I thought I’d share my supposed fix for this. All 3 chips were reflowed ramp up temperature of 120c stage 2, 180c @2mims stage 3, 280c @1min stage 4 @400c @30seconds ramp down stage 1 280c 2mins stage 2 120c @1mins. Bit of a weird ramp but it was based on the flux bubble activity. I used amtec NC-559-V2-TF and did not clean it off. While still warm I added general thermal paste and let it sit in drier for a few hours. The thermal paste was enough to keep the heat sink sticky to the chips. Reassembled the unit and low and behold it hasn’t shut down after streaming OBS and I’ve been testing it for the last week. I couldn’t even get 15mims out of it without it powering down. And now I’ve saved $400aud and don’t have to buy a new monitor.
you're lucky, 400C is a bit much for the peak
but 400C with airgun is much less below chip, depends on your airflow setting
I never had more than 300-330C on metcal 5000, and the temp was the real temp of the chip, we had thermocouples on top/bottom side of pcb. the air flow and heating power is gigantic for those machines comparing with airgun. of course, it was solder without pb, the laptop/desktop nvidia cards sh...ty years rememberings
I tried for some pcb's up to 400C, the result was a dead or semi-dead video chip
good work, you have 'good hand'