bitmine CoinCraft Series Users Thread - Troubleshooting, Efficiency, O/C - page 3.

Soros Shorts

donator

Activity: 1617

Merit: 1012

Quote from: invain on August 15, 2014, 02:05:56 AM

Hello, thank you for this nice information base.
Two weeks ago I owned a pre-owned CoinCraft with 5 modules and an ANTEC 1000W PSU.
I decided to upgrade the PSU to an LEPA 1600W.
After connecting it and switching it on, the PSU knocked two times. I unpluged all connections except the mainboard connector and switched it on again...
I saw smoke and realized, that the connectoin on the PBC-Board from PIN 24 to C18 to C16 burned!
Since that its not possible to start the unit, neither with the original ANTEC PSU.

.
Now I will try to fix the PBC board, but I don't know the function of the capacitor and the resistor.
So please help me with this questions:
Could this be an issue with the PSU-ON-Pin? Is there a difference between the ANTEC and the LEAP PSU?
Is it possible to buy a new PBC from anywhere if I can't fix it?
What did I made wrong?

Many thanks for your help!

The LEPA 1600 is a multi-rail PSU with six 12V rails (4 x 30A and 2 x 20A). However only four 12V 30A rails are accessible as PCI-e. Plus you need to pick the correct connectors to make sure you don't plug 2 PCI-e cables into the same rail. The Desk needs five PCI-e cables, so for the fifth cable you need to make a custom PCI-e cable connected to one of the remaining 12V rails, which by the way are only 20A rails.

Chances are you did something wrong while doing this.

Wolke

legendary

Activity: 966

Merit: 1000

Quote from: invain on August 15, 2014, 02:05:56 AM

Hello, thank you for this nice information base.
Two weeks ago I owned a pre-owned CoinCraft with 5 modules and an ANTEC 1000W PSU.
I decided to upgrade the PSU to an LEPA 1600W.
After connecting it and switching it on, the PSU knocked two times. I unpluged all connections except the mainboard connector and switched it on again...
I saw smoke and realized, that the connectoin on the PBC-Board from PIN 24 to C18 to C16 burned!
Since that its not possible to start the unit, neither with the original ANTEC PSU.

.
Now I will try to fix the PBC board, but I don't know the function of the capacitor and the resistor.
So please help me with this questions:
Could this be an issue with the PSU-ON-Pin? Is there a difference between the ANTEC and the LEAP PSU?
Is it possible to buy a new PBC from anywhere if I can't fix it?
What did I made wrong?

Many thanks for your help!

i think you killed the pcb or the miner -> never, realy never use the modular cables from an other psu !!

use only the cables from the same psu !!!!! the "pinbelegung" is from manufacturer to manufacturer different Sad

invain

newbie

Activity: 5

Merit: 0

Hello, thank you for this nice information base.
Two weeks ago I owned a pre-owned CoinCraft with 5 modules and an ANTEC 1000W PSU.
I decided to upgrade the PSU to an LEPA 1600W.
After connecting it and switching it on, the PSU knocked two times. I unpluged all connections except the mainboard connector and switched it on again...
I saw smoke and realized, that the connectoin on the PBC-Board from PIN 24 to C18 to C16 burned!
Since that its not possible to start the unit, neither with the original ANTEC PSU.
http://i57.tinypic.com/2s7866c.jpg.
Now I will try to fix the PBC board, but I don't know the function of the capacitor and the resistor.
So please help me with this questions:
Could this be an issue with the PSU-ON-Pin? Is there a difference between the ANTEC and the LEAP PSU?
Is it possible to buy a new PBC from anywhere if I can't fix it?
What did I made wrong?

Many thanks for your help!

yslyung

legendary

Activity: 1500

Merit: 1002

Mine Mine Mine

look carefully, there's a newer one me thinks

glendower

newbie

Activity: 17

Merit: 0

Quote from: yslyung on August 11, 2014, 01:29:25 PM

pull the sdcard out n update FW via desktop/laptop. could be an old image. try that then you may magically get "extra menus"

When I look at the Update Firmware setting page, it says the current version is 201406101155.

