Capacitive transformers ? another ponderance for ye experts

This is done often in RF (read: narrow bandwidth), where you put two (or more) capacitors of various value in series and the voltage divides like a resistive divider, except because the reactive current is canceled with an inductor (the total series capacitance is the capacitor side of a tuned circuit), you can divide it with little power loss.

I thought about this problem once (that is, an electric rather than magnetic transformer), but I don't remember my thought process. I don't think it'll work the same because electric fields have monopoles, whereas magnetic fields are always loops. It also wouldn't work as you'd expect, since voltage and current are reversed: instead of a low impedance (constant voltage) source, you need a low conductance (constant current) source to make an equivalent circuit.

Tim

-- Deep Fryer: a very philosophical monk. Website:

True, not to mention there are already piezo transformers in existance, even high powered ones such as the "Transoner" which I'm hoping will actually become available soon.

But I'm thinking of some sort of stacked plate device with terfenol or possibly some kind of MEMS membranes that vary the capacitance of a single flying capacitor. Of course, the switching circuitry would consume some power, hopefully minimal.

Switch the cap to the source in the expanded low capacitance mode, then switch it over to the load and contract the plates. But then again, I reckon this is DC/DC conversion and not necessarily the equivalent of an AC transformer and would require rectification of an AC input.

I'm simply proposing a device. It's up to ye experts to determine what if anything could be done with it, apart from a door stop or paper weight :-)

-DM-

: : I thought about this problem once (that is, an electric rather than magnetic : transformer), but I don't remember my thought process. I don't think it'll : work the same because electric fields have monopoles, whereas magnetic : fields are always loops. It also wouldn't work as you'd expect, since : voltage and current are reversed: instead of a low impedance (constant : voltage) source, you need a low conductance (constant current) source to : make an equivalent circuit.

I think I see what you mean. IOW for near 100% efficient power conversion.

But wouldn't it work for wideband lossless voltage conversion as apposed to a narrow band response from a resonant LC?

Certainly there's gotta be something useful there in some way shape or form :-)

: : Tim : : -- : Deep Fryer: a very philosophical monk. : Website:

What sort of capacitance would you need to do useful work? 10s of uF? How much plate area is that? You now have to provide that much piezo area, or a fair fraction of it, and vibrate it all at audio frequency. Easy to roll capacitor plates into a can, but for this you'd need them flat, which is none too convenient. And you'd need power input to the piezo element.

that would drop the V. V boosting is worth more money, though has many less apps.

a diode is no problem. I guess youre thinking of replacing wall warts.

Let me propose another equally problematic design option, one that bypasses the need for costly piezo or drive power.

The capacitor uses a domed film as the dielectric. This behaves a bit like the rubber mebrane on a keyboard: with no applied pressure it stays thick, when force is progressively applied it suddenly collapses. It springs back when applied force is removed.

Now, apply a charge to the capacitor through a diode. At some point, as V increases and the interplate forces rise, the domed film collapses, the plates get much closer, and C rises considerably. You now have a lower V with the same charge as before. Yes, youve reduced v, but you havent added i, and this is the problem in both applications. No transforming effect. Just a mechanical loss.

Theres also no isolation like this, so no use for wallwart apps.

FWIW discharging the cap remove interplate forces, allowing the film to recover.

So the problems are high cost, large size, operating energy consumption, no transformation, no isolation, noise and vibration, and the realiability issues that tend to go with mechanical devices. R&D's a b--ch!

NT

There was a mechanically vibrating capacitor system used to convert the output of a pH electrode from DC to AC in the Pye Dynacap pH meter. It had the advantage that it drew no continuous current from the electrode.

The AC signal was then amplified by a more-or-less conventional (valve) amplifier and detected synchronously. A proportion of it was fed back to the other capacitor plate to give overall DC feedback.

It was a beautifully designed and constructed piece of analogue kit with a staggeringly high level of accuracy.

Even with Terfenol you will likely get less than 50% efficiency at resonance.

Make an electric version of a magnetic generator. Rotate a variable cap that has interleaved plates. Charge when max capacitance, discharge when min.

I've done a little mechanical design, and I wouldn't want to tackle this project. I don't know what "terfenol" is, but it sounds like some kind of drug. ;-) I've heard of MEMS, and I'm not really qualified to opine on them, but I'm thinking some kind of hydraulic piston or something, or maybe something like one of those old 365 pF variable caps that they used in the old AM radios - the mechanics would be prohibitive.

And from the other end, your idea here is just a charge pump, which they use routinely here and there. :-)

Cheers! Rich

The "Dynacap" unit had two plates on the tines of a tuning fork which was kept in vibration electromagnetically. The frequency was higher than you would get with a rotating cap and there was nothing to wear out.