OT: transformerless step-down DC power supply

How much current/ power? diode bridge and a big ass-zener?

Maybe a few of them in series. George H.

Usually the "drop" is accomplished using a capacitor.

See, for example, "CAP_Dropper-type_Supply.pdf" on the S.E.D/Schematics Page of my website. ...Jim Thompson

There are various ways to do it, but I can't think of one using diodes alone - not unless you use an awful lot of them! Well, I suppose you could put some very long legs on your BR and wrap them round the mains lead.

Tim Williams I think it was posted a counter to the usual diode-capacitor multiplier that effectively transformed the output down, but it requires a bunch of inductors.

NT

No, it was nothing like that, Jim. There was a distinct ladder matrix topology of diodes. I'll wager I still have that schematic on an external drive somewhere, but which drive, in which subdirectory and under what name....? I remember the designer's web site. He was American, quite a prolific inventor and listed most of his creations on the site. The thing I recall most from his home page was he was pictured in tee shirt & shorts, bouncing up and down on a large circular trampoline. Not one of our 'regular regulars' here, but maybe that image will jog someone's memory. It was a really clever design using jelly bean devices that would have totalled no more than half a dollar tops. It'll take me an absolute age to find it if no one here can assist. :-/

Many problems with that. Normally it would be limited to current considered safe for human "consumption". If not, since the neutral and Earth are bond ed at the service panel, it could be done if polarity is strictly observed but you can't count on that with your life in the US with all the incompete nce in some "handyman" type "helpers". I could tell you stories that would put hair on your chest, curl it and take it off in one fell swoop.

Actually I don't see the big problem, for lower power levels, of just getti ng a two buck transformer and leave no doubt that it is at least somewhat s afe.

If weight is a problem, brute force (but not much of it) push pull series a rranged transistors fed by a voltage amp to oscillate at about the resonant frequency and be dome with it. It can probably be done for about six bucks plus the cost of a case, which anything else would need anyway.

This diode contraption is not likely to have any type of good regulation or anything anyway. With all the limitations I would deem it simply not worth doing. And I don't mean like idiot number ten T who doesn't think an SMPS transformer is a transformer. (remember him ? He seems to have disappeared, good)

Or a Royer with a secondary ???

Oh, I was about to killfile you. Your SNR (among other criteria) is running exactly zero in the last three months.

Anyway, there's the series-parallel transformation of the Cockroft-Walton voltage multiplier: the current multiplier. But then you need N diodes and N (independent) _inductors_.

Which, as soon as you need one inductor at line frequency, you might as well put a smaller winding on it, and boom, you've got a transformer, and you only need four diodes, regardless of the ratio.

There's also a waveform chopper circuit, which uses the low voltage portions of the mains waveform to run the output, and is disconnected otherwise. It's like a rectifier, except with a window comparison rather than a single inequality. An exceptifier? But that's an IC, you can't do that with diodes.

Doing it with _just_ diodes, and it not being just a dumb diode dropper, sounds like woo to me. *shrug*

Tim

-- Seven Transistor Labs, LLC Electrical Engineering Consultation and Contract Design Website:

"I remember the designer's web site. He was American, quite a prolific inventor and listed most of his creations on the site. The thing I recall most from his home page was he was pictured in tee shirt & shorts, bouncing up and down on a large circular trampoline."

Yeah I'm pretty convinced it didn't work.

You could probably do something cool with emitter followers connected to zeners as regulators in series to push the dissipation into a couple transistors instead of a huge stack of diodes, maybe use both PNP and NPN devices somehow to take advantage of both phases in some way but without any bulk energy storage it's always going to be pulsed DC

It sounds reasonable that it should, but I can't visualize how it works. Can you point to a sketch somewhere?

Cool! If you take the classic from here,

George H. (not that I can ever imagine wanting to use such a beast.)

I have one (spoilers!), but it might be entertaining to do the transformation yourself. It's pretty tricky, if you aren't used to it. Hint: consider one component, pair of nets (converts to pair of meshes) and branch (converts to net) at a time. Make careful to note that voltage and current are swapped between the two forms.

Otherwise, if you can't get the puzzle, I have the solution below. :)

Tim

Nope- the transformation is much more involved than that!

The diodes aren't even in the same directions; but why, will make sense when you see. :)

Tim

Sounds like bitterex quality >:-} ...Jim Thompson

Google for: voltage multiplier and set the page to pictures.

Thanks I'll sleep on it and see if all your hints lead to something in the morning... then I'll cheat. :^)

George H.

Cursitor, The trash element have either decided to hijack your post by slight renaming it; or they have such crap newsreader's that they have to manually insert the subject.

Either case, they're still trash, without a clue of the issues involved.

If you have questions that I might be able to answer, use the envelope icon on my website and we'll just take this TECHNICAL discussion private >:-}

(I have killfiled the original thread and _all_ derivatives.) ...Jim Thompson

the only way I can think of doing that is if one set of diodes are IR emitters and the other set photovoltaic cells.

Back to back diodes, not conducting, using the junction capacitance, maybe several parallel pairs, to limit the AC, then another diode or four to rectify, several in series forward conducting to set the output voltage, then a whole bunch of reversed diodes to use their junction capacitance for smoothing.

Cheers

OK well seeing it didn't help me much. I'll have to simulate it or stare more.

George H.