Looks like a later version than Bitmine offer for download...

yslyung

legendary

Activity: 1500

Merit: 1002

Mine Mine Mine

pull the sdcard out n update FW via desktop/laptop. could be an old image. try that then you may magically get "extra menus"

Wolke

legendary

Activity: 966

Merit: 1000

Quote from: glendower on August 11, 2014, 12:05:43 PM

Quote from: Wolke on August 11, 2014, 11:35:49 AM

Quote from: glendower on August 11, 2014, 11:18:10 AM

I've just received two Coincraft Desks, both with 8 modules.

I've just been looking to see what can be done to overclock these, but two things are holding me back so far:

1. The temperature gauge on the LCD panels of both units reads 0.0C. Obviously wrong... any idea why for both of these units there would be no temperature data being displayed?

2. There is no option to change the power mode. On the mining page is just says: "Device speed is factory calibrated to the best possible setting.".

Any suggestions folks? Huh

Thanks.

you mean rigs and not desk right ?

Well, funny you should ask that.

I ordered desks... but handles on front, couple of little brackets in the box. Look more like I would expect rigs to look.

I'm being unintentionally dense. They've physically got 4 modules each, so ergo are Desks not Rigs.

Its just because as I was typing, the dashboards were up in front of me, which report 8 modules, even though it appears that physically there are 4 cards in each.

i think that mean -> you've got 4 modules per desk -> with 8 chips per module -> 32 chips each miner -> per chip 31,25 gh/s -> its like "turbo mode"

i have the "old" desks -> with 40 chips each miner.... -> per chip 25 gh/s in normal mode....

it can be possible that this miners are not oc compatible -> but i'm not sure

glendower

newbie

Activity: 17

Merit: 0

Quote from: Wolke on August 11, 2014, 11:35:49 AM

Quote from: glendower on August 11, 2014, 11:18:10 AM

I've just received two Coincraft Desks, both with 8 modules.

I've just been looking to see what can be done to overclock these, but two things are holding me back so far:

1. The temperature gauge on the LCD panels of both units reads 0.0C. Obviously wrong... any idea why for both of these units there would be no temperature data being displayed?

2. There is no option to change the power mode. On the mining page is just says: "Device speed is factory calibrated to the best possible setting.".

Any suggestions folks? Huh

Thanks.

you mean rigs and not desk right ?

Well, funny you should ask that.

I ordered desks... but handles on front, couple of little brackets in the box. Look more like I would expect rigs to look.

I'm being unintentionally dense. They've physically got 4 modules each, so ergo are Desks not Rigs.

Its just because as I was typing, the dashboards were up in front of me, which report 8 modules, even though it appears that physically there are 4 cards in each.

Wolke

legendary

Activity: 966

Merit: 1000

Quote from: glendower on August 11, 2014, 11:18:10 AM

I've just received two Coincraft Desks, both with 8 modules.

I've just been looking to see what can be done to overclock these, but two things are holding me back so far:

1. The temperature gauge on the LCD panels of both units reads 0.0C. Obviously wrong... any idea why for both of these units there would be no temperature data being displayed?

2. There is no option to change the power mode. On the mining page is just says: "Device speed is factory calibrated to the best possible setting.".

Any suggestions folks? Huh

Thanks.

you mean rigs and not desk right ?

glendower

newbie

Activity: 17

Merit: 0

I've just received two Coincraft Desks, both with 8 modules.

I've just been looking to see what can be done to overclock these, but two things are holding me back so far:

1. The temperature gauge on the LCD panels of both units reads 0.0C. Obviously wrong... any idea why for both of these units there would be no temperature data being displayed?

2. There is no option to change the power mode. On the mining page is just says: "Device speed is factory calibrated to the best possible setting.".

Any suggestions folks? Huh

Thanks.

DjangoSi

newbie

Activity: 44

Merit: 0

If I am upgrading the SD Card to the new Firmware I get this error:

root@raspberrypi:~# cgminer -T
[2014-04-25 17:31:13] Started cgminer 4.3.0
[2014-04-25 17:31:13] Failed to write to fdesc 4: Input/output error
[2014-04-25 17:31:13] No CoinCrafd Desk backplane detected.
[2014-04-25 17:31:13] Failed to write to fdesc 5: Input/output error
[2014-04-25 17:31:13] All devices disabled, cannot mine!

Any idea?

