|
Raptor2213
|
|
January 22, 2014, 07:39:17 AM |
|
Don't they use 5v though, not 12v?
|
Did something I say help you out? BTC - 18oTipf66z8dbwTgRCiPjbdPmqEP7zuCFb
|
|
|
HammerHedd
|
|
January 22, 2014, 01:12:03 PM |
|
Hi guys, long time reader first time poster i am having issues with running 1 antminer with my 3 BE's, it seems i can either run my antminer (@2gh/s) or if i run all 4 i get 1.4Gh/s i admit i am really new to this and copied fractalbc's config file(from page 13), but i still cannot get it to work right and i am at a loss i am using cgminer, and win7 i am really struggling, any help would be appreciated, many thanks in advance Try using bfgminer 3.10.0. Install that silicon labs driver first though. Also make sure you uninstall all the zadig bullshit. It should run with the following batch file just fine: bfgminer -o stratum+tcp://stratum.mining.eligius.st:3334 -u 19gu3jok7czATPb5TR8EMMuQu8qrQ5e34B_2 -p x --no-opencl-binaries -S antminer:all --set-device antminer:clock=x0981 pause An explanation for the above code, as I hate it when someone posts something, but doesn't explain why they did something some way: 'bfgminer -o stratum+tcp://stratum.mining.eligius.st:3334' starts bfgminer and connects it to the eligius mining pool. '-u 19gu3jok7czATPb5TR8EMMuQu8qrQ5e34B_2' is the username, with eligius, your payout address is your username. '-p x' is the password. On eligius anything works for a password. '--no-opencl-binaries' keeps it from trying to mine with my GPU. '-S antminer:all' tells bfgminer to scan for the antminers upon startup, otherwise you would have to hit "m""+""all" to add them. '--set-device antminer:clock=x0981' sets the clockspeed of the hashing chip, using some kind of internal lookup table. x0981 sets it to 2GHs, the clockspeed of which is derived by dividing your hashrate by the number of cores on the chip (unknown.) Just had to say, I really appreciate the CLI breakdown! Even after replacing the drivers, tho, bfgminer isn't detecting them... they do work with cgminer, but it's really unstable (Antminer version, something like 3.8.5)
|
DRK: XepkHLT2MYTXSFDc2muiGeA9eRzG6ytpSy P2Pool: stratum+tcp://darkcoin.kicks-ass.net:7903 BTC: 1LVE3pFpAhSrHbiK5hAUWDeVrB5UrPXRkJ http://darkcoin.kicks-ass.net
|
|
|
doctormandrake
|
|
January 22, 2014, 02:54:40 PM |
|
Hi guys, long time reader first time poster i am having issues with running 1 antminer with my 3 BE's, it seems i can either run my antminer (@2gh/s) or if i run all 4 i get 1.4Gh/s i admit i am really new to this and copied fractalbc's config file(from page 13), but i still cannot get it to work right and i am at a loss i am using cgminer, and win7 i am really struggling, any help would be appreciated, many thanks in advance Try using bfgminer 3.10.0. Install that silicon labs driver first though. Also make sure you uninstall all the zadig bullshit. It should run with the following batch file just fine: bfgminer -o stratum+tcp://stratum.mining.eligius.st:3334 -u 19gu3jok7czATPb5TR8EMMuQu8qrQ5e34B_2 -p x --no-opencl-binaries -S antminer:all --set-device antminer:clock=x0981 pause An explanation for the above code, as I hate it when someone posts something, but doesn't explain why they did something some way: 'bfgminer -o stratum+tcp://stratum.mining.eligius.st:3334' starts bfgminer and connects it to the eligius mining pool. '-u 19gu3jok7czATPb5TR8EMMuQu8qrQ5e34B_2' is the username, with eligius, your payout address is your username. '-p x' is the password. On eligius anything works for a password. '--no-opencl-binaries' keeps it from trying to mine with my GPU. '-S antminer:all' tells bfgminer to scan for the antminers upon startup, otherwise you would have to hit "m""+""all" to add them. '--set-device antminer:clock=x0981' sets the clockspeed of the hashing chip, using some kind of internal lookup table. x0981 sets it to 2GHs, the clockspeed of which is derived by dividing your hashrate by the number of cores on the chip (unknown.) Just had to say, I really appreciate the CLI breakdown! Even after replacing the drivers, tho, bfgminer isn't detecting them... they do work with cgminer, but it's really unstable (Antminer version, something like 3.8.5) If your ants are running on cgminer then they do not have the right driver for bfgminer (I prefer cgminer usually, but bfgminer is just a far better experience with these miners, no errors, steady hashes etc). Uninstall the win/*.* zadig driver from every ant, from device manager, then restart the computer and plug them back in, the silab driver should autoinstall if you have it on your computer. if not make sure to download the silab driver, there are links in this thread I believe. Then run one of the .bat or .cmd lines in here that is appropriate for your configuration and it should work. I used the 3.8.5 cgminer for way too long myself, it is much better on bfg, just the error rates being 0% is, but the compatability with other miners is nice too. I had trouble switching at first, but it was well worth it. If you are dedicated to cgminer, get one of the ones with 3.9.0 and the ant commands added, there are links in here also. But i think consensus is bfg ftw for ants.
