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

[cromsps]

Hello,

Yes, both are... The one to the PI its from a nice brand and the one to the miner its a 1600W Bitmain and I already tested with a 2200W from another miner.

On one 741 I have

Code:

ECHU[0 128 0 0], 

on the other I have,

Code:

ECHU[0 512 0 0]

But it could not be Temp cause it happens on both the same thing,
and for test I put

Code:

--avalon7-voltage-level 6

and LIMIT FAN MINIMUM at 4000RPMs, so temp is around 21 and still reboots

[HagssFIN]

Have you opened a support ticket at www.canaan.io ?

I have not seen a 741 behaving like this before.

[cromsps]

Have you opened a support ticket at www.canaan.io ?

I have not seen a 741 behaving like this before.

Hello,

No, cause I bought them used from cryptouniverse(?)...
In this thread, theres another one complaining with the same things... I changed the PI power supply and does the same thing.

@Hagss now you're going to think that I'm crazy that this problem its from the Lan lol... I've seen people complaining about the same thing, but they dont stop mining so no one cares about it, if we have a lot of them.

But with Static IP and with Google DNS it can take about 20 to 30 minutes whitout restarting, with DHCP, it takes only 4 to 6 minutes... I think, when it losees Lan for some reason, it it disconnects from the miners for a second or two, do know if this is possible?... But now I can only test it on the weekend with other Router... Also dont know why it needs IPV4 BroadCast on the network But I am going to test it....

Well after testing the images on other 2 places it worked... For some how, one of my routers of one of my ISPs does not work with all the canaan PI images except with the nicehash image.

But still keeps restarting on Nicehash. On Kano, Slush and other pools, it stays working for 24h without restarting the CgMiner.
I think thats because the nicehash server address is redirecting to the server that the nicehash client asks at th moment. So it works 5, 6 or 7 minutes on one client, then stops (the 741 comes down, restarts cgminer) and connects to another nicehash Client.

If this happens to somenone, the solution for now (and for me), is mining on a "real" pool :-) it works with nicehash, but its always restarting and we will keep listening the miner boots up after one work :-)  but the shares are there and its counting :-)

If someone knows how to solve this without compiling the OpenWrt, with a "keep connection alive", or whatever, please tell me :-)

[Steamtyme]

Thanks for the update. The only other time I remember someone having an issue like this they didn't come back with an answer. Glad everything is up and running.

My theory was that it had to do with Nicehash switching orders or something. It was just a guess as I've never used NH.

[cromsps]

It's me who should thank you for all your help :-)

I think your theory is right. The M3 and the S9 dont do this, but on the Pools page they only have Server, Worker and Pass. On the OpenWrt beside the pools we have the Fan Speed, Voltage and other settings, so if we apply a change only on the server (NH switching Job) it has to restart the system cause it doesnt know if we change voltage or other settings. just my thoughts :-)

Edit: the most strange thing, its the ECHU giving codes on NH after restart, and on other pools gives all Zero

[spooderman]

Hi all,

I have a set up consisting of 3 avalon miners. A 721 and two 741s.

I have not been able to get any mining done whatsoever. The Avalon controller was bought from blokforge and I have been able to connect to it, set up a pool and I have changed the network protocol to DHCP for easier access.

Following the guide (and I've read just about every post in here) has not helped. I have several AUC3 converters and two different AUC3 I2C 5PIN Cables. I have tried them in multiple configurations to no avail.

The miner powers up fine. The controller loads up fine. I can open up the interface fine. But no workers show up in the settings. "Avalon devices" remains empty.

I am not sure where the error is, I have multiple redundancies for everything, including miners, though only one controller. This is a legit avalon controller bought from blokforge flashed for 7 series.

Does anyone have any suggestions?

Regards,

Spooderman

[HagssFIN]

Please try to reflash the Raspberry Pi and make a setup with one miner first.

Use the A741 openwrt firmware for the Raspberry Pi.


When you get it running, add the second miner.

[spooderman]

Is there specific firmware for 741?

Is this the right place?

https://canaan.io/downloads/software/avalon7/openwrt/20161229/

[HagssFIN]

Is there specific firmware for 741?

