6V to 5v ~1mA max. typical 200uA. efficiency >90%

Hi, how can i build a high efficiency(>90%) voltage converter for converting 5.2~6v (4*AA) to 5v and 3.3v (dual output)? 1mA max, 200µA typical(each output) /non-isolated . any suggestion?


thanks.

Re: 6V to 5v ~1mA max. typical 200uA. efficiency >90%

For high efficiency low power converters, charge pumps are pretty much the only option. But getting arbitrary conversion ratios can become very complicated. For example, converting between 5V and 3.3V isn't too difficult (there are chips that are designed to do this), but converting something like 6V to 5V is much more complicated, and would require many more switches and flying capacitors. And when the voltages change, either the ratios need to change (requiring more capacitors and switches, and a smart controller), or you have to accept decreased efficiency or output voltage error.

What are your ripple voltage requirements on the outputs?

A few special designed micropower switchers like Linear LTC3652 are going in the asked direction, but most likely not exactly fitting your needs. You can take it as a proof that's not completely impossible to design a buck converter for similar specifications.

As mtwieg explained, charge pumps can't provide a continuous voltage variation without losses, fractional charge pumps can at least provide a theoretically losless conversion for rational voltage factors. Unfortunately, the integer numbers in the voltage ratio are corresponding to charge pump capacitor and switch counts.

Re: 6V to 5v ~1mA max. typical 200uA. efficiency >90%

FvM said:

A few special designed micropower switchers like Linear LTC3652 are going in the asked direction, but most likely not exactly fitting your needs. You can take it as a proof that's not completely impossible to design a buck converter for similar specifications.

Click to expand...

I think you mean the LTC3632? That gets pretty close to the OP's requirements, but its efficiency seems to drop off below 1mA. Maybe if it's used to convert to 5V, then a charge pump is used to convert 5V to 3.3V, it could get close to 90%.

Re: 6V to 5v ~1mA max. typical 200uA. efficiency >90%

mtwieg said:

What are your ripple voltage requirements on the outputs?

Click to expand...

up to 0.01v is ok. but frequency should be less than 50khz.

mtwieg said:

I think you mean the LTC3632? That gets pretty close to the OP's requirements, but its efficiency seems to drop off below 1mA. Maybe if it's used to convert to 5V, then a charge pump is used to convert 5V to 3.3V, it could get close to 90%.

Click to expand...

most of the times(~90% - 1.9sec) my circuit only consume 200µA and according to datasheet LTC3632 consume at least 125µA it self. (active mode)
So efficiency is less than 60% (regardless of converting efficiency)

Re: 6V to 5v ~1mA max. typical 200uA. efficiency >90%

carp3 said:

up to 0.01v is ok. but frequency should be less than 50khz.

Click to expand...

Well this probably makes a buck converter impossible, since most low power bucks get high efficiency by using a burst mode (which means high ripple).

The closest you can probably feasibly get is using a LDO to regulate down to 5V (so your efficiency from that would be 83-96% max), then using a charge pump cascade to go from 5V to 3V. Overall I wouldn't expect efficiency higher than 80%.

also more than 3mhz is ok.