|
|
|
|
sase007
|
|
January 22, 2014, 03:15:44 PM |
|
Hi, What usb hub is best for 4 ants and one medium fan? Can you send me some links? Thanks...
|
|
|
|
|
Walking Glitch
Sr. Member
Offline
Activity: 252
Merit: 250
Amateur Professional
|
|
January 22, 2014, 09:24:29 PM |
|
Hi guys, long time reader first time poster i am having issues with running 1 antminer with my 3 BE's, it seems i can either run my antminer (@2gh/s) or if i run all 4 i get 1.4Gh/s i admit i am really new to this and copied fractalbc's config file(from page 13), but i still cannot get it to work right and i am at a loss i am using cgminer, and win7 i am really struggling, any help would be appreciated, many thanks in advance Try using bfgminer 3.10.0. Install that silicon labs driver first though. Also make sure you uninstall all the zadig bullshit. It should run with the following batch file just fine: bfgminer -o stratum+tcp://stratum.mining.eligius.st:3334 -u 19gu3jok7czATPb5TR8EMMuQu8qrQ5e34B_2 -p x --no-opencl-binaries -S antminer:all --set-device antminer:clock=x0981 pause An explanation for the above code, as I hate it when someone posts something, but doesn't explain why they did something some way: 'bfgminer -o stratum+tcp://stratum.mining.eligius.st:3334' starts bfgminer and connects it to the eligius mining pool. '-u 19gu3jok7czATPb5TR8EMMuQu8qrQ5e34B_2' is the username, with eligius, your payout address is your username. '-p x' is the password. On eligius anything works for a password. '--no-opencl-binaries' keeps it from trying to mine with my GPU. '-S antminer:all' tells bfgminer to scan for the antminers upon startup, otherwise you would have to hit "m""+""all" to add them. '--set-device antminer:clock=x0981' sets the clockspeed of the hashing chip, using some kind of internal lookup table. x0981 sets it to 2GHs, the clockspeed of which is derived by dividing your hashrate by the number of cores on the chip (unknown.) Just had to say, I really appreciate the CLI breakdown! Even after replacing the drivers, tho, bfgminer isn't detecting them... they do work with cgminer, but it's really unstable (Antminer version, something like 3.8.5) If your ants are running on cgminer then they do not have the right driver for bfgminer (I prefer cgminer usually, but bfgminer is just a far better experience with these miners, no errors, steady hashes etc). Uninstall the win/*.* zadig driver from every ant, from device manager, then restart the computer and plug them back in, the silab driver should autoinstall if you have it on your computer. if not make sure to download the silab driver, there are links in this thread I believe. Then run one of the .bat or .cmd lines in here that is appropriate for your configuration and it should work. I used the 3.8.5 cgminer for way too long myself, it is much better on bfg, just the error rates being 0% is, but the compatability with other miners is nice too. I had trouble switching at first, but it was well worth it. If you are dedicated to cgminer, get one of the ones with 3.9.0 and the ant commands added, there are links in here also. But i think consensus is bfg ftw for ants. This is the problem, the zadig drivers interfere with the silicon labs driver.