Is this the right place?

https://canaan.io/downloads/software/avalon7/openwrt/20161229/

Please go visit here, chapter 4: https://bitcointalk.org/index.php?topic=2588105

There are links and everything.

[spooderman]

Hi,

I have followed this guide as stated. I have 2 of literally everything and both fail in the same way, so I assume that I am doing something wrong.

Any help would be appreciated. (Miners are simply not showing up in the interface. Currently in the debug I have "Socket connect failed => Connection refused")

Kind regards

edit: as is often the case, things are somehow working now. i flashed the firmware on the MM myself (which carries a hefty warning against doing so - not sure why, very simple to do).

if someone else can't get theirs working i wouldn't know what to tell them.

[HagssFIN]

Have you tried using another micro USB cable between the AUC3 dongle and Raspberry Pi?

Make sure that it is for both the power and data.

Is there any useful information in the CGMiner API Log?

[spooderman]

As edited above, it now works. I think flashing the MM with new firmware is what finally got me there.

Buying the avalon controller was a waste of money, I ended up using my old RPi and installing the software manually.

Oh well, guess I have a spare ($110) rpi.

I'm happy to be mining now nonetheless. Thanks for the suggestions.

[HagssFIN]

Okay,
great to hear that it is working now.

[kano]

As edited above, it now works. I think flashing the MM with new firmware is what finally got me there.

Buying the avalon controller was a waste of money, I ended up using my old RPi and installing the software manually.

Oh well, guess I have a spare ($110) rpi.

I'm happy to be mining now nonetheless. Thanks for the suggestions.

The Avalon controller also comes with:

1 X SD Card
4 X AUC3
5 X Micro USB Cable

... so it wasn't a $110 RPi3 ... that's also now $100

[s_falahati]

You can easily troubleshoot if it is the AUC3 which is faulty, if you have a ten miners or more in your setup.

six miners or more?

[HagssFIN]

six miners or more?

You're right. Six or more miners, and you'll have at least two AUCs in your setup.

Thanks.

Sorry guys, I wonder what I've been smoking when I wrote that part.

[s_falahati]

I think you should add this to the main post:

Never buy unshielded USB cables for connecting AUCs to the controller. Keep a multimeter around and check if the resistance of shielded USB cable you want to use is lower than 1ohm. If possible, check all connections, power should be around 0.1ohm, data (two pins in the middle) should be lower than or around 0.2ohm. Unfortunately, there is no way to find out if the cable is actually shielded or not unless you want to cut the actual cable and see the inside. A lot of USB cables out there just short the shield to the ground to make it impossible to know if it is actually shielded or not. Almost ALL cheap USB cables are not shielded properly.

Wasted a week trying to find out the reason behind random disconnects as I used two 3meter USB cable from Anker (yeah!!) to connect two of AUCs to the controller.

[HagssFIN]

@s_falahati
Thank you very much, I have updated the opening post.

[cromsps]

If someone knows how to solve this without compiling the OpenWrt, with a "keep connection alive", or whatever, please tell me :-)

Hello,

Well you won't believe how did I solve this :-) ... If I tried to add #xnsub to the the server, it wont work, it refuses connection,
so I test it on the Shell the miners did not restarted... After a while I started from UI and everything works normaly without restarting.
And then, I could Also add the #xnsub to the pool... Since than no more restarts on my Avalons 741... I dont know if running from the shell, it updates or installs something needed for the first time... But after this I added 3 more PIs and and runing from the Shell one time, than restart to GUI and put the XNSUB and they were ok.

Now after these months I got 841s,  ... Same thing, restarting on nicehash, I tought I could do the same thing that I did it on the 741s, but it does not work... Keeps Restarting if I start it from the GUI, If I start it from Shell, it will work normaly without restarts. but if I put it #xnsub, it wont connect.

Anyone know how to add #xnsub to the Avalon 841? or to update the cgminer on it ??

Thanks.

[PassThePopcorn]

I've used the 851's on nicehash without a real issue (hashrates all seemed normal), I didn't set xnsub because that wouldn't work but they seemed to work fine without it. I also selected a specific price to mine for and would highly suggest putting in the difficulty parameter when using more than one avalon on a pi.
Most pools will allow for parameters to be passed in the password field.
d: difficulty (ie d=14000)
p: price(ie p=0.030) this is a nicehash specific parameter.