On 16/12/11 07:42, Electron wrote:
> Nice (although quite distressing for You ;( ) experience, thanks for
> sharing. Just curious..: if the board was in a metal box, with a wire
> going from circuit ground to the metal box, would (when the box was
> closed with a metal lid) the circuit have behaved well even with the
> lamp on?

If it was in a grounded tin box (Altoids tin?), it probably would have=20
worked fine...

> (sure there are possibly still wires going out and acting like antennas,
> but given the very low frequences emitted by the lamp, this should not
> pose a problem)

It sure screwed up the sense amp!

I tweaked a few component values (reduced resistor values in the=20
feedback loop) which has mostly eliminated the effect.

Anyway -- this is all part of a battery analyser. I'm basing the charger=20
logic on the Atmel AVR450 appnote (and a similar Microchip appnote whose=20
ID# I can't remember at the moment) but redesigning the switchmode power=20
supply. AVR450 has a bit of a... "reputation". Specifically, the FET=20
gate driver falls over at high frequencies, and usually ends up blowing=20
the drive FET outright, or running it in a linear mode and cooking it.

My "V2.0" design is based on a SEPIC converter, a National Semi=20
synchronous gate driver and a pair of FETs. "If it's worth doing, it's=20
worth OVERdoing."

Then there's the discharger and relay driver to deal with... Ideally I=20
want some kind of reverse-polarity protection on the relay driver -- if=20
the battery is hooked up backwards, the O/P relay will refuse to switch=20
over. That's a job for later though... mainly because I can't think of a=20
sane way to do it.

--=20
Phil.
piclist@philpem.me.uk
http://www.philpem.me.uk/
--=20
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist
.