Zelek Uther

hero member

Activity: 700

Merit: 504

Run a Bitcoin node.

Quote from: nox_ on July 17, 2014, 10:55:31 AM

Quote from: Zelek Uther on July 17, 2014, 12:58:01 AM

Quote from: Zelek Uther on July 10, 2014, 05:56:51 AM

Quote from: Shaolino on July 10, 2014, 05:21:35 AM

Hello, I have the same problem as Zelek Uther mentioned here before. The miner starts hashing and then slowly dies after a few minutes.
After rebooting I can start it again with the same result....after a few minutes it just stops hashing.
Any ideas ?

I still have this issue on one miner. I've tried different SDCards, different firmware (including Debian Desk from a forum user), I have even swapped out the RaspPi. Still have the same issue where cgminer fails with a segmentation fault. I see this by invoking cgminer from the command line.

What I'm trialling now is running just one module at a time. I unplugged the PCIe cable from every module except the first one, and it ran fine on one module for 24 hours. Power down, try the same with just module 2... and so on. I run each for at least 24 hours as a test. Still working through this. Then I'll try two modules... then three etc until I can run stable with as many as possible.

I've run each module by itself, then combinations of two, then three, and finally have it hashing stable on four modules.

So the result of my troubleshooting is that I have to leave one module unplugged (PCIe cable disconnected).

What PSU are you running for the five modules?

I bought three 1 TH/s units (each had five modules) and originally I had three broken backplane boards which were replaced. I then had the issue of four boards running at full speed with one at 25 - 29 GH/s.. This was fixed by replacing the stock 1000W PSU with a 1250W PSU then putting the miner into 'turbo moderate' - This kicked the final module into life, cleared my segmentation faults and I am hashing at 1 - 1.1 TH/s on two of the units.

Still need to solder on a capacitor to one of the replacement miner boards I received (had to ship two back for repairs) as they arrived with broken heatsinks :-(

I've got 4x 1TH/s Desks (5 modules) each has the stock Antec 1000W PSU.

I could try a higher W PSU to see if I can get that last module stable, thanks for the idea. Trouble is I'd have to buy one, I have none spare at the moment.

nox_

newbie

Activity: 51

Merit: 0

Quote from: Zelek Uther on July 17, 2014, 12:58:01 AM

Quote from: Zelek Uther on July 10, 2014, 05:56:51 AM

Quote from: Shaolino on July 10, 2014, 05:21:35 AM

Hello, I have the same problem as Zelek Uther mentioned here before. The miner starts hashing and then slowly dies after a few minutes.
After rebooting I can start it again with the same result....after a few minutes it just stops hashing.
Any ideas ?

I still have this issue on one miner. I've tried different SDCards, different firmware (including Debian Desk from a forum user), I have even swapped out the RaspPi. Still have the same issue where cgminer fails with a segmentation fault. I see this by invoking cgminer from the command line.

What I'm trialling now is running just one module at a time. I unplugged the PCIe cable from every module except the first one, and it ran fine on one module for 24 hours. Power down, try the same with just module 2... and so on. I run each for at least 24 hours as a test. Still working through this. Then I'll try two modules... then three etc until I can run stable with as many as possible.

I've run each module by itself, then combinations of two, then three, and finally have it hashing stable on four modules.

So the result of my troubleshooting is that I have to leave one module unplugged (PCIe cable disconnected).

What PSU are you running for the five modules?

I bought three 1 TH/s units (each had five modules) and originally I had three broken backplane boards which were replaced. I then had the issue of four boards running at full speed with one at 25 - 29 GH/s.. This was fixed by replacing the stock 1000W PSU with a 1250W PSU then putting the miner into 'turbo moderate' - This kicked the final module into life, cleared my segmentation faults and I am hashing at 1 - 1.1 TH/s on two of the units.

Still need to solder on a capacitor to one of the replacement miner boards I received (had to ship two back for repairs) as they arrived with broken heatsinks :-(

Zelek Uther

hero member

Activity: 700

Merit: 504

Run a Bitcoin node.

Quote from: Zelek Uther on July 10, 2014, 05:56:51 AM

Quote from: Shaolino on July 10, 2014, 05:21:35 AM

Hello, I have the same problem as Zelek Uther mentioned here before. The miner starts hashing and then slowly dies after a few minutes.
After rebooting I can start it again with the same result....after a few minutes it just stops hashing.
Any ideas ?

