Print

[stepleton]

I've seen people in a few venues now describe resurrecting their DataPower 1.8A PSUs by replacing the SG3524 regulator chip found on the riser card in the middle of the PSU. I've done the same thing myself, in fact. Here's someone discussing it on Mastodon; here's another on 68kmla; I think I've seen more, even here, but my search skills aren't good enough to recover those conversations if we've had them.

I'm getting a feeling that these parts are now nearing their end-of-life. Fortunately, replacements appear to be widely available --- on eBay as well as major vendors like Digi-Key.

The failure seems to me common enough that perhaps this is a good thing to mention in the Lisa FAQ.

[jamesdenton]

The failure seems to me common enough that perhaps this is a good thing to mention in the Lisa FAQ.

Thanks, Tom.

I checked out those threads and couldn't find any definitive symptoms that might point someone to that regulator. Happy to add this to the LisaFAQ (lisafaq.sunder.net) but some additional details besides "replace the regulator if all else fails" might be useful. Or, if it's good practice to replace the regulator no matter what, that's cool too.

[stepleton]

I'm not sure and hope one of the wiser folks on the forum might offer some suggestions for what to say.

The symptoms that confronted me were basically "PSU does not go". The +5V standby was available, but you couldn't turn the system on.

The SG3524 regulator IC will be on the secondary side of the PSU, so in principle it's not the most dangerous place to be poking around with a 'scope, but when I was troubleshooting my DataPower PSU, I wasn't feeling that brave. I first checked discrete components with the PSU off and caps drained, and when I found nothing suspicious (diodes appeared to be diode-ing, etc.) I started pulling ICs and trying simple ones in a breadboard. All the comparators worked fine, at least in response to me twiddling the knobs on a bench PSU (so a very low-frequency signal of course). I didn't think for too long about how to test the SG3524, which seemed like it would be a lot more complicated; if I was going to take it out anyway, I might as well just replace it. So I did and everything was working again.

[jamesdenton]

That works for me. If someone's going through the trouble of replacing capacitors and other serviceable components, replacing the IC wouldn't be too much of a stretch.

Is this the one?

https://www.digikey.com/en/products/detail/texas-instruments/SG3524N/380218

[stepleton]

It should be; it looks like the pinout matches the DataPower schematic, though some signals have different names. I think it's correct.

[nekonoko]

I checked out those threads and couldn't find any definitive symptoms that might point someone to that regulator.

In my case the power light would come on when the switch was pressed, and the Lisa's speaker made a crackling, almost electric noise (similar to a Tesla coil) before shutting off. I found it described as a "machine gun" noise from the speaker in another thread suggesting replacement of the SG3524, and it seemed similar enough to give it a try (and it wound up fixing it). The replacement SG3524 I used was made by TI.