For what it is worth, I thought I'd tell everyone that I've built my first properly functioning FM radio (of very simple design). I built one a couple of days prior to this one, but it changed stations every time I'd move an eye lid. It was all over the place, and mostly just dead.

This radio uses regeneration. It is based upon the 12AT7 FM receiver that you can search for on the internet. Instead I thought I'd try a 6GH8A, with the pentode as the cathode supplied antenna coupler. Thought the pentode would give it more gain, but there really is no gain when you feed the signal in through the cathode. This is supposed to give the antenna circuit more stability, though, which it does (otherwise antenna placement affects frequency). I'm thinking about adding another stage of RF amplification, though the unit does quite well as it is. Some stations are a bit hard to receive. For the output, I use two 6EB8/6HF8 tubes (one of each, because that's what I had at hand). The triodes are the drivers, and the pentodes are push-pull output. Not bad, considering that these tubes are designed for deflection circuits in a TV. They work well. For power I have two Radio Shack transformers. One supplies power to the heaters and to the secondary of the other transformer. The other transformer's primary supplies the high voltage (120 volts), which is fed to a voltage doubler (raises to 300 without a load...at load, it's about 180). Amazing! The circuit almost has potential. Stereo subcarriers can be annoying on some stations, especially weaker ones. At the very least, it is quite the novelty.
Thomas

How about some pictures Thomas??? I tried the simple 12AT7 fm regen and was very disapointed with it, as you said it drifted terribly, even standing in the romm in differnt place casued significant change. But i did by a Westinghouse tube radio that has FM and i like it very much.

:For what it is worth, I thought I'd tell everyone that I've built my first properly functioning FM radio (of very simple design). I built one a couple of days prior to this one, but it changed stations every time I'd move an eye lid. It was all over the place, and mostly just dead.
:
:This radio uses regeneration. It is based upon the 12AT7 FM receiver that you can search for on the internet. Instead I thought I'd try a 6GH8A, with the pentode as the cathode supplied antenna coupler. Thought the pentode would give it more gain, but there really is no gain when you feed the signal in through the cathode. This is supposed to give the antenna circuit more stability, though, which it does (otherwise antenna placement affects frequency). I'm thinking about adding another stage of RF amplification, though the unit does quite well as it is. Some stations are a bit hard to receive. For the output, I use two 6EB8/6HF8 tubes (one of each, because that's what I had at hand). The triodes are the drivers, and the pentodes are push-pull output. Not bad, considering that these tubes are designed for deflection circuits in a TV. They work well. For power I have two Radio Shack transformers. One supplies power to the heaters and to the secondary of the other transformer. The other transformer's primary supplies the high voltage (120 volts), which is fed to a voltage doubler (raises to 300 without a load...at load, it's about 180). Amazing! The circuit almost has potential. Stereo subcarriers can be annoying on some stations, especially weaker ones. At the very least, it is quite the novelty.
:Thomas

Well, this one doesn't drift at all. It's actually quite good. It could use an extra RF amplifier. I made a major improvement last night, though. Since the primary antenna triode (pentode in my case) doesn't really do much (it couples the antenna via the cathode, and does this by injecting the signal into the cathode, which swings it a little bit positive and negative with respect to the grid, which is the same thing as making the grid more positive or negative, but the cathode isn't at all as sensitive, so there is little gain), I decided to see what would happen if I took a little bit of the signal at the regen. cathode and fed it back to the grid of the antenna triode. I'm not sure whether the cathode of the regen. triode is in or out of phase with the grid of the antenna triode. At any rate, I didn't care...just wanted to see what would happen. I removed the .001 MFD condenser from the grid. Instead I connected a 10K resistor from that point to the same point as the antenna triode's cathode choke connects to (slightly above the chassis). I also tried B- (which is below the chassis), but liked the previous bias point better...you can try either. I then connected a 10 pF condenser from the grid of the antenna triode to the cathode of the regenerative triode. Signal strength did increase just slightly (larger condensers cause severe station shifting...which can be corrected by series condensers on the tuning condenser, but the larger condensers also cause unwanted feedback, causing the receiver to be unstable). What was also a miracle is that most of the stereo subcarrier noise and whistles that were present are just about gone on all stations! Some are crystal clear, and some have very quiet swishing in the background. Weaker stations of course have the rushing sound...not sure if this is the regeneration or noise in the air. I assume the former. I also added a shield to the tube, which reduced some rushing sounds. I would really like to see how this thing performs with an additional RF amplifier. I may add an AVC circuit to prevent overloading, or I may add a limiter instead. I'm may use another triode/pentode set-up, which will allow flexibility. For starters, I will use a 7 pin pentode, since I only have 7 pin sockets at hand. I had 3 9 pin sockets, but used them all already.

Anyway, this receiver is very promising. It receives incredibly well with very little circuitry. It is interesting how each station can be picked up twice due to slope detection. Actually, you're only picking up the station once, but there comes a point where you pick up both sides of the frequency shifting (frequency modulation), and the two cancel eachother, making it seem as though the station went away. The station comes back when you tune to either side of this point. It would be very interesting to build a regenerative ratio detector. I wonder if anyone has tried this. You know...the circuit at the end of a superheterodyne that detects with two diodes.... Well, let's say that you build a tuner with the coils set up like the secondary of a ratio detector. Then, instead of feeding them to diodes, feed them to regenerative triodes. In this way you could possibly detect like a superheterodyne, but not with all of the circuitry, due to regeneration. Hmmm.... I'll have to try this one. If it worked properly, you could pick up from the center of an FM station, like you're supposed to. I'm impressed with the current circuit which I put together, though. It's quite nice.

I will try to come up with pictures.

T.

Here's the original circuit.

http://www.users.bigpond.com/cool386/12at7/12at7.html

I used the battery version (first schematic). I did not use the audio reflexing circuit. In place of the headphone transformer, I used a 10K resistor. The antenna triode, in my case, is a pentode. I will try to draw up a schematic soon.

T.