Search results

That'd be expected, the capacity is 1/10th of the original, with much lower voltage. Still not bad. I remember some CR2032s not lasting longer in PCs anyway.

Any return on how long these machines keep time with a coin cell? I though I left one in a Performa here but it seems not.

So, just confirmed on my PB520c, Vbatt is drawn almost directly to charge the battery (it drops at 11.7V), but it's disconnected when it thinks it's fully charged, so I only have to implement current limiting.

Just tested myself for the sake of it, it happily triggers overcurrent protection at 1.3A if I let it charge on a bench supply. So definitely needs some circuit to calm it down.

Updated the model to make it easier to assemble with the meter.

Needs some more margins but it works:

So, the only thing I hate more than people forcing me to published unfinished work because they are redoing what I spend nights already doing, is people wasting time doing what I already did… https://github.com/mmuman/cad-mac-pb500-pwrconn Oh, A voltmeter variant, because people love useless...

So, someone confirmed using 15V Vmain without trouble for a year, I guess we're pretty safe there. That's not a voltage problem but a current issue: with full brightness you'll draw about an amp already idle, so using anything more will try to pull more amps, but since Vbatt is capped at 1A...

Well I guess I'll design a PCB for that to add inside, since all the boards I find on aliexpress are too large. More fun. Already found a possible chip to use.

Just checked: the charge circuit pulls 1A from Vbatt for 10s when initially plugged in, so it goes down to 12V. So yeah, quite braindead. Also, it only starts charging if Vmain is also present, probably to avoid pulling from Vbatt for the CPU. Now oddly enough, if you remove Vmain and the...

No, it's only Vmain. While I did rebuild my battery, I think most of them around are dead anyway, so for those who don't care that'd be enough. I know there are 2 separate PSUs, now I don't know exactly why. Are you sure it's because the laptop doesn't cap the current or just because it was...

I really want to confirm from schematics/PCB, but at least testing with AAUI+VGA output, HDD, full brightness, starting Netscape and using the floppy works OK.

OK, I was reluctant because I wasn't sure it'd like 15V but it seems people are doing it regardless, so might as well make it without extra wires… Still a few tweaks to make assembly easier but it works.

Oh dear, one more thing I spend entire nights on just for people to redo again… I know, I should publish stuff someday… I was waiting for the resin printer around to be set up, I'd have published mine then… In any case, no need to cut metal, the shield of an RCA plug works just fine.

The specs are not very clear how much of the MMU each core actually implements…

Interesting, but if you expect to run OSes on this unmodified, it won't work: Embedded PPC chips only support compatibility with the userland programming model, not the supervisor one, specifically regarding things like the MMU. So you'll need some things to emulate it.

The other way I'd say yes, this way I'm not sure, you'd have to check the developer notes for both models, I recall they changed a pin function somewhere in the line because the ROM wasn't on the board anymore.

Maybe you can find a hackerspace or a fablab near you that have the required tools?

I suppose nobody has schematics for those… I also have an MDD which doesn't start and was given to me in this state…

Btw, having seen this thread about C22, I just confirmed mine has the + pin of C22 at -5V, which is not really a good idea. It still works but it will probably shorten its lifespan. Not sure what it would do to tantalums either…