Re: Samsung LE32M86BD continuous power cycle
This board only has two small BGAs on it: the Samsung DNIe engine, and a Micronas device. The Samsung relied just on the shield for cooling, the Micronas had a small heatsink attached with gooey thermal adhesive. There's another larger chip under a heatsink but that's a QFP package.
I decided to reflow the Micronas. The board is cooling now, we'll see what the outcome is. Edit: Same thing. It's marked as "OPTION" in the service manual anyway... On the other hand, i pressed with my finger on the DNIe engine (didn't bother to put the shield back on the board), and it stayed on after just a minute of being powered up. Garbled screen tho. So, the DNIe engine BGA is up next. Funny thing is - the DNIe chip barely gives out any heat.
Edit 2: Finished with the DNIe too. Powered up first time, but stupor: Garbled screen. Heating up the board while it is working and power cycling the set afterwards brings the picture nice and clean. So the power issue and the garbled screen issue aren't one and the same. Grr. I'll now heat the devices one by one using a smaller nozzle, and see which one is buggy. The only one left is the big TQFP... Or maybe that crystal oscillator on the board is bad... We'll see.
Edit 3: Not entirely sure as i got pissed and removed the smaller nozzle and heated the whole board, but it went like this:
Stupid capacitors. Gonna pull them one by one and check for ESR.
This board only has two small BGAs on it: the Samsung DNIe engine, and a Micronas device. The Samsung relied just on the shield for cooling, the Micronas had a small heatsink attached with gooey thermal adhesive. There's another larger chip under a heatsink but that's a QFP package.
I decided to reflow the Micronas. The board is cooling now, we'll see what the outcome is. Edit: Same thing. It's marked as "OPTION" in the service manual anyway... On the other hand, i pressed with my finger on the DNIe engine (didn't bother to put the shield back on the board), and it stayed on after just a minute of being powered up. Garbled screen tho. So, the DNIe engine BGA is up next. Funny thing is - the DNIe chip barely gives out any heat.
Edit 2: Finished with the DNIe too. Powered up first time, but stupor: Garbled screen. Heating up the board while it is working and power cycling the set afterwards brings the picture nice and clean. So the power issue and the garbled screen issue aren't one and the same. Grr. I'll now heat the devices one by one using a smaller nozzle, and see which one is buggy. The only one left is the big TQFP... Or maybe that crystal oscillator on the board is bad... We'll see.
Edit 3: Not entirely sure as i got pissed and removed the smaller nozzle and heated the whole board, but it went like this:
- TQFP to 80 deg C? Garbled.
- Micronas to 80 deg C? Garbled.
- DNIe to 80 deg C? Garbled.
- Whole area, including TQFP, Micronas and SMD electrolytic capacitors to 80 deg C? Running in spec...
Stupid capacitors. Gonna pull them one by one and check for ESR.
Comment