PolyMorph Receiver

A minimalist direct conversion radio receiver. 4 general purpose transistors and a Polyakov 2 diode mixer.

Similar projects worth following
This is a project that anyone can build from recycled electronics. The goal is to keep components to a minimum while avoiding ICs and be able to build this receiver from almost any old electronic device such as a discarded tube TV set. The local oscillator runs at half the receiving frequency and is remarkably stable even without a proper ground plane or shielding. The VFO employs an 2n3904 NPN transistor instead of an JFET. The audio stage is also implemented using 2 2n3904s and drives well a set of in-ear earphones. I would like to make a transceiver eventually. All of Europe can be heard on the 20m amateur band with my ~5m random wire antenna from my home in Finland. Hawaii (11000km), Alaska (6500km) and Yokohama (7800km) have also been heard on the PolyMorph receiver!

Put everything in an enclosure
Make a transceiver with a bilateral RF amp
Improve audio filtering without adding much
Get a radio amateur licence - DONE!

  • Solder Haze

    morph05/24/2020 at 11:41 0 comments

    We have a working prototype! Almost all the components are scavanged from PCBs that were meant for a completely different purpose, it makes me happy that they get a new purpose in this radio receiver.

    Initially I built the VFO with salvaged varactor diodes as varicaps. It worked well and I could cover the 20m amateur band completely. I had two 1n4004 diodes back to back, and fed the control voltage at the cathode interconnection. Then I remembered that some years ago I had made this receiver topology with a ceramic resonator and I really didn't want to make this build about fighting VFO drift (which there certainly was). I'm trying to keep the components to a minimum and a trade off is that the VFO doesn't have a buffer stage. No problem though, super VXO to the rescue. Two 3.58 MHz  ceramic resonators in parallel promises to cover the whole 40m amateur band, currently that's what I can see on the scope, I haven't tried this in the wild yet.

    The diode mixer RF termination pot also serves as volume control. The amplification is pretty good for using two NPN transistors. In fact all the transistors are from the scrap boards, I had to twist the BJT legs around to match them to my board pads, this wasn't a problem. For anyone building this receiver from scrap, most of the component values are not critical and I'm sure mine aren't a perfect fit. For the transistors you should check online which ones are best for HF gain (RF amp and LO) and the others can be used in the audio amplification.

    I tested the receivers 'ears' with a si5351 signal and it feel very promising, next I need to find a scrap enclosure and take the Polymorph receiver for a walk. I will probably add some form of a simple RF bandpass filter, a good chance to show some impedance matching tricks. Below is a sa612 superhet transceiver prototype.

  • Board Design

    morph05/08/2020 at 17:24 0 comments

    Since I ordered my small kit CNC machine after I saw a blog post about it on Hackaday I've been fine tuning my rapid prototyping techniques. Above is the machined PolyMorph receiver pcb board. Currently I'm experimenting with the muppet style construction with my own twist; I keep everything very close and squarey.

    I start the design process on Inkscape. In the document properties I set the grid size to 2.54 by 2.54 mm, major lines every 4. I then make adjacent squares of 5.08 mm depending on what I want to lay on the board. For example on the far left of the pcb image there is a tetris T shape, that's for the base of a common emitter RF amp. The top long strip is the vcc. 

    Next, I open up JSCut in my browser, load the svg from Inkscape and generate some Gcode. Then it's time for Chilipeppr, still on the browser. Finally, voila a machined pcb!

    I've made quite a lot of boards this way. I love the speed of it once you get the knack of it. I think of a design and in 30 minutes I'm soldering components onto the board. One big advantage of this unorthodox method is that I can mostly fix any mistakes by extending the through hole components to where they need to go and split squares by hand, or a blade... my hands are not that sharp :D

    In the next update I will post some pictures of some scavenged components on the board and some component trickery. There'll some tips on how to get some serious 'pff' out of standard varactor diodes as varicaps and a relief for those of you who either don't have ferrite toroids on hand or would like to know another method of notorious e-imp-matching-dance courtship practice.

  • A New Build

    morph05/01/2020 at 11:09 0 comments

    Beginning of 2020 I decided that I'd start building electronics projects out of what I have and put ordering new parts on hold. It feels like now it's time to build the PolyMorph Receiver on a new board. I'll try to recycle as many parts as I can from old electronics that I have lying around. I'll test different configurations and tricks to make the most out of the sub-harmonic mixer. Nevertheless, the idea is still the same, anyone should be able to build this. I'll also document my construction methods.

  • Colpitts Oscillator

    morph01/09/2019 at 13:58 0 comments

    I added some pictures of just the colpitts oscillator of the circuit. In the picture the BJT is 2n3904 and it's collector resistor is 470 ohm, base to voltage supply rail is 150k ohm. 100 ohm and 100k ohm will work respectively. The supply voltage is 5v.

