S17e 64TH bad board. or PSU?

[scissors14]

I've been trying to repair this s17e i bought of alibaba.

I managed to take it apart and reconnect the ribbon cables and even do some re-soldering around the ribbon connections.

The middle hashboard is not hashing, it appears to pick up all 135 asic chips but it says bad voltages or no temp sensor?

Here's the log. please any help would be much much appreciated.

https://drive.google.com/file/d/1pHKyRV1g-ZFMP7w5SWJ3nzC4SNUPmFtD/view?usp=sharing I have spent many hours trying to figure this out, I was hoping someone with some experience would be able to identify what is wrong. this is my first ASIC machine and i was  hoping to get it working right,

2021-02-11 01:07:46:temperature.c:754:get_temp_info: ERROR: chain 1 can get NONE temp info or temp value abnormal, power it off, is that the problem?

Cheers,
Scissors

[BitMaxz]

It seems that the middle hashboard has a faulty temp sensor but it could be also a PSU issue.

I suggest you if you are going to share the logs here on the forum use this https://pastebin.com/ make sure to copy the whole kernel logs and paste the Pastebin URL here.

Can you try to run only one hashboard(The middle one) and let see if it will run without any issue. If it runs then you have a PSU issue or your power source is not giving enough power. It should be run only on 200-240v according to their specs. Also, try to directly plug the unit into the wall outlet because a very thin extension has limited watts it can also cause a fire if you continue using a thin extension.

And I suggest you physically check the middle hashboard maybe there is some loose heatsink(This is a common issue when it was not carefully delivered) that's why it stop hashing/low hashrate.

You can find another troubleshooting guide from this link: https://support.bitmain.com/hc/en-us/articles/360037629194-Antminer-S17e.

[mikeywith]

2021-02-11 01:07:46:temperature.c:754:get_temp_info: ERROR: chain 1 can get NONE temp info or temp value abnormal, power it off

is that the problem?

It is a problem, the control board doesn't seem to be receiving any signal from 1 or more of the 4 temp sensors on that chain, the default firmware needs 4 out of 4 sensors to work, custom firmware like Vnish will need at least 1, but I am not sure if they have developed something for the S17e version, you will have to check asic.to and AwesomeMiner.

Assuming you indeed have a bad temp sensor then that will fix it, but keep in mind that it's not so likely, the more likely is you have a problem with one of the chips/heatsink, I have explained this issue in great detail here.

[wndsnb]

Looks like it is measuring the voltages in for the 15 voltage domains on each board and isn't liking what it sees. I haven't worked on a S17e yet (although I have a dead one waiting in line for my attention), but other 17 series miners I've worked on normally show voltage variation between domains when there is a problem ASIC. Could be a damaged ASIC, or just a bad connection to the PCB.

From my experience, the temperature sensor error normally has nothing to do with the actual temperature sensor, as mikeywith said. I think the variation in voltage domains caused by other issues (failed chip or bad connection) break the interface to the sensor so the control board can no longer read it.

What is interesting is that both Chain 0 and Chain 1 have an issue with the same voltage domain, domain 12 on both of them. Makes me wonder if a heatsink on one is loose and shorting to the next board.