I'm using a 28X1 to control 40-odd RGB LEDs using eight PCA9532 i2c drivers, powered by 4 AAA NiMHs. It all looks most trippy, but testing has found a failure mode that I'm trying to find a way around.
When the batteries are tired, the system runs just fine, but is no longer programmable. What I think is happening is that the Picaxe has enough voltage to run, but not to be programmable. Is this possible? Switching all the LEDs on can kick the battery voltage from 5.5V down to 3.6V instantaneously. The charger that I'm using has enough grunt to stiffen the voltage so that the kit is programmable, provided the batteries are not too discharged.
I'm wanting to use the Picaxe to monitor the battery state (via a voltage reference diode), so that it can turn off the shiny lights before the batteries get munted. However, the drivers seem to be able to cope with low voltages better than the Picaxe, raising the frightening possiblity that the Picaxe might die with the drivers still on full-whack, leaving the battery pack to get ruined.
So, my current idea for sorting this out is to provide the Picaxe with it's own regulated power supply, so that it's always programmable and can switch the thing off before the Picaxe falls over. I've been digging around on the Maxim and Linear sites to find a dc-dc chip that can take in 6-3V, give out a nice steady 5V and 100 mA so keep the Picaxe happy. Haven't quite worked out how to find what I'm looking for yet. Oh, and it'd be nice to have a bit more current for the i2c memory chip, SRF005 range finder, and the couple of op-amps in there for input processing. Also, it needs to be tiny, coz everything's going inside a 25 mm diameter tube and there's quite a bit in there already. Surface mount isn't a prob, in fact it'll probably be needed to fit everything in.
Has anyone had any experience of such a power supply before? Any pointers or pitfalls? Or is there a better way around this problem?
When the batteries are tired, the system runs just fine, but is no longer programmable. What I think is happening is that the Picaxe has enough voltage to run, but not to be programmable. Is this possible? Switching all the LEDs on can kick the battery voltage from 5.5V down to 3.6V instantaneously. The charger that I'm using has enough grunt to stiffen the voltage so that the kit is programmable, provided the batteries are not too discharged.
I'm wanting to use the Picaxe to monitor the battery state (via a voltage reference diode), so that it can turn off the shiny lights before the batteries get munted. However, the drivers seem to be able to cope with low voltages better than the Picaxe, raising the frightening possiblity that the Picaxe might die with the drivers still on full-whack, leaving the battery pack to get ruined.
So, my current idea for sorting this out is to provide the Picaxe with it's own regulated power supply, so that it's always programmable and can switch the thing off before the Picaxe falls over. I've been digging around on the Maxim and Linear sites to find a dc-dc chip that can take in 6-3V, give out a nice steady 5V and 100 mA so keep the Picaxe happy. Haven't quite worked out how to find what I'm looking for yet. Oh, and it'd be nice to have a bit more current for the i2c memory chip, SRF005 range finder, and the couple of op-amps in there for input processing. Also, it needs to be tiny, coz everything's going inside a 25 mm diameter tube and there's quite a bit in there already. Surface mount isn't a prob, in fact it'll probably be needed to fit everything in.
Has anyone had any experience of such a power supply before? Any pointers or pitfalls? Or is there a better way around this problem?