  • OH6UAV

    morph12/23/2018 at 18:25 0 comments

    One of the goals of this project has been accomplished, I'm now a licensed radio amateur! I got my license over a year ago and have a lot of fun on the HF bands. My current DX record is a contact to Suriname from Finland, ~9000 km on the 15m band with an 13.5 m end fed antenna a 9:1 unun and only 10 watts. My rig is an 'old' Kenwood TS-50, but it works like a charm. Timing is key on the air and so is chasing your dreams!

  • RF Amp

    morph05/22/2015 at 08:05 0 comments

    Added another common emitter amp, this time for incoming RF.

    The performance boost is immense, small signals are much easier to hear and large ones boom (CW and SSB).

    Unwanted outward RF leakage seems to have been greatly reduced.

    I use an atennuator pot to better hear stations, diminish AM breakthrough and generally control audio volume.

    Updated schematic and pictures coming soon!

  • Log

    morph03/10/2015 at 15:41 0 comments


    Uploaded modified schematic. There was a 100k resistor missing on the vfo transistors base. Without it the the vfo doesn't oscillate. The 1n4148 diodes seem to like a 1-1.2 peak to peak oscillator voltage for optimal performance. Currently I have a 4.7uH inductor as the vfo tank and a 1.5uH antenna tuning/impedance matching inductor next neck on neck with it (this is one of the places where the magic happens). The 120p antenna capacitor should be reduced to something like 20p to avoid hum/motorboating in the audio chain. especially when you use external speakers. Keep the audio transistor emitter connection AS SHORT AS POSSIBLE to ground. I'll put a demo video on youtube.

View all 7 project logs

Enjoy this project?



morph wrote 12/25/2018 at 08:31 point

G3XBM made a wspr transceiver using a Polyakov mixer. One difference between the Polymorph receiver and his design is the lo feed point. I have had success with both topologies but should make a comparison between them.

  Are you sure? yes | no

aurelw.wiz wrote 01/03/2019 at 15:48 point


I still have problems with this oscillator colpits

So i am looking around and found one oscillator with

common base for micro SDR receiver it use PNP transistor

so i will try that way..

  Are you sure? yes | no

aurelw.wiz wrote 01/08/2019 at 16:15 point


Do you can explain .

In one schematic you have collector resistor 100 ohm

and in another you have 10k ..

so which one is proper?

Also i still don't have luck with your circuit...

simply not work , oscillator oscillate when i use it as ordinary regen

but nothing with diode mixer...

maybe i must wind antenna coil over oscillator coil ?

  Are you sure? yes | no

morph wrote 01/09/2019 at 12:57 point

You might be referring to the addition of the 100 kilo-ohm resistor in the newer schematic, that is necessary for lo oscillation. I will delete the picture where I forgot to add it.

Can you get my lo to oscillate with only the lc tank coil? Try to get it working before adding any other part of the circuit.

I have been playing with this topology in the past days and will hopefully soon post something new.

  Are you sure? yes | no

aurelw.wiz wrote 01/08/2019 at 16:16 point

Also i don't have such a varible capacitor .

Anyone else maybe build this receiver ?

  Are you sure? yes | no

aurelw.wiz wrote 01/09/2019 at 19:43 point

no i dont tried with just LC tank i always use emitter caps

and fixed emitter resistor of 1k. I can try that .

I also tried with 2N3904 and nothing.-

Also tried to wind antenna coil over VFO coil and again nothing.

Only case when work is like a ordinary regen with

output on collector, trough cap 22n and RFC 4.7uH.

  Are you sure? yes | no

aurelw.wiz wrote 12/24/2018 at 19:37 point

In last day i think that i found what is a problem.

Because i us 3V it looks that general purpose transistors refuse to oscillate on such a low

voltage so i use good old BF199 and i finally get oscillator to work.

On my small speaker i heare typical "hiss".

I use modifayed JLH design for AF amplifier.

Another thing ...when i push ferite rod inside coil i can change and pick up some stations.

AF signal is very low and propbably i need one more preamp stage

or better to use RF preamplifier.

Just to say my oscillator use 220pF caps in emiter devider and

bias is adjusted with pot on transistor oscillator base not on emitter.

schematic is here:

all best

  Are you sure? yes | no

morph wrote 12/25/2018 at 07:48 point

Very good, you are now in the experimentation zone! You could try a higher supply voltage. The rf preamp amplifies af, I've used a common emitter there as well successfully. Interesting, I need to try adjusting the lo power with a base resistor. Can you post your circuit diagram somewhere that doesn't require a login?

  Are you sure? yes | no

Dr. Cockroach wrote 12/23/2018 at 20:23 point

Very interesting circuit. I have been away from ham radio for several years but always enjoyed building my own regen or dc receivers. I have not had the chance to try a Polyakov circuit so perhaps I need to get back in the game and give it a try :-)

  Are you sure? yes | no

morph wrote 12/24/2018 at 16:00 point

Building radio receivers is indeed a lot of fun :) Give the Polyakov mixer a try, there is a lot of material online. This project is made with 'junk' components and the tolerances are very loose. If you don't couple the lo and rf inductively I recommend feeding the lo to the audio side of the diode mixer. Its quite interesting to use a 3.58 MHz ceramic resonator as a vxo. You need to have resistance to ground at the antenna side of the diodes for impedance matching if you omit the rf inductor. I've also added a 1 nF capacitor in parallel with the 100k resistor in the first audio amp, this filters and stabilizes operation so you can drive the AF chain harder while having less adjacent signal breakthrough and high frequencies.

