whither regenerative amplifiers?

Rarely, if ever. Gain and selectivity are cheap nowadays, and regeneration trades them for stability.

John

IIR filter.

John

Yes, there are some kits available to build, seen them somewhere on the net. There is also quenched regen. which works better than just regeneration. Regen has the best sensitivity, but poor selectivity It was popular in the first radios 1910 to 1925, then the superhets came in, with better stability, but cost more.

Hi Alan,

Last time I used it was, oh, about 30 years ago. It worked great and spared me the expense of a crystal filter. Nowadays it is possible to use some cheap mixer chip, filter with a crystal at a frequency where it is low enough in cost and then mix back to where the signal needs to be.

Also, I believe the art of such fine analog schemes is to a large extent lost. Younger engineers often wouldn't even know what you mean by that question. Thinking back to my days back at the university, have we ever been taught regen or Q-multiplication? Nope.

Regards, Joerg

Hi John,

So after trying to mimick tube amps via algorithms the next big thing would be the emulation of a Q multiplier on a DSP :-)

Regards, Joerg

Most RC+opamp filters include some sort of positive feedback.

--
--
kensmith@rahul.net   forging knowledge

Yes, course there is. Needed a 300Hz bp filter for an I.F stage. too wide for a crystal, too narrow for a tuned circuit. Ended up using a Twin-T with feedback to give Q enhancement and flat response. regards john

I guess that depends on the application. I don't know how true it is, but I believe I saw that some of the 433 MHz ISM band receivers use either regenerative or superregenerative detectors. They are tiny and have sensitivities better than -106 dBm.

There has been new work on superregens (September/October 2000). An article was published in QEX, a ham radio journal. The first file at

is that article but it requires membership to get it. However, on the same Web page, I see another that appears to be available without membership. It is An Ultra-Simple VHF Receiver for 6 Meters .

You can probably contact the ARRL and buy a reprint of the first article. It is well worth it. The author, Charles Kitchin, says the gain in a superregen detector is about a million. He discusses his discovery of shaping the quenching waveform so that selectivity is not lost. And he says it is possible to receive FM and NBFM as well as AM. It is a very enlightening article.

I plan to build a superregen to play with. After I get one going, I think I will try to make it work at 450 MHz. Like some other posters here, I built my previous regen 40 years ago.

Cheers, John

Yes, if that is the optimal way to meet the requirement. However, I'd have thought a DSP based solution would be more likely.

--
Brian Reay
www.g8osn.org.uk
www.amateurradiotraining.org.uk
FP#898

Not exactly what you asked for, but you might be interested to know that this is exactly what is going on in your inner ear... and it's partly mechanical! The sensory cells in your cochlea sit on the basilar membrane, which vibrates from the incoming sound. The sound enters at one end of the cochlea, where the basilar membrane is narrow, stiff, and has low mass. The membrane becomes wider, more massive, and less stiff as you go toward the apex of the cochlea. The incoming sound generates a travelling wave on the membrane, with a peak at a frequency-specific location.

But it wouldn't be very sharply tuned just from the mechanics. As it turns out, there are 2 different kinds of sensory cells on the basilar membrane: one row of Inner Hair Cells which actually tranduce the sound into neural signals to the brain, and 3 rows of Outer Hair Cells which are electromechanical amplifiers. They change length according to the voltage across them, and they also change the flow of current and hence the voltage across themselves when this happens. The net effect is that they form a regenerative amplifier that provides a huge Q-multiplication for the mechanical tuning of the basilar membrane, and it is this enhanced vibration that excites the Inner Hair Cells to give enhanced sensitivity to low sound levels. Without this effect we'd lose about 60 dB of sensitivity at low sound levels. (The effect saturates at higher levels.)

Just thought you'd like to know!

Bob Masta dqatechATdaqartaDOTcom D A Q A R T A Data AcQuisition And Real-Time Analysis

John Larkin

Selectivity at the upper decades of Mhz is cheap nowadays? Where do I buy one of these flea-market tuneable multi-ganged, several-octave HF-VHF preselectors?

I wasn't thinking of regenerative detectors. I was thinking of a tuneable front-end preselector which would emulate the Q characteristics of a crystal-lattice bandpass filter.

don't we have nowadays, high-power, low-noise FETs which are almost as good as the vacuum tubes made in the late 60's, early 70's.

Stability? RF stability is like aircraft stability.... if your design and implementation are excellent, it would be hard to get into an unstable area of operation.

Or SAWs, or ceramic resonators, or DSP stuff.

John

If only you could explain cochlear hydrops in a way that suggested a fix.

-- So those 380 tons of missing explosives were moved by Saddam before all those expert inspectors noticed, eh?

No wonder twelve years of inspections found NOTHING.

Umm. Er Don't you mean negative?

Steve, K9DCI.

Hi John, I built a few superregens about 40 years ago too!! They were on 288 MHz when Australian amateurs had the one metre band. Surprisingly good sensitivity for a one-valve receiver (6J6 usually) but of course they put out a fair bit of rubbish over a fair bit of the band. I often wonder if the quench frequency could be better controlled - perhaps an external injection oscillator or something like that. Anyway, best of luck with your 450 MHz. Alan VK2TWB

Ham fest, build your own, can get more than 10:1 over frequency.

Can get to a xal selectivity at all. You can do an elecrically tuneable with Pin diodes and Fets. but it is at least 10 to 100 times wider.

12 or 14 bit, 80 MSPS ADCs are cheap and plentiful nowadays, and consume a few hundred milliwatts.

John

Well, sure, but it's probably not going to be a regenerative amp stage.

John

Hi Alan,

Only using conversion - filter - conversion back. Otherwise it's cavity resonators and beaucoup Dollars.

Don't expect lattice characteristics. I used a Q multiplier at IF frequencies quite a bit when I was young. You can "dial in" almost any bandwidth you like but the 6-60dB shape factor was never even close to what a four or six crystal filter could do. You get the characterics of a single tuned LC circuit which has a selectable Q. No more, no less. Of course, you could gang Q multipliers but that is going to be very complicated when it comes to aligning them.

From a noise point of view they are better. But if you ever had to go up a slippery roof after a thunderstorm to replace the amp board because the FET had blown you start to appreciate tubes.

The old Q multipliers were kind of fickle. But then again, so is a Ferrari. Things could be automated but this technology enjoyed too short of a life span for that to happen.

Regards, Joerg