David Douard @david@social.logilab.org

Let's see, will this make noise again? The original IC (TAS5342A) failed quite hard and burnt a few traces... (It's a D class power amp from a Marshall Stanmore speaker). Parts of the PSU died also (already fixed them). Not sure if the PMW controller before survived either (it does generate square signals on its PMW outputs, but couldn't see any width modulation).

Tiens j'ai de la compagnie à l'atelier. Voilà-t-y pas que il faut que je fasse gaffe maintenant quand j'ai besoin d'un bout de ferraille !

Victory!

En attendant, nothing like a good ol' brake job! (Avec ma douille 8 pans de 90mm de fortune)

Alors le moteur Lada, il m'embête un peu là... Surtout les pièces introuvables en fait, parce que maintenant qu'il tourne vaguement correctement, il s'avère pas mal fuir d'huile par cet adaptateur pour le filtre a huile, félé, et semble-t-il introuvable...

Okay I think the replacement transformer does not exactly meet the expected specs... It looks OK, it has R and L values for both primary and secondary sides are in the right ball park, but the response seems pretty bad... First pic is the original xformer (which still have one secondary OK), the second is the replacement one... "That's your problem, lady!"

And this CCFL power supply is driving me nuts. Replaced the xformer (which was bad), I fried and replaced the controller chip (FAN7314), checked (in circuit) all (most?) the components (resistors, capacitors, diodes, transistors), checked the CCFL with another screen PSU... Still not working

My bench is getting a bit messy...

Let's put the new one (shipped without the O-ring obviously)

Yeah, OK, the throttle valve actuator of my car is a bit dirty. Well, the electronic module is. Full of oil. And dead.

And the "finished" product

The (almost) never ending parcel delivery

And it works with vMix!

I even put these pinouts correctly. How insane is that?

As a result, I've been able to assemble a working (!) board

ok so I finally received the parts (the boards have been delivered more than a week ago). Since I forgot to put a nice ICSP header on these boards, I had to find a hack^Wsolution. I've used one of the boards as a socket. I've populated it with the bare minimum (cristal and its pair of capacitors, the reset pull-up resistor and a couple of decoupling caps) and I've soldered wires on pads used for ICSP; a good ol' Uno et voila, I could burn a booloader (MiniCore's one)
https://github.com/MCUdude/MiniCore#bootloader-option

And the culprit was this tiny capacitor (the CR2032 is to give a scale, it's 1.2x0.55mm), it was mesuring 1.1kohms

The EC/LPC IO ship, which the CMOS battery powers directly, is an IT8728F, and if I have not been able to find a schematic for this motherboard, I could find a few gigabyte ones using this chip, thus having the same schematic around it.

Just repaired this motherboard, from a Shuttle DS437. It was booting and working ok, but the problem was it could not keep BIOS settings. In fact, it was depleting the CMOS battery within a few weeks (maybe a couple of month)

That moment you realize you forgot to put the ICSP header on the PCB you sent to manufacuring yesterday...😭