Generate 200V at 10mA from 24V

There's probably an off-the-shelf device that does that. For my home telephone system I found a 5V -> 48V converter for ~$10. ...Jim Thompson

Hi Tim, have you trolled DK for DC-DC converters?

Here's one from Recom.

George H.

Hi, Tim:-

Have you looked at controller chips- parts with an external switch and current limit on the low side?

Eg. LM3478 (just picking one at random).

--sp

I've used an Offline switcher in a pinch..

If you look at something like the NCP011 (onSemi) series or equal, it gives you minimum connections. You still need to deal with magnetics however, this is a very simple xformer because feed back is done via optical with a zener in line of your required voltage. This means the circuit will operate on a wide range of input voltage with minimum conponents.

I think there also could be a simulation file for this family, I would have to check that.

Jamie

I'll look some more. Mostly I was wondering what the crew thought about topologies -- i.e., are there any big pitfalls to using boost as opposed to flyback, etc.

"Don't care too much about efficiency" is an interesting phrase -- I'd think that my threshold is somewhere around 80% efficient, but super easy, 70% efficient, and still not burning up the board would probably work for me (but -- where does that watt go?).

The example board that I'm looking at is interesting: they've got a 20W input converter the size of a matchbook on a board that can't use more than 4W total, $90 DC-DC converters, etc. On the plus side, I think there are some efficiencies and size reductions I can find. On the minus side, any time I see a board that's so far from my expectations I have to ask "was this really designed by an idiot, or did they know about a pitfall that I'm going to fall into?"

Some little switcher chip and a flyback transformer sould work.

I posted this a while back

to make 1200 (or now, 1400) volts from 24. The HV1 output is 280 volts, so you could strip the rest of the multiplier chain.

That DRQ127 dual-winding inductor is cheap and gives you a 2:1 boost for free, and lets you use a lower voltage fet. You could also get 200 from 24 with just an inductor, classic boost.

The fet could be something tiny at this power level. You could use a chip with the output transistor built in, but then you'd probably need a higher-ratio transformer, maybe custom.

Or for really cheap, roll your own:

Thin on ICs? Roll your own ;-)

Or chuck in whatever controller chip you like. You'll probably want a

And as mentioned, boost is an option too. I'd avoid direct boost, but a tapped boost (boosted boost..?) would be fine, like JL's.

Tim

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

A customer board is currently generating 200V at 10mA from 5V, using a $90 part. I'm changing enough stuff that it's very likely I'll be able to start from a 24V rail for that same result, and their production volume is high enough that a day or two spent designing in a solution with less BOM impact could be worthwhile.

IC's that do this seem to be thin on the ground, and magnetics seem to be all custom. The solutions that I did find revolved around photoflash stuff, which are on the wimpy side, and one Linear part (LT3751) that is designed for much bigger circuits.

Is there a known solution for this out there that uses off the shelf parts and answers my need? Or is my initial assessment accurate?

--

Tim Wescott Wescott Design Services

Just change Q1, Q3, Q3 & Q5 to HV devices >:-} ...Jim Thompson

-- | James E.Thompson | mens | | Analog Innovations | et | | Analog/Mixed-Signal ASIC's and Discrete Systems | manus | | San Tan Valley, AZ 85142 Skype: skypeanalog | | | Voice:(480)460-2350 Fax: Available upon request | Brass Rat | | E-mail Icon at

Q1, _Q2_, Q3 & Q5 :-[ ...Jim Thompson

-- | James E.Thompson | mens | | Analog Innovations | et | | Analog/Mixed-Signal ASIC's and Discrete Systems | manus | | San Tan Valley, AZ 85142 Skype: skypeanalog | | | Voice:(480)460-2350 Fax: Available upon request | Brass Rat | | E-mail Icon at

My google-fu failed me entirely with Digikey -- I missed the category of SMPS transformers. I may be set now.

Pffbt! Mine uses less! ...Semantically speaking!

;-)

Tim

The Baxandall Class-D oscillator uses even fewer

With a 24V rail you can't cross-connect a pair of power-MOSFETs gate to drain - 75V from gate-to-source is a bit much for regular parts.

Biasing a gate-drive winding so that the circuit both starts and doesn't have both MOSFET on at the same time takes some care. I don't much like this solution, but LTSpice says that it works

With bipolar transistors it's a lot simpler, but they can squeg if you make the inductor too big.

Obviously missing the point: Baxandall's circuit is not regulated, nor is it a controlled PWM circuit. To do that requires quite a bit more hardware,

Tim

What's wrong with a simple single-inductor boost circuit? If you use a controller IC with an external MOSFET, the IC can be an inexpensive low-voltage type. We show sixteen candidates in Table 9.6. For example, the LM3478 comes in an 8-pin package and doesn't require many external parts.

The Baxandall circuit can produce a a predictable and stable multiple of th e voltage between the input to the inductor and voltage across the "on" dra in or collector. If you need a regulated output,sense the voltage you want and adjust the voltage at the inductor - PWM it if you don't want to waste any power there.

It may be less efficient than a well-designed PWM circuit, but it is usuall y a whole lot quieter and simpler.

Win, Did you solve your refrigerator light problem? ...Jim Thompson

High boost ratios can get messy. At some point, the unloaded kick reaches some finite voltage, the peak of the half-sine ring. At that point, all the stored energy, less a bit of losses, has been transferred from the inductor into various capacitances, and there's nothing left for the load. All that energy in the capacitances is generally unrecoverable. An inductive boost, to be efficient, has to output a fraction of that peak flyback voltage. Making 200 from 24 is probably OK with an inductive booster. The first 24 is free, so the boost ratio is really about 7.

A high boost ratio also needs a high-voltage mosfet, which suggests high Rds-on. So at some point a real flyback, with a transformer, or adding a C-W multiplier starts to make sense. My center-tapped inductor autotransformer trick is a compromise.

be

is

Baxandall's class-D oscillator is rather less of a compromise, but you have to be willing to wind your own transformers (or to get them wound) if you want to exploit it. And it can pay to have the transformer wound in a way t hat minimises the effects of inter-winding capacitance, with - say - banke d windings (for example).

There's a longer thread about 1.5V to 400V from a year ago:

Cheers

Klaus