Quote from: wndsnb on April 28, 2021, 07:27:42 PM

From what I can tell from the log, the miner starts mining and it takes about 3 minutes for the chip to overheat.

Quote from: amirheavy666 on April 21, 2021, 05:42:09 PM

AntMiner

Chain# ASIC# Frequency(avg) Voltage Consumption (W) GH/S(ideal) GH/S(RT) Errors(HW) Temp(PCB) Temp(Chip)

3 44 450 16.5 494 13,305.60 7,857.16 0 43-49-47-53 71-61-80-62 Auto-tuning

I'm assuming that was captured during the 3 minutes, and one chip temp is already up to 80 dec, significantly higher than the others. So it looks like it runs for a little bit, and that chip just keeps getting hotter until it hits the 90Deg limit and shuts down. You could confirm this by just looking at the status page and refreshing constantly during bootup, monitoring the temperature. If it just gradually rises to failure, then I don't think it looks like a faulty sensor, but just a bad ASIC or a bad connection to its heatsink.

These miners have a lot of issues with heat sinks, many times the copper plating that interfaces the chip to the solder that holds the heatsinks on delaminates, which can cause a little gap that will kill the efficiency of the heatsink. It is very possible that if you shake that miner a bit, or if you push on that heatsink, it will just fall right off. I assume you've already checked that there isn't a loose or missing heatsink? And that the heatsinks are clogged up with anything?

You may be able to limp along if you block the 2 empty slots somehow. With them blocked, a lot more air will be forced through the heatsinks and it might be enough to keep it below 90.

Dear wndsnb
this status is with one hashboard and two other hashboard was removed
Chain# ASIC# Frequency(avg) Voltage Consumption (W) GH/S(ideal) GH/S(RT) Errors(HW) Temp(PCB) Temp(Chip)

3 44 450 16.5 494 13,305.60 7,857.16 0 43-49-47-53 71-61-80-62