Swedish ASIC miner company kncminer.com

[aasl]

it was "screen -r cgminer" on my Saturn

username is root, password by default is "admin", and your password configured in web admin page will replace the default password

Thanks, I'm in. Now I don't know how to get the stats up.

@user27

yes. since there is only one background process running, it is okay to use just "screen -r" to show the cgminer stats. If you are new with screen command, be careful and never use "ctrl-C" to exit, as it will terminate the mining instance. always use "ctrl-A D" to detach the screen

[1714203709]

1714203709

[user27]

it was "screen -r cgminer" on my Saturn

That has done it, thanks.

Now I get to see 2.9TH and a screen full of "disabled for 300 seconds due to repeated hardware errors". Way more disabled cores than accepted shares.

Is it recommended to update to f/w 1.0? If so, how do I do it? (I have the bin file)

[aasl]

it was "screen -r cgminer" on my Saturn

That has done it, thanks.

Now I get to see 2.9TH and a screen full of "disabled for 300 seconds due to repeated hardware errors". Way more disabled cores than accepted shares.

Is it recommended to update to f/w 1.0? If so, how do I do it? (I have the bin file)

try the "upgrade" tab on the knc web control page.

[user27]

try the "upgrade" tab on the knc web control page.

Tried it, the thing sits there and does nothing.

I just get "Sending request to 192.168.1....." and it carrys on mining.

One die is totally dead and this one is not looking much better:



EDIT:

Dead one:

[aasl]

try the "upgrade" tab on the knc web control page.

Tried it, the thing sits there and does nothing.

I just get "Sending request to 192.168.1....." and it carrys on mining.

One die is totally dead and this one is not looking much better:



EDIT:

Dead one:

it takes time to upload the 17MB file. you need to stay on the upgrade page and wait. after upload, it will upgrade the system, and then get restarted automatically. during the process, do not switch to other tabs. or the upgrade will be terminated.

If after upgrading you still have die issues, you can power off the PSU and power on again. my neppy sometimes also has die issues. using web page to restart the miner does not work, but a hard restart solves the problem.

[user27]

it takes time to upload the 17MB file. you need to stay on the upgrade page and wait. after upload, it will upgrade the system, and then get restarted automatically. during the process, do not switch to other tabs. or the upgrade will be terminated.

If after upgrading you still have die issues, you can power off the PSU and power on again. my neppy sometimes also has die issues. using web page to restart the miner does not work, but a hard restart solves the problem.

It just times out after a while/

Same from different browsers and different devices.

2nd die on first chip has not worked at any point; totally screwed I think.

[tolip_wen]

It just times out after a while/

Same from different browsers and different devices.

2nd die on first chip has not worked at any point; totally screwed I think.

Have you tried a ribbon cable from a working cube on that one?

YMMV

[user27]

It just times out after a while/

Same from different browsers and different devices.

2nd die on first chip has not worked at any point; totally screwed I think.

Have you tried a ribbon cable from a working cube on that one?

YMMV

The other three dies on that cube work for the most part so I think cable issues are unlikely to be the cause; I have tried swapping that cube with another one i.e. different PCI-e cable, different PSU, different ribbon, different beagle socket/

I managed to get the firmware uploaded; must have tried 8 or 9 times that timed out and then it worked for no apparent reason. Now I have no advanced tab and additional errors:

[raskul]

you got something broken.

try the "upgrade" tab on the knc web control page.

Tried it, the thing sits there and does nothing.

I just get "Sending request to 192.168.1....." and it carrys on mining.

One die is totally dead and this one is not looking much better:



EDIT:

Dead one:

[user27]

you got something broken.

Thanks for the insight bud; I sure did.

Plus they will have gone home for the weekend now.

[Tigggger]

Lots of errors are usually accompanied by the affected die showing it's not running with 0v 0a on the advanced page.

I've had that a few times as I have been tweaking the voltages and speeds to get the best out of mine.

When it happens a 'reboot' won't usually fix it, so I would give the affected die more volts and/or drop the speed, apply the settings then physically power off the miner, most times it then comes back to life

[raskul]

you got something broken.

Thanks for the insight bud; I sure did.

Plus they will have gone home for the weekend now.

gutted
how many TH you was? 3TH/s? wots ur wallet address?

[proclivity]

It just times out after a while/

Same from different browsers and different devices.

2nd die on first chip has not worked at any point; totally screwed I think.

Have you tried a ribbon cable from a working cube on that one?

YMMV

The other three dies on that cube work for the most part so I think cable issues are unlikely to be the cause; I have tried swapping that cube with another one i.e. different PCI-e cable, different PSU, different ribbon, different beagle socket/

I managed to get the firmware uploaded; must have tried 8 or 9 times that timed out and then it worked for no apparent reason. Now I have no advanced tab and additional errors:

1.1.x is ASIC 2 die 2, increase the V on that die. As Tiggger said that's an indication that the die/VRMs have turned off.

