Breadboarding the power supply

Carrying on the “I plugged it in and it just worked” theme, this is the test version of the power supply for the main processor of the Mitochondrion. This feeds a nice steady five Volts to the processor. It should solve the problem of the processor telling the rest to turn on, the rest turning on and sucking up all the power, the processor then dying coz everything else is turned on, and nothing gets turned off coz the processor is ded, and then you’re stuck until the batteries run down.

For the non-geeks, the clever part is the tiny chip on the little green board in the middle. For the geeks, it’s a MAX710 step-up, step-down DC-DC converter. 5V, 500 mA, damn high efficiency.

I plugged it in, it worked. The glare is the LED to prove it worked. It probably worked coz I’ve learnt to spend half an hour checking every little thing before plugging it in.

Of course, when I say it worked, I mean I plugged it in to a wall plug and it kicked out five Volts. I have yet to test its stiffness (quiet in the cheap seats, I mean how steady the output is when the input is bouncing around from three to six Volts in about a millisecond). But we’ll save that for another day.

When the final version is built, it’ll be 25 mm by 15 mm. That’s not bad for something kicking out 2.5 Watts.

The Gannt chart for the next step of the Mitochondrion has forty-seven steps. I’ve ticked off another seven this week, taking me up to twenty-one. This project planning thing is psychologically valuable, otherwise I’d be just looking at the Heap of Stuff to Do, and feeling intimidated by its seeming endlessness.

8 thoughts on “Breadboarding the power supply

  1. That’s pretty cool!
    So why have one PS? Why not one tiny one just for the proc and another for the drivers? On the boards I do, we usually provide separate, filtered supplies for the power section and the processing section, and separate grounds joined together at a single point usually at or near the biggest power-drawing component.

    That 710 is a nice chip.

    1. This is the tiny one for the proc, coz it’s also running several other sensors and an audio amp on the same board. Everything else runs off the unregulated battery power, but the drivers for everthing else are happy down to 2.3 V, so should be okay. Or at least, they’d better be, coz there’s no more room for more power supplies.

      1. What sort of battery are you considering? (purely out of curiosity)

        For instance, I would have thought a 6V gell cell (aka SLA) would have enough capacity to minimise voltage bouncing under a significantly varying load. Not that having a regulator for the critical bits is a bad idea…

        1. 8 AAA NiMHs. Given the peak load is only 10 Watts, I’m surprised how much the voltage goes all over the place. But it does.

          I’d love to have more batteries in there, but it’s seriously volume constrained. I’d also love to use lithiums, but that’s saved for the next version, if I’m stubborn to go through all this again.

    1. I haven’t been telling people, though a few have worked it out. Still, I’m interested to hear people’s guesses.

      Anything with a thousand LEDs is going to be cool, but that hat is downright fantastic.

Leave a Reply

Your e-mail address will not be published. Required fields are marked *