|
|
|
|
Mudbankkeith
|
|
January 22, 2014, 09:25:53 PM |
|
|
BTc donations welcome:- 13c2KuzWCaWFTXF171Zn1HrKhMYARPKv97
|
|
|
Walking Glitch
Sr. Member
Offline
Activity: 252
Merit: 250
Amateur Professional
|
|
January 22, 2014, 09:37:03 PM Last edit: January 22, 2014, 10:30:17 PM by Walking Glitch |
|
I have a friend who ran 21 BEs on a raspi using a couple of the other Rosewill 10 port hubs. (Same adapter, and same internally more or less, just different dimensions.) They worked fine, 1 powering the pi, and the other one also powering a fan. https://bitcointalk.org/index.php?topic=378388.0
|
|
|
|
Jay_Pal
Legendary
Offline
Activity: 1493
Merit: 1003
|
|
January 22, 2014, 11:43:35 PM |
|
|
|
|
|
|
HammerHedd
|
|
January 23, 2014, 01:40:54 AM |
|
Hi guys, long time reader first time poster i am having issues with running 1 antminer with my 3 BE's, it seems i can either run my antminer (@2gh/s) or if i run all 4 i get 1.4Gh/s i admit i am really new to this and copied fractalbc's config file(from page 13), but i still cannot get it to work right and i am at a loss i am using cgminer, and win7 i am really struggling, any help would be appreciated, many thanks in advance Try using bfgminer 3.10.0. Install that silicon labs driver first though. Also make sure you uninstall all the zadig bullshit. It should run with the following batch file just fine: bfgminer -o stratum+tcp://stratum.mining.eligius.st:3334 -u 19gu3jok7czATPb5TR8EMMuQu8qrQ5e34B_2 -p x --no-opencl-binaries -S antminer:all --set-device antminer:clock=x0981 pause An explanation for the above code, as I hate it when someone posts something, but doesn't explain why they did something some way: 'bfgminer -o stratum+tcp://stratum.mining.eligius.st:3334' starts bfgminer and connects it to the eligius mining pool. '-u 19gu3jok7czATPb5TR8EMMuQu8qrQ5e34B_2' is the username, with eligius, your payout address is your username. '-p x' is the password. On eligius anything works for a password. '--no-opencl-binaries' keeps it from trying to mine with my GPU. '-S antminer:all' tells bfgminer to scan for the antminers upon startup, otherwise you would have to hit "m""+""all" to add them. '--set-device antminer:clock=x0981' sets the clockspeed of the hashing chip, using some kind of internal lookup table. x0981 sets it to 2GHs, the clockspeed of which is derived by dividing your hashrate by the number of cores on the chip (unknown.) Just had to say, I really appreciate the CLI breakdown! Even after replacing the drivers, tho, bfgminer isn't detecting them... they do work with cgminer, but it's really unstable (Antminer version, something like 3.8.5) If your ants are running on cgminer then they do not have the right driver for bfgminer (I prefer cgminer usually, but bfgminer is just a far better experience with these miners, no errors, steady hashes etc). Uninstall the win/*.* zadig driver from every ant, from device manager, then restart the computer and plug them back in, the silab driver should autoinstall if you have it on your computer. if not make sure to download the silab driver, there are links in this thread I believe. Then run one of the .bat or .cmd lines in here that is appropriate for your configuration and it should work. I used the 3.8.5 cgminer for way too long myself, it is much better on bfg, just the error rates being 0% is, but the compatability with other miners is nice too. I had trouble switching at first, but it was well worth it. If you are dedicated to cgminer, get one of the ones with 3.9.0 and the ant commands added, there are links in here also. But i think consensus is bfg ftw for ants. This is the problem, the zadig drivers interfere with the silicon labs driver. Yup, you are correct - they work fine with the CP210x drivers. Also, for the edification of the group, I found it works best to unplug and replug (i.e. restart) the antminers each time you change the frequency if you want to experiment with OCing
|
DRK: XepkHLT2MYTXSFDc2muiGeA9eRzG6ytpSy P2Pool: stratum+tcp://darkcoin.kicks-ass.net:7903 BTC: 1LVE3pFpAhSrHbiK5hAUWDeVrB5UrPXRkJ http://darkcoin.kicks-ass.net
|
|
|
Blackbird0
|
|
January 23, 2014, 02:15:37 AM |
|
Has anyone been able to OC them successfully past 2.06?
I run the 0981 argument, and it's fine. I run 0A81, and I get lots of HW errors that make the effective rate below 2.0. I'm cooling with a fan.
I have heatsinks I can adhese to the U1's, and I could even try swapping out resistors, but I'm curious if there's an easier way and/or if someone's tried it first.
I'm using the standard BFGMiner, btw.
|
|
|
|
pgminer01
Newbie
Offline
Activity: 31
Merit: 0
|
|
January 23, 2014, 02:17:58 AM |
|
I ran 2.2 gh , but with lot of hw error , 1 drop to 1.4 / 1.5 gh , nopoint to oc it
|
|
|
|
Walking Glitch
Sr. Member
Offline
Activity: 252
Merit: 250
Amateur Professional
|
|
January 23, 2014, 02:21:07 AM |
|
They're not stable above 2GHs (250MHz) without more voltage. You need to change the resistors to raise the voltage, but that requires better cooling. The solder mask on these conducts heat better than the Block eruptors from what I can tell, so the heatsink does more on these, but it would still be beneficial to remove it. I am going to find a circuit board to practice removing it on first, then once I get my heatsinks in for these, I will remove the mask, and adhere them to the heatsinks, and see how far I can take mine.
|
|
|
|
Blackbird0
|
|
January 23, 2014, 02:35:28 AM |
|
They're not stable above 2GHs (250MHz) without more voltage. You need to change the resistors to raise the voltage, but that requires better cooling. The solder mask on these conducts heat better than the Block eruptors from what I can tell, so the heatsink does more on these, but it would still be beneficial to remove it. I am going to find a circuit board to practice removing it on first, then once I get my heatsinks in for these, I will remove the mask, and adhere them to the heatsinks, and see how far I can take mine.
I see people advertising on ebay sticking heatsinks to the *front* where the chips are. Think that does anything or are some ebay folks selling snake oil?
|
|
|
|
Walking Glitch
Sr. Member
Offline
Activity: 252
Merit: 250
Amateur Professional
|
|
January 23, 2014, 02:43:12 AM |
|
They're not stable above 2GHs (250MHz) without more voltage. You need to change the resistors to raise the voltage, but that requires better cooling. The solder mask on these conducts heat better than the Block eruptors from what I can tell, so the heatsink does more on these, but it would still be beneficial to remove it. I am going to find a circuit board to practice removing it on first, then once I get my heatsinks in for these, I will remove the mask, and adhere them to the heatsinks, and see how far I can take mine.
I see people advertising on ebay sticking heatsinks to the *front* where the chips are. Think that does anything or are some ebay folks selling snake oil? They're selling Dr. Joe Wallach's miracle dirt. These chips are flip-chip BGA packages, and likely there is an air gap between the chip itself and the top of the plastic package itself (read almost guaranteed), which would act as an insulator, so without the heatsink cooling the PCB itself off (the PCB itself is an aluminum or copper substrate which has the layers of electronics and wired "printed" onto it, and so acts as a heatsink) it would actually get hotter. The only time a heatsink on top of the chip would be useful is if you are trying to push it to its utmost limits, as it would draw off a SMALL amount of heat. EDIT: An exception to this rule is the VREG, which would benefit from a ram chip heatsink on top of it if you are pushing it past specification.
|
|
|
|
dmpotter
|
|
January 23, 2014, 02:51:59 AM |
|
They're not stable above 2GHs (250MHz) without more voltage. You need to change the resistors to raise the voltage, but that requires better cooling. The solder mask on these conducts heat better than the Block eruptors from what I can tell, so the heatsink does more on these, but it would still be beneficial to remove it. I am going to find a circuit board to practice removing it on first, then once I get my heatsinks in for these, I will remove the mask, and adhere them to the heatsinks, and see how far I can take mine.