If you can't bring up the advanced tab try clearing browser cache. Also Elenelen reports it can take some time to load so give it a few minutes. Worst case, use putty to login and edit the VRM config file manually then reboot.

[user27]

Also Elenelen reports it can take some time to load so give it a few minutes. Worst case, use putty to login and edit the VRM config file manually then reboot.
If you can't bring up the advanced tab try clearing browser cache.

I managed to get the advanced page; I think you have to initiate it by clicking that tab but it never wants to load the first time after restart. So to get advanced tab just flick to it and then to something else before going back to advanced tab.

Lots of errors are usually accompanied by the affected die showing it's not running with 0v 0a on the advanced page.

Yes, 0A and 0v for those two DC/DC.

I've had that a few times as I have been tweaking the voltages and speeds to get the best out of mine.

I can't even get that one running at 50Mhz.

When it happens a 'reboot' won't usually fix it, so I would give the affected die more volts and/or drop the speed, apply the settings then physically power off the miner, most times it then comes back to life

Thanks for the tip, but no joy.

how many TH you was? 3TH/s?

2.9... KnC still haven't delivered a 3TH machine or a refund.

1.1.x is ASIC 2 die 2, increase the V on that die. As Tiggger said that's an indication that the die/VRMs have turned off.

What is safe voltage to increase to?

u27

[proclivity]

What is safe voltage to increase to?

u27

I have gone all the way up on a few dies for a few minutes. Usually this did not encourage them much. What did work for most dies is power cycling the cube while the controller card runs. There are more steps in Kurt's troubleshooting post that involve resetting the miner. I had to do this about 20-25 times in the first week.

If you reset 25 times and it still doesn't work then you'll be ready to RMA that cube. I had to do this with two of my cubes and a controller card. again with insufficient packaging. I don't know who would think a single cardboard strip would survive transport. Cubes were all over the place in one box, their metal shell bent on some.

[crashoveride54902]

What is safe voltage to increase to?

u27

I have gone all the way up on a few dies for a few minutes. Usually this did not encourage them much. What did work for most dies is power cycling the cube while the controller card runs. There are more steps in Kurt's troubleshooting post that involve resetting the miner. I had to do this about 20-25 times in the first week.

If you reset 25 times and it still doesn't work then you'll be ready to RMA that cube. I had to do this with two of my cubes and a controller card. again with insufficient packaging. I don't know who would think a single cardboard strip would survive transport. Cubes were all over the place in one box, their metal shell bent on some.

I'd put money on they did this on purpose i mean really...do they want all those neppys raising diff. rate so they make less on their DC? i don't think so, I'm just surprised they actually sent out working units...well for the most part...I mean they already designed to fail as fast as possible pretty much with running everything as close to max as they could...surprised they didn't put little jugs of gasoline in there for you guys lol

[tolip_wen]

The other three dies on that cube work for the most part so I think cable issues are unlikely to be the cause; I have tried swapping that cube with another one i.e. different PCI-e cable, different PSU, different ribbon, different beagle socket/

I managed to get the firmware uploaded; must have tried 8 or 9 times that timed out and then it worked for no apparent reason. Now I have no advanced tab and additional errors:

Might one die be overheating before you even have a chance to look

Did you visually inspect the ASIC/heatsink mating?
Mis-installation is possible, piece of grit got in there?.
The aluminum plate needs inspected too the VRMs at the end of the "E" need cooled.

If you have similar FW upload issue in the future, possibly run controller without cubes.
It might be too busy freakin out about something to pay attention.

YMMV

[The Avenger]

I'd put money on they did this on purpose i mean really...do they want all those neppys raising diff. rate so they make less on their DC? i don't think so, I'm just surprised they actually sent out working units...well for the most part...I mean they already designed to fail as fast as possible pretty much with running everything as close to max as they could...surprised they didn't put little jugs of gasoline in there for you guys lol

lol +1

edit
A man pointing a plastic gun is telling you not to buy from KNC!
http://www.maxkeiser.com/2014/07/psa-do-not-buy-the-3-ths-knc-box-sustainability-of-small-scale-bitcoin-mining/comment-page-1/

[aasl]

true! my 850w PSU does not shut down, but the PCI-e cables become very hot (especially near the PSU connector). I am worried that such huge current will make the cable/connector melted.

I just come across an idea: many graphic cards have a 6pin-to-double-molex cable.


If we have two molex cable connecting to the PSU, and then connect to this cable, will it solve the problem?

anyone is using a 6pin-to-molex adaptor to distribute the current to prevent overheat? does it work?

[dropt]

anyone is using a 6pin-to-molex adaptor to distribute the current to prevent overheat? does it work?

There already stands concern with drawing 400W of power over 3 wires, I would hazard a guess that trying to pull that type of power over 2 wires is not very smart.