Nothing else I could do for that rig. Just died. Oh well.
Topic: ANTMINER S5: 1155GH(+OverClock Potential), In Stock $0.319/GH & 0.51W/GH - page 15. (Read 451116 times)
November 25, 2015, 09:56:36 PM
November 25, 2015, 06:16:27 PM
I've fixed this before, but forgot what I've done. Can't remember at all. All dashes on the first row, and then sometimes a few X's on the second row. Sometimes starts hashing around 300-400G, but never above. Tried another PSU. This is my fathers rig, so I haven't asked him to try another controller. I'm guessing controller, hash board, or possibly a cable, but I doubt it is a cable. Any guesses? I just have remote access to it right now, but can ask my Dad to try a few things.
First thing i would do is turn the controller 180 degree and effectively swap the blade. If its now the bottom blade (on the webgui) thats dead, then you know its not the controller.
Assuming he tried a power cycle already.
You have only 26 chip on this board.
November 25, 2015, 05:50:25 PM
One of the weirdest issues. If one hashing board dies out, you'd figure at least the other one will work. All configurations were tried. Different controller, cables, PSU. Ok, so concluded that one hashing board died. Fine. Thought could get the other "good" board hashing, but that won't even hash. Restored image from SD card. Tried different firmwares. The webGUI pages come right up fine but hashing speed just indicates 0. The other "good" board doesn't want to hash either, but indicates all zeroes. The controller was put on another rig and hashed fine. The only thing I could possibly think is this rig got ZAPPED electrically somehow.
http://photopiks.com/minerdashes.png
After flipping the hash boards.
http://photopiks.com/minerdashes3.png
This is the log, hard to dicypher if you aren't a programmer.
http://pastebin.com/7izjejfX
http://photopiks.com/minerdashes.png
After flipping the hash boards.
http://photopiks.com/minerdashes3.png
This is the log, hard to dicypher if you aren't a programmer.
http://pastebin.com/7izjejfX
What happens if you just disconnect the bad board?
November 25, 2015, 05:35:51 PM
One of the weirdest issues. If one hashing board dies out, you'd figure at least the other one will work. All configurations were tried. Different controller, cables, PSU. Ok, so concluded that one hashing board died. Fine. Thought could get the other "good" board hashing, but that won't even hash. Restored image from SD card. Tried different firmwares. The webGUI pages come right up fine but hashing speed just indicates 0. The other "good" board doesn't want to hash either, but indicates all zeroes. The controller was put on another rig and hashed fine. The only thing I could possibly think is this rig got ZAPPED electrically somehow.
After flipping the hash boards.
This is the log, hard to dicypher if you aren't a programmer.
http://pastebin.com/7izjejfX
After flipping the hash boards.
This is the log, hard to dicypher if you aren't a programmer.
http://pastebin.com/7izjejfX
November 25, 2015, 05:02:37 PM
Heylo folks,
Looking for a little help. I didn't notice when, but sometime within the past month or so, one of my S5's stopped hashing or would produce minimal hashrate. Today I reset it back to the factory default and it won't hash w/ the default config. On the Status dashboard, it shows 30 asics on one board/chain and 26 asics on the other.
There's nothing visually wrong with it, to my eyes. Can you provide me with some guidance?
Looking for a little help. I didn't notice when, but sometime within the past month or so, one of my S5's stopped hashing or would produce minimal hashrate. Today I reset it back to the factory default and it won't hash w/ the default config. On the Status dashboard, it shows 30 asics on one board/chain and 26 asics on the other.
There's nothing visually wrong with it, to my eyes. Can you provide me with some guidance?
Swap PSUs, cables, and/or controllers with other miners, try to isolate the issue. It would be unlikely for both hashboards to go bad although not impossible. Take the plastic guards off and look closely for any physical damage/burns. That's roughly where my knowledge ends
, maybe others can help more. 
November 25, 2015, 04:26:26 PM
Heylo folks,
Looking for a little help. I didn't notice when, but sometime within the past month or so, one of my S5's stopped hashing or would produce minimal hashrate. Today I reset it back to the factory default and it won't hash w/ the default config. On the Status dashboard, it shows 30 asics on one board/chain and 26 asics on the other.
There's nothing visually wrong with it, to my eyes. Can you provide me with some guidance?
Looking for a little help. I didn't notice when, but sometime within the past month or so, one of my S5's stopped hashing or would produce minimal hashrate. Today I reset it back to the factory default and it won't hash w/ the default config. On the Status dashboard, it shows 30 asics on one board/chain and 26 asics on the other.
There's nothing visually wrong with it, to my eyes. Can you provide me with some guidance?
November 22, 2015, 08:30:14 PM
It was easier to just move a good working beagle-bone to another miner. Booted right up and started hashing good, so this does confirm a bad hashing board. 
