I think that the noise in this regen is caused by shifting internal capacitances of the two BJTs used in this circuit. These capacitances shift at the amplitude of the oscillatorsignal. To minimize this effect, you must use RF transistors in this regenerative circuit.

When the receiver is oscillating, the base-collector junctions of the two BJTs which are DC shorted by the tankcircuit coil will start conducting, imposing damping of the tank circuit. That destroys the Q of the tank circuit, causing a poor selectivity and adds extra noise too.

very interesting circuit Alan! is that a differential amp configuration? i'm very interested in knowing how such configuration works, would appreciate it if you could explain a bit.

i couldn't resist but built it on my breadboard with the intention of covering the FM broadcast band, albeit i don't know anything about electronics. the only high frequency BJT's i have are 2 varieties of NPN's, so i just flipped the circuit upside down.

a pair of 2SC9018 (Ft=1.1GHz) was tried first, they seemed to oscilate, but did not squeal even when the 25K pot used, the only pot i have, was turned to zero. so i changed the emitter resistor from 100K to 51K, it squealed. after some fiddling with the tuning coil and cap, some stations came in, loud enough for a table top set, with a simple 2 stage AF amp followed by an LM386 played through a small 3" speaker. great!

then i tried a pair of PH2369's (Ft=500MHz), tuning positions of the stations changed quite a bit, and not as loud, it dind't squeal either. so i reduced the emitter resistor again from 51K to 22K. it squealed, and stations received were as loud as before. i measured the voltage across the emitter resistor, 1.77V across 22K, so the emitter current is about 80uA, correct? is that ok, does that go with your original design, or will that mess up some design aspects that you have given considerations to and i'm not even aware of?

one last question, how come the AF information is already in the emitter current, without any additional parts doing the demodulation, and even an RFC is not required?

thank you for this very interesting circuit, great fun and learning tool for me!

This type of oscillator is often used in single chip FM/AM superhets which are often found in clock radios and portable radios. In these receivers the oscillator tank circuit is often coupled to the oscillator by a link coil to minimize the nasty effects onto the tank circuit.

You can also use a separate detector with this regenerative circuit. On this receiver at http://www.b-kainka.de/bastel3.htm you have the same regenerative stage which you have used in the noisy regen. But there is a seperate audion detector used. The regenerative stage and the detector a coupled onto a tap of the coil.

Cool circuit, some time ago i've built it for shortwave on a breadboard in 15 mins using a polyvaricon cap and a molded inductor, and it worked immediately. Seems to me quite identical to the russian "universal demodulator" (Polyakov, I think), see http://ok1ike.c-a-v.com/soubory/det.htm (in Czech, use google to translare). The audio output is taken from a one stage fet amplifier (or detector?), and it's recommended to use low noise fet. Maybe an improvement for your "noisy?" I've dismantled the noisy after listening a little, but I'm going to try the fet mod. See later, 73 de Michele IZ8JJI

Jason,

I'm glad to hear you built the Noisy Regen and got it working OK. Send me a picture if you happen to take one, or shoot me a URL and I'll link it to this page.

Regards,

Alan

That is a cool little projct i build it from your diagram and it worked thanks