Re: board - T562.jpg

You seem to have some analog credentials. Maybe you could divert some of this thread to answer a question about switched cap voltage converters.

Where I am currently working we need to convert a wide battery voltage range to typical circuit voltages, such as 5, 3.3, etc. volts. The battery voltage varies from 7 to 17 typically. We can used inductive switchers for this, but it is in a radio and we prefer not to use these. They are also not so efficient when running at low currents, at least the ones we can buy as chips. By low I mean

Just kill file any cartoon character you see posting on a newsgroup.

These sweats don't have pockets, and anyhow I'm surrounded by cats. And why would I keep a USB Logitech in my pocket?

Ah, wit. Thanks for pointing it out to us.

John

I didn't know there was any such thing as an *efficient* switched-cap regulator. Their only convenience lies in small size. They are particularly nice when there is no load current... such as the bias for an integrated electret microphone in a hearing aid chip I designed last year ;-)

However, if you're switching large currents... thus low resistance switches, driving the gate capacitance *will* be a major contributor to your losses.

Are you certain you wouldn't be better off with an inductive switcher?

What are your actual I/O requirements.

...Jim Thompson

Why can't a switched cap converter be efficient? The only losses are in the resistance of the swtich (same as an inductive switcher) and the difference in voltage between the switched output and the end circuit. I was thinking along the lines of using multiple topologies to get a good voltage match to the required output voltage at any given input voltage. 2:1, 3:1, 4:1, I even came up with a circuit to provide

Two problems with inductive switchers running from higher voltages; 1) the losses are significant when running at lower currents like < 100 mA. My current board has to run in idle at 100 mW and provide up to 2 watts in full performance mode. With a 9:1 ratio of operation, losses at 100 mW can be significant.

Input ranges from 7 to 17 volts. Outputs are at 5, 3.3 and 1.8 volts. I would only expect the main switcher to generate 5 volts as the others can be made from that. It should run from a 600 kHz reference clock, but that can be divided down if needed to make the circuit work efficiently. The output current for efficient operation would range from 20 mA to 250 mA.

Is this anything like practical?

I've completely lost the point of this now. Have you?

HOLLY CRAP!!! THERE WAS A POINT???

LOL!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!

Late at night, by candle light, "Homer J Simpson" penned this immortal opus:

Actually I was referring to YOU and Fields. Take Jim, Mike and Graham along, make it a slurp-all-you-can fest.

- YD.

Late at night, by candle light, "Homer J Simpson" penned this immortal opus:

Obviously. No.

- YD.

Then what has it to do with finding prime numbers on a microcomputer?

Late at night, by candle light, "Homer J Simpson" penned this immortal opus:

NFI. You should know.

- YD.