I have removed every single cap off the board, still no change. the only few left are sitting on top of the GPU itself out of around 30 on there 5 are less then 1 ohm
I tested the GPU itself for a general short (tho it isn't foolproof) but never tested the card itself.... Most graphics cards fail due to bad BGA connections as we know.... So getting rid of the lead free crap and putting leaded solder on is a first on my agenda... Can usually remedy any issues as well as expanding the cards life.
that's pretty funny.
as someone who does a lot of work on stuff in unstabe enviroments i remove leaded solder & replace it with tin/copper solder to increase the joint strength so it wont crack again.
people constantly blame lead-free solder because of poor cooling or poor choice of flux during manufacturing / repair.
Interesting troubleshooting adventure. Nice to be able to check out the board with a microscope, too. What you thought was a solder bridge on one of the chips is just a typical way of connecting 2 pins together. They are so close that solder flows over the trace so you can't see it. But, the card would never have worked at all if an unintended solder bridge like that one was really present.
May I make a suggestion that would save you a lot of desoldering and lifting of parts? If you identify a short between V+ and ground, the quickest way to find it is to feed some current into the short using an adjustable supply. Keep the current low, but just enough to warm up the offending shorted part. After a while you can feel what part is warm (or find it with an IR temperature probe). Not too surprising that the chip itself shorted given the high heat of operation. Big points for perseverance!
OK so removed caps on top of GPU and still throws a short... Replaced and marked all components I have removed so far as to not get confused. Decided to push 12v into the card with a thermocouple against each corner of the GPU... Bottom right of the GPU is getting hot but nowhere else.... Time to rip the GPU off again I guess as seems the short is there somewhere. As I have already lifted and reballed the GPU I very much doubt its the BGA connections causing exactly the same fault.
Is it possible that a short has occurred internally on the PCB as it is a multilayer?
Comment