Magamp oscillator (2023 Update)

Does this have a use? I would think you could make a bang-bang oscillator but you'd need a switch or two.

George H.

mag amps require an AC supply to exhibit amplification, so if you allow an AC supply then yes it's possible.

I think the ring generators used to generate ring voltage for phone exchanges used to work that way, being some sort of phase-locked oscillator. dividing 60Hz to 20Hz or 50Hz to 16.66' Hz.

On a sunny day (Tue, 3 Nov 2015 16:30:15 +0100) it happened Piotr Wyderski wrote in :

I already told ye, 'solid state diode' -> tunnel diode. If you do not know 'tunnel diode' then look up Esaki diode perhaps. An LC oscillator can be made with 1 inductor resonating on its own winding capacitance if you are strict and do not allow capacitors, and some resistor (you will also need those in magnetic amplifiers). So then you are already done. If you MUST use transductors, then use the tunnel diode generated AC to power some, and feedback, for a gain > 1 and the correct phase that will oscillate too.

Unless you want to change the rules when playing, and forbid these diodes... Even than be it powered by AC or DC, ANY amplifier needs a power source, magnetic amplifiers too.

There is no problem.

I don't think so, the question is out of pure curiosity. As one of the other posters has already noticed, it would not be better than a transistor-based oscillator, and not even smaller. Since the smallest reasonable oscillator is probably an ATTiny in SOT23-6...

I am currently playing with magamps and hence the question.

Best regards, Piotr

I know of tunnel diodes and I like the idea. The available ones are low-power, but that can be amplified.

Best regards, Piotr

I recall seeing a late 1950s "all diode audio amplifier" circuit, it was class-D pwm and used only 2 terminal active devices. As well as regular rectifier diodes it used 4 layer diodes as relaxation oscilators!

piglet

There is such a thing as the "parametric transformer" about which I know nothing but assuming it works like a varactor diode parametric amplifier then I suppose it could be made to oscillate, I guess it needs a higher frequency pump to provide the power source?

piglet

Magnetic amplifiers are in the general class of 'parametric amplifiers'; I think some of the ferroresonant tricks were given names such as "parametric transformer", but one would have to search a lot of old sales brochures to be sure.

Just in case that was to me, I think it was just and experiment to see how low he could go on B+ and still make an oscillator. He started with making a very high input impedance transformer for crystal radio use, and that led to this. (I think) He has a site, at the bottom is a list of projects, you will find the transformer project in that list.

Mikek

The telephone ringing generator is such a beast.

The telephone ringing generator used inductors and capacitors; both of those are solid state (no gas, no vacuum).

As i said, Ma Bell did it..

BUT you have...the ring generator in the telephone,designed and made by the millions by Ma Bell.

And the bee cannot fly.

Some things are actually impossible.

Bees can obviously fly. What idiot decided that they can't?

If you can design an oscillator that meets Piotr's requirements, we'd love to see it.

How does that work?

I think you could make an oscillator using a mag amp for gain. But if you use say 60 hz power, the highest frequency that the oscillator could run at would be about 6 hz.

Dan

Feed the 60Hz to an LC & get 600 out of it. Oh, via diodes. BTDT by accident.

NT

I recall seeing (about 10yrs ago) a ref to math showing that it's not possi ble; that even the ideal case, mag amp operating point is outside the gain/ phase for oscillation. I didn't look up the ref though.

It was in an article on historical development of electronics, claiming tha t the first oscillators had to wait for vacuum tubes, since even though mag amps existed earlier, oscillators weren't possible. (But even Faraday cou lda made one, if he had Nyle Steiner's burned zinc oscillator.)

Also: "impractical" means that hobbyists do it all the time for laughs, wh ile "impossible" means that no hobbyist has succeeded, ever.

Separate topic: is an all-inductors diode possible, using magnet-biased sa turable reactors? I recall a Tesla patent claiming this. If real, why all the rotary converters, and why were selenium or copper-oxide rectifiers su ch a big deal?