inverting dc-dc converter circuit

This afternoon we were talking with Phil Hobbs about an electro-optical thingie we're doing. The prime power is a +15 volt wall wart, and we need -30 internally to power some photodiodes and opamps and stuff, 30 mA maybe. Our design currently has a cute homebrew single-inductor flyback converter, which circuit I've posted here some time back. We are concerned about having such a potentially noisy gadget on the same small board with nanoamp signals.

I proposed a different circuit: imagine eight opto-SSRs and two capacitors. A low frequency clock, 400 Hz maybe, switches 4 of them on and 4 off, alternately. The arrangement connects the two caps in parallel to the +15 supply, charging them up. Then it disconnects them and then restacks them in series such as to make -30 to ground. The low frequency and fairly soft switching edges should make this pretty quiet.

Phil named this the Groucho Marx Generator.

John

Reply to
John Larkin
Loading thread data ...

Incredible.

Why a 555 with a trivial diode-capacitor voltage multiplier won't work?

VLV

Reply to
Vladimir Vassilevsky

It's incredible that you think it might. Stick to digital.

John

Reply to
John Larkin

You mean this doesn't work:

formatting link
? Give it a try.

--
Muzaffer Kal

DSPIA INC.
ASIC/FPGA Design Services

http://www.dspia.com
Reply to
Muzaffer Kal

Diode forward drops would be an issue if you really need -30 and are trying to use a doubler circuit. Otherwise, for the benefit of the rest of us, why wouldn't it work?

Sylvia.

Reply to
Sylvia Else

Well, it wouldn't even produce -15v from +15v, let alone -30v.

Sylvia.

Reply to
Sylvia Else

At least one more cap to keep droop below 10% ppk. 25uF

The switched caps would have to be 5 or 10x multiples of this to maintain anywhere near 30V after the transfer.

With half the energy transfer lost in switching, thats 120mA drawn from the 15V rail.

Not small anymore.

There are quieter conversion methods than flyback or switched caps.

Harpo was mute.

RL

Reply to
legg

What make SSRs "fairly soft switching edges"? May be better to make a sin generator and a transformer?

Reply to
Robert Elson

The PV-powered gate drivers are very weak, so the switching edges are slow, far slower than a 555's. And the fets go from resistive to constant-current at higher voltage drops, which further softens things up.

And SSRs are fast-break, slow-make, so there's no shoot-through.

John

Reply to
John Larkin

There's no point to trying it. It's obvious that it wouldn't meet my requirements. Stick to digital.

John

Reply to
John Larkin

You can get opto-MOSFET switches as SSRs

formatting link

If you are careful to not have the source legs be the ones that move, you can use a capacitively coupled gate drive on a low Vth MOSFET and save on the current for driving the LEDs.

If you use a sine wave to drive the MOSFETs, you can have their switching be very soft and hence low noise.

An N channel and P channel MOSFET pair can switch one end of a capacitor from +15 to ground.

An N channel and P channel MOSFET pair can switch one end of a capacitor from ground to -15

An N channel and P channel MOSFET pair can switch one end of a capacitor from -15 to ground.

An N channel and P channel MOSFET pair can switch one end of a capacitor from -30 to -15

It looks like 8 MOSFETs plus two working capacitors would do it.

Reply to
MooseFET

Larkin and Obama are out of the same mold... mediocre talent, yet excessive ego ;-) ...Jim Thompson

--
| James E.Thompson, CTO                            |    mens     |
| Analog Innovations, Inc.                         |     et      |
| Analog/Mixed-Signal ASIC's and Discrete Systems  |    manus    |
| Phoenix, Arizona  85048    Skype: Contacts Only  |             |
| Voice:(480)460-2350  Fax: Available upon request |  Brass Rat  |
| E-mail Icon at http://www.analog-innovations.com |    1962     |
             
I love to cook with wine.     Sometimes I even put it in the food.
Reply to
Jim Thompson

If you still have this circuit handy, please re-post.

don

Reply to
don

How 'bout a 48V power supply and define ground where you want it? Or do you need most of the 'poop' at +15 Volts?

George H.

Reply to
George Herold

The 555 would only swing about 13 volts p-p. Then you get to start in on diode drops.

The 555 edges are ballpark 100 ns, noisy.

555s have supply shoot-through, ditto.

John

Reply to
John Larkin

Here it is:

ftp://jjlarkin.lmi.net/Inverter.jpg

John

Reply to
John Larkin

Sure, it needs an output cap.

I'd just use huge aluminum electrolytics. 1000 uF caps would be no problem.

No. A charge pump like this can approach 100% efficiency.

The nice thing about the Groucho circuit is how slowly it runs, ballpark 400 Hz, with soft switching edges. And no magnetics.

John

Reply to
John Larkin

Two years ago I did the electronics for a super noise-sensitive fiberoptics project. Late in the game a spec change came along (don't they always?) and now we needed +40VDC and +210VDC. There was only 12VDC available to make it from. So I did two homebrew converters, MIC4421's driving FETs and I babied the shoot-through a bit. The 40V got one of those Cooper Versa Pac transformers and the 210V an ordinary CCFL transformer. The rectifiers were well muffled with the usual inductors.

Not one wee whiff of noise. Ok, I had it on a PLL but even when I unclutched that there was no noise. I love magnetics.

Cool :-)

--
Regards, Joerg

http://www.analogconsultants.com/

"gmail" domain blocked because of excessive spam.
Use another domain or send PM.
Reply to
Joerg

Whereas Jim's egomania is justified by his stellar reputation amongst integrated circuit designers? Bob Widlar, Barry Gilbert and Hans Camenzind recognised him as comparable talent? Had every actually heard of him?

-- Bill Sloman, Nijmegen

Reply to
Bill Sloman

How about a couple of LTC1144s?

Reply to
MooseFET

ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.