Never wanted to risk popping out the BBB, thanks for telling me its doable if you have a spare. If its any consolation, the dead PCB is sellable for parts. If you can't fix it, you can sell it easily on the forum.
I would not be surprised if you received PM's about it soon, if you have not already.
I asked my Dad to take apart his bad rig and put it back together. Sometimes just doing that has helped me in the past. The daughter card has 4 connectors on it, so I would imagine it can run 4 hashing boards. I have tested those connectors a year ago with just connecting on the same miner and worked as normal but since they have been running so good I don't touch them. I'm just trying to eliminate everything. There's also a 4-5 pin cable going from the daughter board to the hashing board. Not sure what that is yet.
Yes it can run 4 boards at the same time. But according to Bitmain, thats the way to go (turning the board over). Maybe each side are run through different electronics, so if it dies both board sockets on one side could die.
But since we already know its the board that has a problem. You could try to reflow resoldering. And the 4 pin from one of the board to the controller is there to power the controller.
November 22, 2015, 07:29:48 PM
It was easier to just move a good working beagle-bone to another miner. Booted right up and started hashing good, so this does confirm a bad hashing board.
Never wanted to risk popping out the BBB, thanks for telling me its doable if you have a spare. If its any consolation, the dead PCB is sellable for parts. If you can't fix it, you can sell it easily on the forum.
I would not be surprised if you received PM's about it soon, if you have not already.
I asked my Dad to take apart his bad rig and put it back together. Sometimes just doing that has helped me in the past. The daughter card has 4 connectors on it, so I would imagine it can run 4 hashing boards. I have tested those connectors a year ago with just connecting on the same miner and worked as normal but since they have been running so good I don't touch them. I'm just trying to eliminate everything. There's also a 4-5 pin cable going from the daughter board to the hashing board. Not sure what that is yet.
November 22, 2015, 07:00:20 PM
It was easier to just move a good working beagle-bone to another miner. Booted right up and started hashing good, so this does confirm a bad hashing board.
Never wanted to risk popping out the BBB, thanks for telling me its doable if you have a spare. If its any consolation, the dead PCB is sellable for parts. If you can't fix it, you can sell it easily on the forum.
I would not be surprised if you received PM's about it soon, if you have not already.
November 22, 2015, 06:57:20 PM
It was easier to just move a good working beagle-bone to another miner. Booted right up and started hashing good, so this does confirm a bad hashing board.
November 22, 2015, 06:49:53 PM
I've fixed this before, but forgot what I've done. Can't remember at all. All dashes on the first row, and then sometimes a few X's on the second row. Sometimes starts hashing around 300-400G, but never above. Tried another PSU. This is my fathers rig, so I haven't asked him to try another controller. I'm guessing controller, hash board, or possibly a cable, but I doubt it is a cable. Any guesses? I just have remote access to it right now, but can ask my Dad to try a few things.
First thing i would do is turn the controller 180 degree and effectively swap the blade. If its now the bottom blade (on the webgui) thats dead, then you know its not the controller.
Assuming he tried a power cycle already.
I've re-flashed from the SD card and also updated firmware. Done all the "software" related procedures so far. I think it is the controller, but just a guess right now. I'll ask him to try that. The data cables are cheap and do fall apart easily.
Then i guess software side you can't shut it off for 20 seconds, with PSU unplugged and then replug it. That usually solve minor issues like a couples of X. But a blade line full of X, not so sure. Still worth a try.
I have indicated in the original post that another PSU was tried, hence it has been power cycled a few times.
You actually cannot rotate that beagle-bone board 180 degrees. The Ethernet square drops down into that green board so when you turn the beagle-bone board the Ethernet jack hits the circuit board below since there is no hole/cutout there for it. Yes, miner has been power cycled several times and a BLAH.
Don't pop out and rotate the beagle bone, i'm not suggesting you pop it out and pot in back on the daughter board. The controller, that is the BBB+The daughter board, can be rotated just fine. Its just to swap the PCB's. Its the only way to do it if you don't have long 18 pins cables.
November 22, 2015, 05:01:10 PM
I've fixed this before, but forgot what I've done. Can't remember at all. All dashes on the first row, and then sometimes a few X's on the second row. Sometimes starts hashing around 300-400G, but never above. Tried another PSU. This is my fathers rig, so I haven't asked him to try another controller. I'm guessing controller, hash board, or possibly a cable, but I doubt it is a cable. Any guesses? I just have remote access to it right now, but can ask my Dad to try a few things.
First thing i would do is turn the controller 180 degree and effectively swap the blade. If its now the bottom blade (on the webgui) thats dead, then you know its not the controller.
Assuming he tried a power cycle already.
I've re-flashed from the SD card and also updated firmware. Done all the "software" related procedures so far. I think it is the controller, but just a guess right now. I'll ask him to try that. The data cables are cheap and do fall apart easily.
Then i guess software side you can't shut it off for 20 seconds, with PSU unplugged and then replug it. That usually solve minor issues like a couples of X. But a blade line full of X, not so sure. Still worth a try.
I have indicated in the original post that another PSU was tried, hence it has been power cycled a few times.
You actually cannot rotate that beagle-bone board 180 degrees. The Ethernet square drops down into that green board so when you turn the beagle-bone board the Ethernet jack hits the circuit board below since there is no hole/cutout there for it. Yes, miner has been power cycled several times and a BLAH.
November 22, 2015, 04:30:53 PM
I've fixed this before, but forgot what I've done. Can't remember at all. All dashes on the first row, and then sometimes a few X's on the second row. Sometimes starts hashing around 300-400G, but never above. Tried another PSU. This is my fathers rig, so I haven't asked him to try another controller. I'm guessing controller, hash board, or possibly a cable, but I doubt it is a cable. Any guesses? I just have remote access to it right now, but can ask my Dad to try a few things.
First thing i would do is turn the controller 180 degree and effectively swap the blade. If its now the bottom blade (on the webgui) thats dead, then you know its not the controller.
Assuming he tried a power cycle already.
I've re-flashed from the SD card and also updated firmware. Done all the "software" related procedures so far. I think it is the controller, but just a guess right now. I'll ask him to try that. The data cables are cheap and do fall apart easily.
Then i guess software side you can't shut it off for 20 seconds, with PSU unplugged and then replug it. That usually solve minor issues like a couples of X. But a blade line full of X, not so sure. Still worth a try.
November 22, 2015, 04:27:16 PM
I've fixed this before, but forgot what I've done. Can't remember at all. All dashes on the first row, and then sometimes a few X's on the second row. Sometimes starts hashing around 300-400G, but never above. Tried another PSU. This is my fathers rig, so I haven't asked him to try another controller. I'm guessing controller, hash board, or possibly a cable, but I doubt it is a cable. Any guesses? I just have remote access to it right now, but can ask my Dad to try a few things.
First thing i would do is turn the controller 180 degree and effectively swap the blade. If its now the bottom blade (on the webgui) thats dead, then you know its not the controller.
Assuming he tried a power cycle already.
I've re-flashed from the SD card and also updated firmware. Done all the "software" related procedures so far. I think it is the controller, but just a guess right now. I'll ask him to try that. The data cables are cheap and do fall apart easily.
November 22, 2015, 04:17:37 PM
I've fixed this before, but forgot what I've done. Can't remember at all. All dashes on the first row, and then sometimes a few X's on the second row. Sometimes starts hashing around 300-400G, but never above. Tried another PSU. This is my fathers rig, so I haven't asked him to try another controller. I'm guessing controller, hash board, or possibly a cable, but I doubt it is a cable. Any guesses? I just have remote access to it right now, but can ask my Dad to try a few things.
First thing i would do is turn the controller 180 degree and effectively swap the blade. If its now the bottom blade (on the webgui) thats dead, then you know its not the controller.
Assuming he tried a power cycle already.
November 22, 2015, 04:15:13 PM
I've fixed this before, but forgot what I've done. Can't remember at all. All dashes on the first row, and then sometimes a few X's on the second row. Sometimes starts hashing around 300-400G, but never above. Tried another PSU. This is my fathers rig, so I haven't asked him to try another controller. I'm guessing controller, hash board, or possibly a cable, but I doubt it is a cable. Any guesses? I just have remote access to it right now, but can ask my Dad to try a few things.
November 22, 2015, 10:21:49 AM
Hello,
I also have several problems with some antminers recently.
Does anyone have an idea how this can happen?
All ASICs seem to work, but the miner does not start hashing.
I also have 2 miners where suddenly only hashboard 1 (chain 1) shows all ASICs working, while all ASICs on chain 2 are not shown (only "-").
I am quite confused by this behaviour, testing around with different firmware & resetting to default values did not solve it.
I am open and happy to any suggestions.
Kind regards
I also have several problems with some antminers recently.
Does anyone have an idea how this can happen?
All ASICs seem to work, but the miner does not start hashing.
I also have 2 miners where suddenly only hashboard 1 (chain 1) shows all ASICs working, while all ASICs on chain 2 are not shown (only "-").
I am quite confused by this behaviour, testing around with different firmware & resetting to default values did not solve it.
I am open and happy to any suggestions.
Kind regards
November 17, 2015, 12:39:41 PM
Hey guys,
I need some advices with a S5 that I bought with some damage.
First, I got one blade that have a short (PSU won't start if plugged in and with both PCI-E) and I will probably remove the heatsink and have a look at the thermal paste to see if it's the problem (has I red on a link posted somewhere here to an external site to do so)
Then my other blade does hash most of the time. I managed to make it hash one time at 100M then I manage to push it to 250M but i had 3x (first 3 of the chain) then I rebooted and it doesn't hash anymore at any freq I set.
I tried the lastest S5 firmware and now using Nicehash (just because i wanted to be sure it's not the firmware) which bring me a question. Do we have any firmware that allow us to set the fan Freq + stop the beeper. I couldn't find a firmware that do both at the same time.
I tested with a PSU that is not single rail (I will try a single rail tonight) but my guess is that this blade is about to die and doesn't respond very well to his power input.
Any inputs appreciated !
Thanks
Herb
EDIT: Figured out the rest
First, I got one blade that have a short (PSU won't start if plugged in and with both PCI-E) and I will probably remove the heatsink and have a look at the thermal paste to see if it's the problem (has I red on a link posted somewhere here to an external site to do so)
Then my other blade does hash most of the time. I managed to make it hash one time at 100M then I manage to push it to 250M but i had 3x (first 3 of the chain) then I rebooted and it doesn't hash anymore at any freq I set.
I tried the lastest S5 firmware and now using Nicehash (just because i wanted to be sure it's not the firmware) which bring me a question. Do we have any firmware that allow us to set the fan Freq + stop the beeper. I couldn't find a firmware that do both at the same time.
Any inputs appreciated !
Thanks
Herb
EDIT: Figured out the rest
November 17, 2015, 10:23:21 AM
Soft restarts and reboots didn't help. Ended up having my wife power it down. Think I'll just throw a RPi on it for now.
November 17, 2015, 06:06:50 AM
i've created a script just to soft restart the miner overtime if it drops below a certain hashrate. you can set the minimum hashrate in "minhash" (80000 = 800GH/s).
ssh into the miner and save it in the /config folder. any name you want. example hashcheck. then chmod 755 hashcheck.
to run the script, enter "nohup /config/hashcheck &". this will allow the script to run after you exit from ssh
to view the log enter yourmineraddress/hashcheck.txt in your browser.
ssh into the miner and save it in the /config folder. any name you want. example hashcheck. then chmod 755 hashcheck.
to run the script, enter "nohup /config/hashcheck &". this will allow the script to run after you exit from ssh
to view the log enter yourmineraddress/hashcheck.txt in your browser.
Code:
#!/bin/sh
echo $(date) "| hashcheck monitor started" > /www/pages/hashcheck.txt
restartcount=0
minhash=80000
while true
do
var=$(cgminer-api | grep 'GHS 5s] => ' | cut -c 16- | tr -d '.')
if [ $var -lt $minhash ]; then
echo $(date) "| rechecking in 60s" >> /www/pages/hashcheck.txt
sleep 60
var=$(cgminer-api | grep 'GHS 5s] => ' | cut -c 16- | tr -d '.')
if [ $var -lt $minhash ]; then
restartcount=$((restartcount+1))
restarthash=$var
restarttime=$(date)
echo $(date) "| failed: " $restarthash >> /www/pages/hashcheck.txt
echo $(date) "| restarted cgminer, waiting for 90 seconds: " >> /www/pages/hashcheck.txt
/etc/init.d/cgminer.sh restart
sleep 90
echo $(date) "| miner restarted " $restartcount " times: " $restarthash ", at " $restarttime > /www/pages/hashcheck.txt
else
echo $(date) "| recovered: " $var >> /www/pages/hashcheck.txt
sleep 10
fi
else
echo $(date) "| normal operation: " $var >> /www/pages/hashcheck.txt
sleep 10
fi
done
echo $(date) "| hashcheck monitor started" > /www/pages/hashcheck.txt
restartcount=0
minhash=80000
while true
do
var=$(cgminer-api | grep 'GHS 5s] => ' | cut -c 16- | tr -d '.')
if [ $var -lt $minhash ]; then
echo $(date) "| rechecking in 60s" >> /www/pages/hashcheck.txt
sleep 60
var=$(cgminer-api | grep 'GHS 5s] => ' | cut -c 16- | tr -d '.')
if [ $var -lt $minhash ]; then
restartcount=$((restartcount+1))
restarthash=$var
restarttime=$(date)
echo $(date) "| failed: " $restarthash >> /www/pages/hashcheck.txt
echo $(date) "| restarted cgminer, waiting for 90 seconds: " >> /www/pages/hashcheck.txt
/etc/init.d/cgminer.sh restart
sleep 90
echo $(date) "| miner restarted " $restartcount " times: " $restarthash ", at " $restarttime > /www/pages/hashcheck.txt
else
echo $(date) "| recovered: " $var >> /www/pages/hashcheck.txt
sleep 10
fi
else
echo $(date) "| normal operation: " $var >> /www/pages/hashcheck.txt
sleep 10
fi
done
Jump to: