[Guide] Dogie's Comprehensive ASICMiner Cube Setup [HD]

[hagenuk]

0.01 BTC if someone will resolve my problem before 23.00 UTC of 02/01/2014

It's the power supply. I had the exact same problem with 3 cubes, many others here in this thread same, problem, usually with a Seasonic and also with some of the Corsair's with an "X' in the model name. Usually a rebadged Seasonic.  It would run for 1.75 seconds then die once turned on or once the fuze was inserted.

Don't waste your time openeing up(well you can check that everything is tight and the cards are in their slot) , it is not the reason for the problem you are having now. If you go back in the thread you can see there is a solution by putting some Capacitors in line with the power at the back of the cube, but it is a pain in the ass and can be fiddly.  

The two power supply's i have tested and are good to power 2 cubes at a time are the Corsair GS 800  and the thermaltake 840. Both worked fine for me. My Seasonic 1050w was a royal pain in the ass.

Changed the psu and now everything is working well! Thank you!!!

Drop me an address

Isn't it me the 1st one to give you the answer for your problem ?   No forced donation....hahahahaha.... But good to know your problem was solved.

Actually I have the same problem with some of my cube.

For me I'll just needed to make a few quick like 3-5 attempts of jump starting the cube with the paper clip(plug in and out a few time with 1seconds each) trick on my Seasonic X850W. Used to do with the capacitor but not anymore now since I have been successfully with my own trick

But not sure does it work for other PSU.

[1715163229]

1715163229

[1715163229]

1715163229

[1715163229]

1715163229

[1715163229]

1715163229

[1715163229]

1715163229

[otsaku]

Glad to see it's hashing.  Now that it's doing that but only with three blades, I've only seen this under a few circumstances:

1)  Powersupply is too weak, regardless of what the numbers say (some are sensitive to initial draw, etc)
2)  Some blades aren't seated properly and/or are running too hot - if you haven't already, open it up, tighten the heatsink screws and reseat the blades solidly.
3)  (rarely) Cube is simply "confused" and clicking on the Update button on bottom right can restart it

thanks.

Heatsinks are screwed tight, None of my blades were in the grooves of the case, two weren't seated properly. I think it might be the PSU - it's a 550w corsair, i just pulled one of the pci-e plugs that has  a removable extra two prongs and used a p6(?) > pci-e adapter. Now have five out of six blades working, none of them running all chips or at full speed. Still only running 5ghs. I'm going to try with two psu's - unless of course someone tells me that is a bad idea...

[jdot007]

Anyone have any experience with these power supplies & the cubes?

http://www.amazon.com/dp/B00CGYCNG2/ref=pe_385040_30332190_TE_M3T1_ST1_dp_1

I had a melted/exploded fuse==now damaged unit on one cube, running 2 cubes on high off one powersupply. The other cube
will run on high but will just fall back to low. 

Thanks

-J

I blew my fuses as well because the ones that came with mine were 32V.  Once I purchased 12V 30A ones I was fine.  Just my two cents, hope that helps!

Thanks for the info. The fuses all say 30 on them. I only have an issue with one cube. Has anyone been able to get a replacement controller board? The way this thing looks I'll probably need one

The other cubes that I have are humming along fine (low setting)
may have to reinvest in some of those 850 corsairs to run at high though

 

[Damnsammit]

Thanks for the info. The fuses all say 30 on them. I only have an issue with one cube. Has anyone been able to get a replacement controller board? The way this thing looks I'll probably need one

The other cubes that I have are humming along fine (low setting)
may have to reinvest in some of those 850 corsairs to run at high though

They should say 30 on them.  They are 12V 30A fuses.  Contact your supplier about a controller board... they should have extras or will be able to swap one out for you and then get a replacement from friedcat.

[jdot007]

Glad to see it's hashing.  Now that it's doing that but only with three blades, I've only seen this under a few circumstances:

1)  Powersupply is too weak, regardless of what the numbers say (some are sensitive to initial draw, etc)
2)  Some blades aren't seated properly and/or are running too hot - if you haven't already, open it up, tighten the heatsink screws and reseat the blades solidly.
3)  (rarely) Cube is simply "confused" and clicking on the Update button on bottom right can restart it

thanks.

Heatsinks are screwed tight, None of my blades were in the grooves of the case, two weren't seated properly. I think it might be the PSU - it's a 550w corsair, i just pulled one of the pci-e plugs that has  a removable extra two prongs and used a p6(?) > pci-e adapter. Now have five out of six blades working, none of them running all chips or at full speed. Still only running 5ghs. I'm going to try with two psu's - unless of course someone tells me that is a bad idea...

Well I'm finding that the cubes are finicky with power supplies and I know for a fact that they're finicky if you have android phones or tablets ect on the same network. Be sure not to overlook that when troubleshooting.

[quetzalito]

Hi again guys!

I posted more than a week ago that my cube worked for almost 9 hours "High" on a corsair TX750 V1, and then found it dead on the morning with the "clicking sound" and immediate shutdown when I tried to  turn it on.
Glad to tell you that I finally received the replacement Corsair RM750, and it turned on normally, and all the chips are "o". Will be hashing on low for stability for the time being and let you know if anything goes wrong.

Also I just read a few posts back someone said that the board is faulty if you remove all the boards and it still doesn't turn on, but I can say this is false because before I bought new the PSU I tested that, and I still got the immediate "clicking sound".

So if you are having problems change your PSU. And if you just received your cube, make sure to open it and check all the screws on the heatsinks.

I tested:

Working:
Corsair RM750

Failed:
Seasonic X650 - Immediate clicking sound
Corsair TX750 V1 - failed after 9 hours, then immediate clicking sound

[alexrossi]

0.01 BTC if someone will resolve my problem before 23.00 UTC of 02/01/2014

I have a big problem: i'm trying to fire up a cube, i have a coolermaster v700 (58A on the 12v rail), with the fuse the miner start and after a few second it shutdown, without the fuse the miner starts but the leds are all off and also if i try to navigate on 192.168.1.254:8000 the browser doesn't load anything...

If the cube shutdown within 1-2sec after powering up, your PSU probable got overloaded by the surge in the current draw. Try with a different PSU if you have.

It's true

[stewdk]

Just wanted to share my experience with fan replacements.  I put some Cougar Vortex HDB fans (http://www.newegg.com/Product/Product.aspx?Item=N82E16835553005) in my cubes, and they sure do run nice and silent but at the expense of pushing a lot less air than the stock fans. The cubes do run on high with these fans (I ran them for over a day like that) but it gets way too hot for my liking.  I touched a thermometer to one of the heatsinks and it read over 90°C (ambient is ~21°C).  So I put the original fans back in and they're running much cooler now.

afaik the network interface on blades and cubes is 10 MB.  some network equipment may not deal with this properly.  if it's a managed switch, try setting the port config to 10MB.  however, a good switch should autosense to 10 MB.  I do not know if they are half or full duplex.

My router reports that they are 10mbps half duplex, if that matters to anyone.  I have two cubes running on a proxy on an Ubuntu machine, and system monitor shows that the bandwidth doesn't usually goes above 40 kB/s (320 kbps).

I know for a fact that they're finicky if you have android phones or tablets ect on the same network.

I think that a possible way to mitigate this if you have an old router laying around is to put your cubes on their own subnet.  Plug the cubes into the LAN ports of your old router, and use the WAN port of the old router to connect to your existing network. You have to configure the router so its LAN is actually a different IP subnet - for example if your existing network is 192.168.1.X, then the cube subnet has to be something else, like 192.168.2.X.

And for good measure, here are my PSU datapoints:
Rosewill Xtreme RX750-S-B 750W: fail - overcurrent protection kicks in after a minute or so with a single cube on low
EVGA NEX750G Gold 750W: success - runs two cubes on high, though the fan is on max
EVGA 600B 600W: success - runs a single cube on high, nice and silent, but the PCIe power cables get hot

Perhaps dogie would be interested in keeping a table in the OP about which PSUs have been reported working / not working, and how many cubes the PSU can power?

[Damnsammit]

Anyone hook a voltameter up to their cube yet to see how much the PSU is drawing from the wall?

For some reason, I suspect the 200-230W rating from ASIC Miner is bullshit.  I'm probably wrong though, just not sure why so many PSUs are having issues with the Cube.

[ksami]

Hi, I'm new & got one Cube.

I have a Corsair RM850 PSU... The PSU is starting & working but I have an issue : If I set the same IP for the proxy it works for a few times & stuck and keep restarting after 2 minutes. I figured it out, so I have launched 2 separated proxy but I have a lot of rejected packets. Do you know why

I have also checked the screws of heat sinks of all the blades... All are OK !

It seems that if I switch Clock from Low to High, it reboot the Cube with X X X X X X in all ASIC chips... The temperature is OK by the way... Do you know why ?

Here some screenshots :

http://img15.hostingpics.net/pics/164151581.png

http://img15.hostingpics.net/pics/478079602.png


Thank you !

[babaji.ca]

0.01 BTC if someone will resolve my problem before 23.00 UTC of 02/01/2014

It's the power supply. I had the exact same problem with 3 cubes, many others here in this thread same, problem, usually with a Seasonic and also with some of the Corsair's with an "X' in the model name. Usually a rebadged Seasonic.  It would run for 1.75 seconds then die once turned on or once the fuze was inserted.

Don't waste your time openeing up(well you can check that everything is tight and the cards are in their slot) , it is not the reason for the problem you are having now. If you go back in the thread you can see there is a solution by putting some Capacitors in line with the power at the back of the cube, but it is a pain in the ass and can be fiddly.  

The two power supply's i have tested and are good to power 2 cubes at a time are the Corsair GS 800  and the thermaltake 840. Both worked fine for me. My Seasonic 1050w was a royal pain in the ass.

Changed the psu and now everything is working well! Thank you!!!

Drop me an address

Glad You got it working, but it was actually hagenuk who solved it for you in post # 662 . I was just emphasizing it because it seemed like you missed it.

[twib2]

Hi, I'm new & got one Cube.

I have a Corsair RM850 PSU... The PSU is starting & working but I have an issue : If I set the same IP for the proxy it works for a few times & stuck and keep restarting after 2 minutes. I figured it out, so I have launched 2 separated proxy but I have a lot of rejected packets. Do you know why

I have also checked the screws of heat sinks of all the blades... All are OK !

It seems that if I switch Clock from Low to High, it reboot the Cube with X X X X X X in all ASIC chips... The temperature is OK by the way... Do you know why ?

Sometimes the wrong 850 is still not good enough.  I have an 850 that just can't cut the mustard and it's a good name.  OTOH, I have 2 Dell poweredge PSUs at 750, each running 2 Cubes.  Meh.

One other consideration is the network - I've had some situations whereby a particular machine was being nasty on the network and killing my blade/cube/miner performance.  Just for kicks, you might want to ensure there are no other routers, computers, homenet adapters, wifi APs, etc, on your subnet and see if it still has the same problem.  If it does, it's probably power supply.  If it doesn't keep happening, keep adding devices until performance trashes and you'll have found the culprit.  In my case, it was one wireless access point running DD-WRT that was somehow spamming the network in a way I never noticed before.  I removed it and voila', max hashing again.

As another poster mentioned, these things are very sensitive to wireless denizens, especially android phones and also QoS settings (you might also try turning off all QoS in any wireless devices on your subnet.

Good luck!

[ksami]

Sometimes the wrong 850 is still not good enough.  I have an 850 that just can't cut the mustard and it's a good name.  OTOH, I have 2 Dell poweredge PSUs at 750, each running 2 Cubes.  Meh.

One other consideration is the network - I've had some situations whereby a particular machine was being nasty on the network and killing my blade/cube/miner performance.  Just for kicks, you might want to ensure there are no other routers, computers, homenet adapters, wifi APs, etc, on your subnet and see if it still has the same problem.  If it does, it's probably power supply.  If it doesn't keep happening, keep adding devices until performance trashes and you'll have found the culprit.  In my case, it was one wireless access point running DD-WRT that was somehow spamming the network in a way I never noticed before.  I removed it and voila', max hashing again.

As another poster mentioned, these things are very sensitive to wireless denizens, especially android phones and also QoS settings (you might also try turning off all QoS in any wireless devices on your subnet.

Good luck!

Thank you, I will give a shot because I have a lot of network devices and virtual servers & 2 routers connected

[otsaku]

ASIC_01-16: O O O O O O O O O O O O O O O O
ASIC_17-32: O O O O O O O O O O O O O O O O
ASIC_33-48: O O O O O O O O O O O O O O O O
ASIC_49-64: O O O O O O O O O O O O O O O O
ASIC_65-80: O O O O O O O O O O O O O O O O
ASIC_81-96: O O O O O O O O O O O O O O O O
M_01-16: 329 209 284 254 449 464 434 239 254 209 299 344 344 254 329 374
M_17-32: 254 299 254 464 254 449 284 344 314 404 239 314 269 614 344 119
M_33-48: 494 164 344 284 269 404 224 389 404 269 284 194 179 344 284 209
M_49-64: 494 284 344 329 164 299 299 194 179 164 194 374 194 359 314 329
M_65-80: 224 329 314 239 299 179 314 254 344 194 149 179 209 119 089 209
M_81-96: 059 044 134 074 059 059 029 029 014 000 000 000 044 000 000 000
Jobs:0000001654 Accepted:0000001588 Rejected:0000000005 (0:2) F1 F2 F3
MHS:23798 Utility:325 Efficieny:096.00%


Still not running perfectly. Not too sure what the solution was:

Rebooted the router.
Unplugged my Jalapenos.
Turned off wifi on my Ipad.
Turned off wifi on my galaxy note 3.
Turned off mac mini.
Rebooted RPi.

Ordered a new 750w PSU so we will see if that grabs the last few ghs.

Thanks for the help.

[Damnsammit]

ASIC_01-16: O O O O O O O O O O O O O O O O
ASIC_17-32: O O O O O O O O O O O O O O O O
ASIC_33-48: O O O O O O O O O O O O O O O O
ASIC_49-64: O O O O O O O O O O O O O O O O
ASIC_65-80: O O O O O O O O O O O O O O O O
ASIC_81-96: O O O O O O O O O O O O O O O O
M_01-16: 329 209 284 254 449 464 434 239 254 209 299 344 344 254 329 374
M_17-32: 254 299 254 464 254 449 284 344 314 404 239 314 269 614 344 119
M_33-48: 494 164 344 284 269 404 224 389 404 269 284 194 179 344 284 209
M_49-64: 494 284 344 329 164 299 299 194 179 164 194 374 194 359 314 329
M_65-80: 224 329 314 239 299 179 314 254 344 194 149 179 209 119 089 209
M_81-96: 059 044 134 074 059 059 029 029 014 000 000 000 044 000 000 000
Jobs:0000001654 Accepted:0000001588 Rejected:0000000005 (0:2) F1 F2 F3
MHS:23798 Utility:325 Efficieny:096.00%

Yeah, it still looks like a PSU issue... you don't have any blades running completely efficient.  Although that 614 looks quite sexy!

Here's mine right now:

ASIC_01-16: O O O O O O O O O O O O O O O O
ASIC_17-32: O O O O O O O O O O O O O O O O
ASIC_33-48: O O O O O O O O O O O O O O O O
ASIC_49-64: O O O O O O O O O O O O O O O O
ASIC_65-80: O O O O O O O O O O O O O O O O
ASIC_81-96: O O O O O O O O O O O O O O O O
M_01-16: 386 389 393 391 389 390 385 393 398 395 392 395 392 396 395 399
M_17-32: 392 400 394 396 402 395 394 396 401 400 397 399 404 400 400 401
M_33-48: 403 404 402 403 402 407 404 401 403 399 399 406 401 410 408 400
M_49-64: 401 411 401 407 405 405 402 399 412 407 405 401 410 409 405 407
M_65-80: 405 407 406 407 403 400 403 402 405 403 403 402 400 402 403 397
M_81-96: 398 400 403 394 401 395 396 397 396 390 390 393 387 394 386 385
Jobs:0004071359 Accepted:0003958288 Rejected:0000000910 (5:4) F1 F2 F3
MHS:38385 Utility:525 Efficieny:097.22%

[Donkezz]

Hello please i have problem i want setup my cube miner but speed is very slov only 50-60 MH/s

https://i.imgur.com/Ys2Jh6h.jpg?1

https://i.imgur.com/Ys2Jh6h.jpg?1

I have atx power 550W.

I tried bitcoin.cz, btcguild, givemecoins etc.

Thanks

[stex2009]

Hello please i have problem i want setup my cube miner but speed is very slov only 50-60 MH/s
I have atx power 550W.
I tried bitcoin.cz, btcguild, givemecoins etc.
Thanks

This is the slowest cube ever recorded here!

[Apteryx]

I have atx power 550W.

I tried bitcoin.cz, btcguild, givemecoins etc.

Thanks

I would suspect that there is an issue with boards not being seated on this one

[alexrossi]

To hagenuk: https://blockchain.info/it/tx/1c6f1fe4b9b8d2cc75bd8905f4efad3f4d88bf754387b585b3404a0b5d4bd1cf

[Donkezz]

Please anyone can help me? I buy new asicminer cube but not work good. Very slow speed or nothing.

https://i.imgur.com/ESGD52p.jpg

I install teamviewer 9 if anyone want help please write me on email donkezz@gmail.com i give teamviewer ID and pass.

Thanks