Author

Topic: Antminer s5 board issue (Read 3602 times)

newbie
Activity: 2
Merit: 0
June 09, 2017, 10:34:51 PM
#12
Are there any points to check to verify this? I have taken the heatsink off to check the thermal paste and no chip looked burn or moved. If I can find something ill reflow it. In the log its saying it cant set pins for uart.... rx and tx. Dunno how the srting is set up but with the s1 if the first chip was damaged or lost power it wouldnt set clock and you would be a board down.


Edit:
 voltage I believe on vdd1is .6-.65. Dunno what the first 2 chips (with the crystal) voltage are at. I can poke around more when I get home from work.


I know this is old thread, but wanted to let yall know that if you power on the miner, and while running start pushing on the asic chips on the board it starts to function again.
sr. member
Activity: 453
Merit: 250
May 18, 2015, 06:25:24 AM
#11
Are there any points to check to verify this? I have taken the heatsink off to check the thermal paste and no chip looked burn or moved. If I can find something ill reflow it. In the log its saying it cant set pins for uart.... rx and tx. Dunno how the srting is set up but with the s1 if the first chip was damaged or lost power it wouldnt set clock and you would be a board down.


Edit:
 voltage I believe on vdd1is .6-.65. Dunno what the first 2 chips (with the crystal) voltage are at. I can poke around more when I get home from work.
hero member
Activity: 924
Merit: 500
May 18, 2015, 03:52:49 AM
#10
one board literally does not hash no matter what port or cable i use or clock speed, i tested this with the other fully disconnected. I was looking in the kernel log and its reporting power error. I tested the vdd's and they are all the same as the working board, so maybe its something else.

edit:
Its not hitting 80c, its barely breaking 55c for me. I'm just saying the guy said it got hot and shut off on him.


Your boards are gone. This is big mistake of BITMAIN, their S5 even they stop mining, the are the same power of like when mining. If your miner overheat, the chips are very good piece of silicon, they survive, but they melt soldering and desolder power pins. At least two of them are desoldered. You can use regulated power supply, set it max. to 8A and 9V, set 100MHz and start your S5.

Try to press every bottom side of each ASIC, you will probably find two desoldered which pressed are temporarily working. Do not try it at full Amps or at 12V, it may brake your PCB.

So basically, if you are right, adding a little bit of solder paste and reflowing should put it back to life.

can you put a pic online with which power pins are desoldered on yours?

I didn't search, so maybe you already did.

Mostly GND is gone, but it has GND in the center and Vdd on corners. Check sidehack's thread https://bitcointalk.org/index.php?topic=995675.300 with PCB.
hero member
Activity: 840
Merit: 1000
May 18, 2015, 03:43:12 AM
#9
one board literally does not hash no matter what port or cable i use or clock speed, i tested this with the other fully disconnected. I was looking in the kernel log and its reporting power error. I tested the vdd's and they are all the same as the working board, so maybe its something else.

edit:
Its not hitting 80c, its barely breaking 55c for me. I'm just saying the guy said it got hot and shut off on him.


Your boards are gone. This is big mistake of BITMAIN, their S5 even they stop mining, the are the same power of like when mining. If your miner overheat, the chips are very good piece of silicon, they survive, but they melt soldering and desolder power pins. At least two of them are desoldered. You can use regulated power supply, set it max. to 8A and 9V, set 100MHz and start your S5.

Try to press every bottom side of each ASIC, you will probably find two desoldered which pressed are temporarily working. Do not try it at full Amps or at 12V, it may brake your PCB.

So basically, if you are right, adding a little bit of solder paste and reflowing should put it back to life.

can you put a pic online with which power pins are desoldered on yours?

I didn't search, so maybe you already did.
hero member
Activity: 924
Merit: 500
May 18, 2015, 03:33:45 AM
#8
one board literally does not hash no matter what port or cable i use or clock speed, i tested this with the other fully disconnected. I was looking in the kernel log and its reporting power error. I tested the vdd's and they are all the same as the working board, so maybe its something else.

edit:
Its not hitting 80c, its barely breaking 55c for me. I'm just saying the guy said it got hot and shut off on him.


Your boards are gone. This is big mistake of BITMAIN, their S5 even they stop mining, the are the same power of like when mining. If your miner overheat, the chips are very good piece of silicon, they survive, but they melt soldering and desolder power pins. At least two of them are desoldered. You can use regulated power supply, set it max. to 8A and 9V, set 100MHz and start your S5.

Try to press every bottom side of each ASIC, you will probably find two desoldered which pressed are temporarily working. Do not try it at full Amps or at 12V, it may brake your PCB.
legendary
Activity: 1456
Merit: 1000
May 16, 2015, 10:39:39 AM
#7
I had a new s5 do the same thing. Only 1 board would hash. Bitmain gave me a place to send it to and have it replaced. I was the original owner though and they asked for the order number to verify. I also had two units with bad control boards. Not the best luck from the last batch.

They have a good support team, assuming you bought it from them and it's still in warranty is best option.

They will still help out of warranty but parts will cost.
newbie
Activity: 9
Merit: 0
May 16, 2015, 10:23:54 AM
#6
I had a new s5 do the same thing. Only 1 board would hash. Bitmain gave me a place to send it to and have it replaced. I was the original owner though and they asked for the order number to verify. I also had two units with bad control boards. Not the best luck from the last batch.
hero member
Activity: 840
Merit: 1000
May 16, 2015, 02:00:20 AM
#5
one board literally does not hash no matter what port or cable i use or clock speed, i tested this with the other fully disconnected. I was looking in the kernel log and its reporting power error. I tested the vdd's and they are all the same as the working board, so maybe its something else.

edit:
Its not hitting 80c, its barely breaking 55c for me. I'm just saying the guy said it got hot and shut off on him.

I have a board with similar issue that will come back unclaimed/refused by the RMA recipient after 1 month.
I'm glad they did cross shipping on this one.
The board failed within 2 weeks of delivery from Bitman, temps were never over 60°C
I'll have a look at it to see what can be done to revive it since Bitmain refused the board.

Also, I noticed that the board was losing chips whenthe S5 was unplugged while "hot" like 60°C and not restarted immediatly.
legendary
Activity: 1456
Merit: 1000
May 16, 2015, 12:16:33 AM
#4
one board literally does not hash no matter what port or cable i use or clock speed, i tested this with the other fully disconnected. I was looking in the kernel log and its reporting power error. I tested the vdd's and they are all the same as the working board, so maybe its something else.

edit:
Its not hitting 80c, its barely breaking 55c for me. I'm just saying the guy said it got hot and shut off on him.

I would follow the first one on checking the cords they do a great job on possible faults.

With it not hashing if cords and board look good I would look at second guide. (Be aware always risks with firmware updates).  But that is the way to try to fix it if it's software.  When you update make sure keep settings is checked to make your life easy.
sr. member
Activity: 453
Merit: 250
May 15, 2015, 11:03:33 PM
#3
one board literally does not hash no matter what port or cable i use or clock speed, i tested this with the other fully disconnected. I was looking in the kernel log and its reporting power error. I tested the vdd's and they are all the same as the working board, so maybe its something else.

edit:
Its not hitting 80c, its barely breaking 55c for me. I'm just saying the guy said it got hot and shut off on him.
legendary
Activity: 1456
Merit: 1000
May 15, 2015, 10:26:24 PM
#2
If you are hitting 80c that is not good.  It will eventually hurt your miner to run so hot, the 80c temp shutdown is there for a reason.  But what speed are you hashing at?  Depending on this anwser I would look at these two Bitmain guides:

First I would look at this and check connections - https://bitmain.zendesk.com/hc/en-us/articles/204875015-How-to-diagnose-S5-hashing-PCB

If that does not work then look at software - https://bitmain.zendesk.com/hc/en-us/articles/203559969-My-S5-hashes-at-about-600GH-S
sr. member
Activity: 453
Merit: 250
May 15, 2015, 06:58:20 PM
#1
i bought an s5 off ebay and the seller said in the listing that only one board hashes (which is fine because it was a decent package deal). Supposedly it over heated and shut down and only one board works now. Shut down over 80c was selected but who knows if that was the case before hand. I have tried all the reset buttons and swapping the control board around as well as the cables. i have poked around in the log and it seems to see the chips but they wont accept work or set clock? Bitmain also has the chip count set to 32 on the s5's (lol?). So maybe there is a burnt out component? I am willing to use this board to test and poke around so we can get as much info like the s1's. Anyone know what the D1-D4 leds are for? If its dead its dead but it wont hurt to poke around.


Little info:

1. With the board plugged in it draws about 25-35 extra watts
2. All the chips will show up as "----"
3. If I leave it on enough one will come online "O"  after a little while it will error out "X"
4. I tried setting clock pretty low, 2 had an error "XX" right away while the others were "---" (different than the one that will come online and x out after)
5. The red light "RF" never turns on or blinks

pastebin of bad board:
http://pastebin.com/VJF0Mecq
Jump to: