Troubleshooting and repair guide for AvalonMiner models 721, 741, 761, 821 & 841

[Steamtyme]

Sorry I'm not an expert at combing through the logs. I can generally figure out an issue with the Miner itself when that's the problem. In regards to what happened after the move hard to say, it could be a slightly different line voltage in your new location making it so the Pi isn't getting enough juice to support 5 miners per AUC3 - this is just a guess.

Have you tried firing it up again as a test with say 3 or 4 miners per AUC3? If not that should be your first step.
Do you have something like a Samsung phone charger wall brick?? That's what I was using for my PI's in the past - I've never had to run 5 machines per AUC3 though.

[1714308859]

1714308859

[1714308859]

1714308859

[1714308859]

1714308859

[jingil]

Thank Steamtyme ,
I was also thinking about chaining less units and see what happens as you said.
The bricks will be here by tomorrow , but im not so sure if it is rasbperi power,  as pack of 10s are doing the same thing too.
Regards

[Steamtyme]

Forgive my ignorance if I get this wrong. Your pack of 10's are still 5 miners/AUC3,
The Pi get's it's power then it branches off to the 4 USB ports to supply each potential chain. The power supplied to each chain would be proportional to what is incoming to the PI, so if that has dropped off due to your line voltage it could be a chain effect.
I believe with what you have you should try 1 PI with 12 miners - 3 per AUC to start. If that fires up do another PI with 16 miners - 4 per AUC. If the first test doesn't work try starting at 8 miners with there being 2 miners per AUC. Also just double check that all your cables are seated properly.

Just play with it until you get miners hashing consistently. Good luck.

[jingil]

Thats exactly as you said man.
I am going to test what you said but i guess somehow this will solve the problem as i had tested something like you said , powering up device by device and at some point it become totally unstable.
Im wondering what does canaan officially announced as the raspbery power requirements for up to 20 miners per  raspbery.
Also powered USB hub is an option ? how would it connect to raspbery ?
Regards

[NotFuzzyWarm]

If you mean using powered hub to power the RasPi it is generally NOT an option unless it specifically says it will supply at least 2.5A per-port. (Canaan calls for at least 2.5A rating).

If you mean using a hub between the Pi and the AUC's --no one has EVER needed to do that. The issue with running more than 3 miners per AUC is not related to power, it is related to longer communications times needed to talk to all the miners in the chain. That increased time leads to higher number of rejected shares.

For full info on all Avalon miners use the CGMiner API log. There is a ton of data there on the AUC's including data speed, voltages, temp, etc.

[philipma1957]

1) chains of 3 are better then chains of 4 or 5.

2) next what is the exact make and model of the power brick on your rasp pi?

3) here is a link to a good power brick

https://www.amazon.com/CanaKit-Raspberry-Supply-Adapter-Listed/dp/B00MARDJZ4/

which make and model  rasp pi do you have?

go back to basics

do 1 rasp pi to 1 miner  see if it works

[jingil]

If you mean using powered hub to power the RasPi it is generally NOT an option unless it specifically says it will supply at least 2.5A per-port. (Canaan calls for at least 2.5A rating).

If you mean using a hub between the Pi and the AUC's --no one has EVER needed to do that. The issue with running more than 3 miners per AUC is not related to power, it is related to longer communications times needed to talk to all the miners in the chain. That increased time leads to higher number of rejected shares.

For full info on all Avalon miners use the CGMiner API log. There is a ton of data there on the AUC's including data speed, voltages, temp, etc.

Thank you ,
Can you give me some hints at the key and values i must check in the cgminer log and where to compare  those values with correct ones ?
I can also paste the log here , if you have sometime to check .
Regards

[...]

The raspberry models are : Raspberry Pi 3 Model B Rev 1.2
Firmware  version is : OpenWrt Designated Driver 12009 / LuCI Master (git-19.094.07392-b9dea8e)
MM version is : 8211904-1ab9760
Power brick is : 5V ,2500MA for two sets of units and one 5V ,2000MA for one set.

How can i check for power surge or line voltage drop ?
I guess it should be possible in cgminer logs.
Thank you everyone for help.

[NotFuzzyWarm]

This Pi is running 7 miners, 4 on 1 AUC and 3 on a 2nd one.

1st AUC reported shows:

Code:

CRC[0 0 0 0],MM Count=4,Smart Speed=1,Connecter=AUC,AUC VER=AUC-20151208,AUC I2C Speed=400000,AUC I2C XDelay=19200,AUC Sensor=11129,AUC Temperature=43.21,Connection Overloaded=false,Voltage Level Offset=-1,Nonce Mask=24,USB Pipe=0,USB Delay=r0 0.000000 w0 0.000000,USB tmo=0 0|STATS=1,ID=AV81,Elapsed=16062,Calls=0,Wait=0.000000,Max=0.000000,Min=99999999.000000,MM ID1=Ver[8411812-d33ea90]

and the 2nd AUC reports:

Code:

CRC[0 0 0 0],MM Count=3,Smart Speed=1,Connecter=AUC,AUC VER=AUC-20151208,AUC I2C Speed=400000,AUC I2C XDelay=19200,AUC Sensor=11677,AUC Temperature=41.12,Connection Overloaded=false,Voltage Level Offset=-1,Nonce Mask=24,USB Pipe=0,USB Delay=r0 0.000000 w0 0.000000,USB tmo=0 0|STATS=2,ID=POOL0,Elapsed=16236,Calls=0,Wait=0.000000,Max=0.000000,Min=99999999.000000,Pool Calls=0,Pool Attempts=0,Pool Wait=0.000000,Pool Max=0.000000,Pool Min=99999999.000000,Pool Av=0.000000,Work Had Roll Time=false,Work Can Roll=false,Work Had Expire=false,Work Roll Time=0,Work Diff=81853.00000000,Min Diff=4098.00000000,Max Diff=81853.00000000,Min Diff Count=310,Max Diff Count=116532,Times Sent=4768,Bytes Sent=652003,Times Recv=5304,Bytes Recv=873196,Net Bytes Sent=652003,Net Bytes Recv=873196|STATS=3,ID=POOL1,Elapsed=16236,Calls=0,Wait=0.000000,Max=0.000000,Min=99999999.000000,Pool Calls=0,Pool Attempts=0,Pool Wait=0.000000,Pool Max=0.000000,Pool Min=99999999.000000,Pool Av=0.000000,Work Had Roll Time=false,Work Can Roll=false,Work Had Expire=false,Work Roll Time=0,Work Diff=0.00000000,Min Diff=0.00000000,Max Diff=0.00000000,Min Diff Count=0,Max Diff Count=0,Times Sent=2,Bytes Sent=153,Times Recv=5,Bytes Recv=1610,Net Bytes Sent=153,Net Bytes Recv=1610|STATS=4,ID=POOL2,Elapsed=16236,Calls=0,Wait=0.000000,Max=0.000000,Min=99999999.000000,Pool Calls=0,Pool Attempts=0,Pool Wait=0.000000,Pool Max=0.000000,Pool Min=99999999.000000,Pool Av=0.000000,Work Had Roll Time=false,Work Can Roll=false,Work Had Expire=false,Work Roll Time=0,Work Diff=0.00000000,Min Diff=0.00000000,Max Diff=0.00000000,Min Diff Count=0,Max Diff Count=0,Times Sent=2,Bytes Sent=153,Times Recv=5,Bytes Recv=1610,Net Bytes Sent=153,Net Bytes Recv=1610|'

Yours should read about the same regarding AUC IC2 Speed, temps, Connection Overloaded, etc.

[jingil]

Thank you,
I checked them with one the mine AUC3 which is as follow :

Code:

 CRC[0 0 0 0],MM Count=4,Smart Speed=1,Connecter=AUC,AUC VER=AUC-20151208,AUC I2C Speed=400000,AUC I2C XDelay=19200,AUC Sensor=15225,AUC Temperature=28.99,Connection Overloaded=false,Voltage Level Offset=0,Nonce Mask=24,USB Pipe=0,USB Delay=r0 0.000000 w0 0.000000,USB tmo=0 0|STATS=1,ID=AV81,Elapsed=264,Calls=0,Wait=0.000000,Max=0.000000,Min=99999999.000000,MM ID1=Ver[8211904-1ab9760]

Mostly they are the same but for one parameters mine are much higher AUC Sensor=15225 , I have no idea what this implies or means .
But  the other AUC3 contains much more log.

Code:

CRC[0 0 0 0],MM Count=5,Smart Speed=1,Connecter=AUC,AUC VER=AUC-20151208,AUC I2C Speed=400000,AUC I2C XDelay=19200,AUC Sensor=16000,AUC Temperature=26.55,Connection Overloaded=false,Voltage Level Offset=0,Nonce Mask=24,USB Pipe=0,USB Delay=r0 0.000000 w0 0.000000,USB tmo=0 0|STATS=2,ID=POOL0,Elapsed=264,Calls=0,Wait=0.000000,Max=0.000000,Min=99999999.000000,Pool Calls=0,Pool Attempts=0,Pool Wait=0.000000,Pool Max=0.000000,Pool Min=99999999.000000,Pool Av=0.000000,Work Had Roll Time=false,Work Can Roll=false,Work Had Expire=false,Work Roll Time=0,Work Diff=8192.00000000,Min Diff=1.00000000,Max Diff=8192.00000000,Min Diff Count=1,Max Diff Count=1588,Times Sent=1824,Bytes Sent=208055,Times Recv=16,Bytes Recv=8016,Net Bytes Sent=208055,Net Bytes Recv=8016|STATS=3,ID=POOL1,Elapsed=264,Calls=0,Wait=0.000000,Max=0.000000,Min=99999999.000000,Pool Calls=0,Pool Attempts=0,Pool Wait=0.000000,Pool Max=0.000000,Pool Min=99999999.000000,Pool Av=0.000000,Work Had Roll Time=false,Work Can Roll=false,Work Had Expire=false,Work Roll Time=0,Work Diff=0.00000000,Min Diff=0.00000000,Max Diff=0.00000000,Min Diff Count=0,Max Diff Count=0,Times Sent=2,Bytes Sent=149,Times Recv=4,Bytes Recv=1658,Net Bytes Sent=149,Net Bytes Recv=1658|STATS=4,ID=POOL2,Elapsed=264,Calls=0,Wait=0.000000,Max=0.000000,Min=99999999.000000,Pool Calls=0,Pool Attempts=0,Pool Wait=0.000000,Pool Max=0.000000,Pool Min=99999999.000000,Pool Av=0.000000,Work Had Roll Time=false,Work Can Roll=false,Work Had Expire=false,Work Roll Time=0,Work Diff=0.00000000,Min Diff=0.00000000,Max Diff=0.00000000,Min Diff Count=0,Max Diff Count=0,Times Sent=2,Bytes Sent=149,Times Recv=4,Bytes Recv=1658,Net Bytes Sent=149,Net Bytes Recv=1658|

These AUCs logs are from the pack with 2 AUCs and 5 avalons each .


Well Well Well,
So i guess i found what was so f...king wrong.
I was overlooking something very simple (may be simple but not documented any where).
Im mining at slush pool main address  stratum+tcp://stratum.slushpool.com:3333 and everything seemed fine (nearly).
After moving miner to another location and the connect / disconnect story started while the miners where working i could see it is all stale hashes and the pool seems dropping all my hashrates.
I overlooked this and never thought this is the issue , i thought because miners are not stable this is happening.
Well actually it was vise versa totally.
For some reason the  stratum+tcp://stratum.slushpool.com:3333 was counting all my hashes as stall . going around searching for what causes these stall hash rate i found that there is kind of logic in the pools software that when they found to many stall hashes from client they try to somehow communicate with client and inform it about this , either this or stale hash buildup on avalons seems to cause them to restart .
Changing to stratum+tcp://cn.stratum.slushpool.com:3333 seems to solve the stale hashes.
I have got 35 of them stable for about 30 minutes now.A chain of 5 seems to still crash though which im working on .
Best regards

[Steamtyme]

Well glad you got it sorted, when you get the AUC's you might still want to think about spreading the miners out so you don't have to run chains of 5. I'll just throw this in there as it's a good time to ramp up at Kano pool. Lower fees, but it does come with higher variance. Might be worth looking into for yourself if you've never mined there.

[s_falahati]

Hey HagssFIN,

What about the thermal paste at the back of the hashing board?  Hash boards seem physically and thermally connected to the back plate as well and from the look of it; there is thermal paste involved. Should I disconnect it and change the paste at the back as well?

[HagssFIN]

If the old paste is bad,
I would clean it with isopropyl alcohol
and then re-paste it with a good quality thermal paste.

[soymoji]

Soymoji, did you solve your problem?

No i could not . i tried different versions of any firmware available for A921 none worked!

the A921 Modules shows green light but never turn blue! and  in CGminer status shows : " This section contains no values yet "

tricky!   

[HagssFIN]

@soymoji
Are you able to test it with another Rasp Pi?

Are you able to test it with another AUC3 converter?

[NotFuzzyWarm]

@soymoji
Did you also try a different USB cable? Cables that are for charging only are more common than one would think...  It needs one that is not just for power but must also support data transfer, sometimes called a 'sync' cable.

[ogombii]

Hi Guys

I have a 2 841s with issues:

1. the PMU lights are on, i suspect bad pmus, is there anyone who can do an out of warranty repair?
2. one pmu light is red one green on a single side, can i get pmus anywhere? or RMA?

[Steamtyme]

Have you reached out to Canaan to see if they have the parts available for sale? It's really easy to work on the Avalon machines so I imagine you could DIY. If not you might want to check with a distributor like Blokforge or Cryptouniverse. If they don't pan out you could post in the Marketplace you are looking for a couple, iirc they are side specific. Left or Right and marked.
I'll also look through what I have. I can't remember if I have a good spare or not. I actually can't even test it anymore either, gave away the 841. Either way I'll let you know.

[ogombii]

Thanks for the response, I'll email Canaan, to see if they can send me spares, hopefully, I'll get a reply.

[HagssFIN]

Thanks for the response, I'll email Canaan, to see if they can send me spares, hopefully, I'll get a reply.

You can also create an account (if you don't have one yet)
and submit a ticket at their website.

https://canaan.io/submit-ticket/

[ogombii]

ticket opened yesterday, waiting for a response back.  thanks.