Antminer S17 Pro Power_Lost

[Jaeger061011]

Hey guys,
I Need help with my Antminer S17 Pro. After it tries to set the Voltage to 21.0 V it puts out the Status: Error_Power_Lost. First I tought it is the PSU. But after running it with another the problem is still the Same.
What can I do to find the Issue ?

Thank you

[1715087339]

1715087339

[1715087339]

1715087339

[1715087339]

1715087339

[1715087339]

1715087339

[1715087339]

1715087339

[favebook]

I had a similar problem with my S17e. I figured I'd change PSU but problem persisted. Which concludes that it must be control board. If you have another, test with it. Otherwise keep power cycling it and it might work eventually.

[mikeywith]

I had a similar problem with my S17e. I figured I'd change PSU but problem persisted. Which concludes that it must be control board. If you have another, test with it. Otherwise keep power cycling it and it might work eventually.

I still think it's more of a power-related issue, are you sure you tried a working PSU?

OP, if the miner isn't under warranty, open the PSU and give it a good dusting, and then remove two dashboards and keep only one plugged in, start with keeping chain 0 and then try, if the problem persists then remove chain 0 and install chain 1 alone, do this for the 3 boards, there is a good chance that one of teh hash boards is casuing this issue.

Also, you must try the obvious basic troubleshooting which is reset and flash a different firmware.

[favebook]

Let's say I believe him that he tested with new or already good (tested) PSU. Then it would practically be impossible to be anything else. Either control board or connecting cables. I doubt hashboards have control units that can directly talk to PSU - that is why there is control board in middle of them afaik.

But yeah, he should totally try different firmware.

Also, it would help a lot if he tells us how old the unit is, when did he buy it, how long it worked for him, in what conditions, at what hashrate and have previous owners ever had this problem before (if he is not the first owner).

Would be helpful if OP knew what triggered this event. Maybe it was blackout or power surge?

[philipma1957]

so if i have a running s17pro and a dead s17pro.

and i moved the psu from the good one to the dead one.

and it does not turn on and mine.

gives the same error .

what do i do next.

i take the dead psu move it to the running unit.

if the unit runs the issue is not the psu.

this means controller boards or hashboards. or connecting white wires.

next step for me would be move the controller from the dead miner to the good miner

if the miner works it is not the controller.

i would then pull one connecting wire from the dead unit.

put it to the good unit.

if the good unit works it was not the connecting wire 1

i would leave the swapped wire 1 on the good unit
then use wire 2 from the dead unit
along with wire 1.

if it runs add wire 3.

at this point if the good unit runs

you have tested the psu from the  dead unit
you have tested the three white units from the dead unit
you have tested the controller from the dead unit.

all work

so now do fan 1 then fan 2 then fan 3 then fan 4

all that is left is hash boards

do them one at a time.

yeah it sucks as you may need to do

1 psu swap dead to good
1 controller swap dead to good
3 white swaps one at a time dead to good.
4 fan swaps one at at a time dead to good.
3 hashboard swaps dead to good.

sooner or later you will find the dead part.

[mikeywith]

[...]

You make perfect sense, many things in these miners are hard to be explained because they don't make sense, but I have personally had issues that did seem to be PSU related and ended up being hash boards related, I initially thought of it the way you do now, it's pretty simple, the control board controls both the PSU and the hash boards, so any power related issue must be PSU or control board, but well, shit happens.

I encoured a similar power issue on a T17, I don't remember the exact error in the kernel log, but it was "power xx error", I tried a working PSU and a working control board and that didn't fix it, I played with the hash boards and the miner worked, it turned out that one of the hash boards was causing the issue, the miner worked perfectly with 2 boards.

What I did next was installing that "bad" hashboard on another T17, and guess what? it worked without an issue, I installed the hash board which I took out from the second T17 on the first T17 and it worked as well! do I have a technical or scientific explation for this? No, and i am sure even Bitmain themselves don't.

[Jaeger061011]

Thanks for your help guys. I do the troubleshooting as you said. The Power_Lost passed away ( I have no clue how) but I found out that the hashchains maybe caused this error. I put the chains in another working Miner and found out that all three hashchains were dead. 1-2-3 0 Asics 0 Asics and 40 Asics. I had to contact the seller, I bought them used and this one never works good from the beginning.

Thank you and keep the good work up, other miners needs your help too  

[philipma1957]

I want to thank you for followup. So bad hash boards it is.

[Jaeger061011]

I got another Question. Can hashboards get too cold ?  Its 12 c° / 54 F now in my area. The big fan blows air from the outside directly inside to the miner inlet fans. After I let one miner blow up the hot air in the one miner with the Issue, the chains came back. 

[favebook]

I got another Question. Can hashboards get too cold ?  Its 12 c° / 54 F now in my area. The big fan blows air from the outside directly inside to the miner inlet fans. After I let one miner blow up the hot air in the one miner with the Issue, the chains came back. 

Well, it can, but 12C is nowhere near cold enough for it to make problems on good unit. I've had my machines run perfectly fine with inlet directly to miners of -5C.

HOWEVER, since your unit is already making problems. It could be possible that cold air can make it even more problematic.

[philipma1957]

Following your idea above

I wonder if two hashboards are good and one is shit.

Have you (op) tested just one at a time?

[mikeywith]

Your ambient temperature could have been causing that,  but that is unlikely since 12c is within the normal temperature range, what i think happened is that the heat from the second miner probably caused the heatsinks in the bad miners to stick well and fixed the loose contact they had with the chips.

This probably won't last long, it's like all the other temporary solutions, if returning the unit is an option, you should go for it.