question, have you used a grounded base pre amp purely to match the lo-z of the mixer o'p/

You see i would'nt mind making a cut down version just for receiving amd i wondered if the oscillator buffer and one audio pre amp could go ?!!


YO5PBG - Adrian said...

-124dbm looks like 0,15uV??? So Roger, do you think this Polyakov design is better than a direct conversion RX for the same components count?
I'm curios also about spurs.. out of freq/band sigs... maybe this is hard to tell for a little bandwidth.
I'm standing by with lot of interests about your feeling about Polyakov design... Maybe you can use it also for TX - i've seen it used in transverters.
73 de Adrian - YO5PBG