I see people advertising on ebay sticking heatsinks to the *front* where the chips are. Think that does anything or are some ebay folks selling snake oil? They're selling Dr. Joe Wallach's miracle dirt. These chips are flip-chip BGA packages, and likely there is an air gap between the chip itself and the top of the plastic package itself (read almost guaranteed), which would act as an insulator, so without the heatsink cooling the PCB itself off (the PCB itself is an aluminum or copper substrate which has the layers of electronics and wired "printed" onto it, and so acts as a heatsink) it would actually get hotter. The only time a heatsink on top of the chip would be useful is if you are trying to push it to its utmost limits, as it would draw off a SMALL amount of heat. EDIT: An exception to this rule is the VREG, which would benefit from a ram chip heatsink on top of it if you are pushing it past specification. Thank you, I always felt the same way, but you put it so much nicer.
|
|
|
|
Blackbird0
|
|
January 23, 2014, 03:00:46 AM |
|
They're not stable above 2GHs (250MHz) without more voltage. You need to change the resistors to raise the voltage, but that requires better cooling. The solder mask on these conducts heat better than the Block eruptors from what I can tell, so the heatsink does more on these, but it would still be beneficial to remove it. I am going to find a circuit board to practice removing it on first, then once I get my heatsinks in for these, I will remove the mask, and adhere them to the heatsinks, and see how far I can take mine.
I see people advertising on ebay sticking heatsinks to the *front* where the chips are. Think that does anything or are some ebay folks selling snake oil? They're selling Dr. Joe Wallach's miracle dirt. These chips are flip-chip BGA packages, and likely there is an air gap between the chip itself and the top of the plastic package itself (read almost guaranteed), which would act as an insulator, so without the heatsink cooling the PCB itself off (the PCB itself is an aluminum or copper substrate which has the layers of electronics and wired "printed" onto it, and so acts as a heatsink) it would actually get hotter. The only time a heatsink on top of the chip would be useful is if you are trying to push it to its utmost limits, as it would draw off a SMALL amount of heat. EDIT: An exception to this rule is the VREG, which would benefit from a ram chip heatsink on top of it if you are pushing it past specification. Thank you. So, if I'm getting this right: unless I'm screwing with resistors, which I am ludicrously unqualified to do, I'm going to be only seeing a successful overclock to 2.0 GH/s, which is with the 0981 command. There's nothing else I can do that would otherwise allow me to overclock beyond that.
|
|
|
|
Walking Glitch
Sr. Member
Offline
Activity: 252
Merit: 250
Amateur Professional
|
|
January 23, 2014, 03:17:19 AM |
|
Thank you, I always felt the same way, but you put it so much nicer.
You're welcome, and thank you. So, if I'm getting this right: unless I'm screwing with resistors, which I am ludicrously unqualified to do, I'm going to be only seeing a successful overclock to 2.0 GH/s, which is with the 0981 command. There's nothing else I can do that would otherwise allow me to overclock beyond that.
Yes. Most of these chips will not do 2.2GHs (275MHz) at stock voltage which is 0.8V. Some will do it, however, they will throw hardware errors, yet still submit over 2GHs (according to the third hashrate column in bfgminer) but like I said, most will throw too many hardware errors to make it worth it. Mine fall into the latter category, and will not run well at 2.2GHs. Keep in mind the nanofuries, which are based on bitfury chips throw around 5-10 percent hardware errors, which is due to the design of the bitfury chips, just like the asicminer chips. The antminer chips however do not exhibit this behavior unless they are voltage-starved, or possibly overheated - however I haven't overheated mine, so it's hard to say. The only way to squeeze more hashrate from these is with more voltage, and therefor better cooling. You would probably be fine with the stock heatsinks at 0.9V and maybe 1 Volt, as long as you have good active cooling in a room temperature environment. (Personally I think 1V is pushing it without a better heatsink with the thermal mask being removed.) However, I would guess that you would be able to beat the nanofury hashrate of 2.2GHs at 0.9V with most chips, without hardware errors, which the nanofuries will have. Also I may be mistaken, but I think the nanofuries and the like also operate at 0.9V, while these operate at 0.8V.
|
|
|
|
|