You know Guru. I'm gonna abandon the regen. receiver. I hooked up a battery pack and could not get the same audio. Those things are just too unstable. They work but I would not expect your average Joe to tolerate such a device. Not anymore. I went back to an inverter transistor configuration and it leveled out the signal but the battery pack deal was very discouraging. I think I will take the DC receiver and break the coupling point to the audio and go ahead and convert it again. I intend to use the least amount of parts as possible. This is not a communications receiver. Just something to enjoy short-wave broadcasts. I want this low power. So I think a good low frequency 2nd local oscillator will do the trick. I should get a high output off of it, then mix it and try an LC filter or maybe scrounge of a 455Hz cermic filter out of an old radio and the run it back to the audio. May 2 or 3 more transistors.
As a side note it was interesting that the first two stages of the regenerative receiver double as an oscillator and I found that the two transistor relaxation type oscillator was rather stable with a high output! Also that same technique (the feedback) could be used as a receiver front end for high sensitivity. Of course you want a fixed bandpass which is not hard at all. A parallel resonant front end filter works quite well and is only two parts with a series coupled capacitor making 3 total.
edit*
At any rate, that will remove the beat note though I might find some birdies along the way.
Oh, response to the synchronis receiver someone mentioned. I looked at it and I do understand the concept. I think after what I just observed with those type circuits. I for see instability problems with a broad tuning range. It would probably require less circuitry just to do another conversion. Yes, overall I thing dual conversion superhets reign supreme.