  Are you sure? yes | no

aurelw.wiz wrote 12/21/2018 at 18:32 point

I don't have 2N3904 so i use 9016  but it not oscillate,then i treied

some others and still nothing ?

what might be a problem?

  Are you sure? yes | no

morph wrote 12/22/2018 at 15:00 point

It might be a good idea to first test only the Colpitts local oscillator circuit by itself. You can test different coils and values with your transistors. An oscilloscope really helps. Then when you have oscillation you can design the lc tank for your target frequency. The lo drive amplitude needs to be adjusted for different diodes. Fast switching diodes have usually worked best for me, but I got the receiver running with two red LEDs also (the sensitivity was low). Anyway, try the standalone Colpitts oscillator circuit first. I'll post some updates to this project next year when I'm at home. Now at the cottage :D

  Are you sure? yes | no

aurelw.wiz wrote 12/18/2018 at 19:30 point

Just a question about variable capacitor in VFO.

It looks to me that receiver need two ganged variable cap

or i was wrong.

Which one is main tuning capacitor?

Also i see that 20+245pF could be replaced with 18pF ?

  Are you sure? yes | no

aurelw.wiz wrote 12/21/2018 at 16:24 point


Is small orange coil antenna coil?

And which variable cap is main tuning?

I use for test ordinary plastic varible cap and one side

connect to antenna coil,and other side to oscillator coil.

Mid point is on GND ...but i cannot pick up anything.

I use 1SS119 switch diode instead of 1N4148?

  Are you sure? yes | no

morph wrote 12/22/2018 at 15:14 point

There are two separate coils next to each other. They both have individual ground connection. The larger is the lo tank coil. The lo signal is inductively coupled to the rest of the circuit via the two coils. 

The smaller rf antenna coil has one variable capacitor to match the incoming rf frequency, you can also try different constant values and see which works best.

The lo needs to be adjusted for half the desired reception  frequency. I use an variable air capacitor, but any capacitance that can be altered is good for tuning. The series capacitors are specific to my setup and you need to experiment with them to adjust your tuning range. In the beginning you only need the main coil and the capacitor for the lc tank to work.

  Are you sure? yes | no

aurelw.wiz wrote 12/17/2018 at 21:09 point

First of all..thank you for reply!

Also , thank you for suggestions and experience with this receiver.

I must say that i don't build this receiver yet i just watch video

BUT i already have SW receiver with ordinary regen which not work very well

and id very noisy...your is NOT !

In the past i build similar VFO for FM super-regen and is really stable .

For RF preamp i will build circuit like is on given image.

I will post update...

all best!

  Are you sure? yes | no

aurelw.wiz wrote 12/16/2018 at 10:35 point

I watch this video and i must say that receiver work.

I have in plan build something similar with RF premplifier

in common base mode with BF506 pnp transistor fom old TV tuners.

So i hope that should work better.

As far as i don't know what kind of coil i must build?

You have 25 turns of green plastic insulated wire for tuning tank

and 5 turns for oscillator coil BUT what is a diametar of coil?



  Are you sure? yes | no

morph wrote 12/16/2018 at 18:35 point

Awesome! The coil diameter is 16mm, the form is a piece of PVC pipe. I did some air core inductor calculation and the receiver main coil is around 4.5uH give or take. 

I've since made some more experiments with the sub-harmonic mixer with good results. A BJT RF preamp is really a good improvement. Another fun design alteration is to use a 3.58 Mhz ceramic resonator instead of the main coil. You can then listen to the 40m band. I completely omitted the smaller value antenna coil and replaced it with an 1 k ohm potentiometer. In a colpitts configuration with the resonator or xtal you want to have a small coupling capacitor to the mixer, 20 pF has worked well. Same goes for the RF preamp coupling. I have had good results with coupling the local oscillator signal straight from the base of the BJT, the emitter also works. The base connection has a very clean sine wave when the drive of the colpitts oscillator is set correctly with a pot going to ground from the emitter (10k). In fact I also had very little drift when using just an inductor instead of the resonator. There might be potential to set some regeneration when using the base coupling method, but I haven't experimented with that, also the mixer diodes have a lo drive sweet spot that makes regeneration trickier.

I should publish the upgrades and variations here. Hopefully you got some ideas from this reply and I'm very happy that you are building this receiver. The topology is very fun and offers high performance for such a low part count :)

  Are you sure? yes | no

Similar Projects

Does this project spark your interest?

Become a member to follow this project and never miss any updates