I still have this issue on one miner. I've tried different SDCards, different firmware (including Debian Desk from a forum user), I have even swapped out the RaspPi. Still have the same issue where cgminer fails with a segmentation fault. I see this by invoking cgminer from the command line.

What I'm trialling now is running just one module at a time. I unplugged the PCIe cable from every module except the first one, and it ran fine on one module for 24 hours. Power down, try the same with just module 2... and so on. I run each for at least 24 hours as a test. Still working through this. Then I'll try two modules... then three etc until I can run stable with as many as possible.

I've run each module by itself, then combinations of two, then three, and finally have it hashing stable on four modules.

So the result of my troubleshooting is that I have to leave one module unplugged (PCIe cable disconnected).

Shaolino

newbie

Activity: 20

Merit: 0

Quote from: Zelek Uther on July 10, 2014, 05:56:51 AM

Quote from: Shaolino on July 10, 2014, 05:21:35 AM

Hello, I have the same problem as Zelek Uther mentioned here before. The miner starts hashing and then slowly dies after a few minutes.
After rebooting I can start it again with the same result....after a few minutes it just stops hashing.
Any ideas ?

I still have this issue on one miner. I've tried different SDCards, different firmware (including Debian Desk from a forum user), I have even swapped out the RaspPi. Still have the same issue where cgminer fails with a segmentation fault. I see this by invoking cgminer from the command line.

What I'm trialling now is running just one module at a time. I unplugged the PCIe cable from every module except the first one, and it ran fine on one module for 24 hours. Power down, try the same with just module 2... and so on. I run each for at least 24 hours as a test. Still working through this. Then I'll try two modules... then three etc until I can run stable with as many as possible.

Interesting, thanks for your information, so it makes no sense for me trying different SDcards etc.....
my miners were delivered end of June, could it be that a whole batch has this problem ?

jimmy3dita

hero member

Activity: 588

Merit: 500

Just a quick note to say that Bitmine has promptly sent me a new PSU (it took just 72 hours, using UPS), so right now I'm back at full speed.

Zelek Uther

hero member

Activity: 700

Merit: 504

Run a Bitcoin node.

Quote from: Shaolino on July 10, 2014, 05:21:35 AM

Hello, I have the same problem as Zelek Uther mentioned here before. The miner starts hashing and then slowly dies after a few minutes.
After rebooting I can start it again with the same result....after a few minutes it just stops hashing.
Any ideas ?

I still have this issue on one miner. I've tried different SDCards, different firmware (including Debian Desk from a forum user), I have even swapped out the RaspPi. Still have the same issue where cgminer fails with a segmentation fault. I see this by invoking cgminer from the command line.

What I'm trialling now is running just one module at a time. I unplugged the PCIe cable from every module except the first one, and it ran fine on one module for 24 hours. Power down, try the same with just module 2... and so on. I run each for at least 24 hours as a test. Still working through this. Then I'll try two modules... then three etc until I can run stable with as many as possible.

Shaolino

newbie

Activity: 20

Merit: 0

Hello, I have the same problem as Zelek Uther mentioned here before. The miner starts hashing and then slowly dies after a few minutes.
After rebooting I can start it again with the same result....after a few minutes it just stops hashing.
Any ideas ?

nhminer

sr. member

Activity: 473

Merit: 250

Quote from: rmb2u on July 09, 2014, 06:48:05 AM

Hi all,

I have received my miner, but I have a strange problem.
After running about 2 minutes the miner front display stops working well, and is only showing strange signs.

Does anyone have an idea what is wrong here?

Kind regards,

RmB2u

Join the club -- common problem, ignore it -- It doesn't cause any trouble

rmb2u

newbie

Activity: 41

Merit: 0

Hi all,

I have received my miner, but I have a strange problem.
After running about 2 minutes the miner front display stops working well, and is only showing strange signs.

Does anyone have an idea what is wrong here?

Kind regards,

RmB2u

Topic: bitmine CoinCraft Series Users Thread - Troubleshooting, Efficiency, O/C - page 3. (Read 24629 times)