OLD: BFGMiner 3.10.0: modular ASIC+FPGA, GBT+Strtm, RPC, Mac/Lnx/W64, AntU1, DRB - page 64.

Laxe

newbie

Activity: 14

Merit: 0

Quote from: drukoz on January 14, 2014, 06:13:13 AM

does anyone else get a problem with their ant miner u1 i get 0.5/0.7/2.0 gh/s out of my ant miner with bfg with cgminer 3.8.5 i get a steady 1.9 - 2.3 gh/s and i have the pool confirm that it seems like ant miners aren't working properly because if i use cgminer 3.100 (newest version) i get the same 700 or 800 mhs so it might be a problem with bfgminer and how it takes the miner if anyone has gotten it to work with BE's in same hub and same program (not 2 instances) with bfgminer then please pm me also i am using the bfgmienr-ant program please and thank you Cheesy

You're not the only one, I too and getting slower speeds with BFG Miner. I've made some improvement but it's a hassle to do everytime. Check out my post here: https://bitcointalksearch.org/topic/m.4503625

drukoz

member

Activity: 67

Merit: 10

Quote from: HellDiverUK on January 14, 2014, 06:04:56 AM

Quote from: retek on January 14, 2014, 05:51:03 AM

Quote from: HellDiverUK on January 14, 2014, 05:46:24 AM

No, because it's not possible unless you change a component on the board.

My erupter is atm mining emark @ 664 Mh/s: of course I did not changed any component on the board.
If you read the link above, you will get the "trick" to overclock it (changing some timing parameter).

Actually it isn't - it's just a "trick" which makes cgminer's bugged interface report a faster speed than it's actually mining.

But, hey, knock yourself out if you think you're getting more speed than is physically possible.

yea it is just a glitch but also i noticed that it does make your miners more steady on bfgminer and also if it was real then it should show in your pool whats ur hashrate before and after ??

retek

full member

Activity: 238

Merit: 100

Quote from: HellDiverUK on January 14, 2014, 06:04:56 AM

Actually it isn't - it's just a "trick" which makes cgminer's bugged interface report a faster speed than it's actually mining.

But, hey, knock yourself out if you think you're getting more speed than is physically possible.

Uhm, so is some sort of joke... I should have to do some real test with the number share =)
Thanks a lot for the real quick reply: I will come back to bfgminer after it.

drukoz

member

Activity: 67

Merit: 10

does anyone else get a problem with their ant miner u1 i get 0.5/0.7/2.0 gh/s out of my ant miner with bfg with cgminer 3.8.5 i get a steady 1.9 - 2.3 gh/s and i have the pool confirm that it seems like ant miners aren't working properly because if i use cgminer 3.100 (newest version) i get the same 700 or 800 mhs so it might be a problem with bfgminer and how it takes the miner if anyone has gotten it to work with BE's in same hub and same program (not 2 instances) with bfgminer then please pm me also i am using the bfgmienr-ant program please and thank you Cheesy

HellDiverUK

hero member

Activity: 1246

Merit: 501

Quote from: retek on January 14, 2014, 05:51:03 AM

Quote from: HellDiverUK on January 14, 2014, 05:46:24 AM

No, because it's not possible unless you change a component on the board.

My erupter is atm mining emark @ 664 Mh/s: of course I did not changed any component on the board.
If you read the link above, you will get the "trick" to overclock it (changing some timing parameter).

Actually it isn't - it's just a "trick" which makes cgminer's bugged interface report a faster speed than it's actually mining.

But, hey, knock yourself out if you think you're getting more speed than is physically possible.

retek

full member

Activity: 238

Merit: 100

Quote from: HellDiverUK on January 14, 2014, 05:46:24 AM

No, because it's not possible unless you change a component on the board.

My erupter is atm mining emark @ 664 Mh/s: of course I did not changed any component on the board.
If you read the link above, you will get the "trick" to overclock it (changing some timing parameter).

HellDiverUK

hero member

Activity: 1246

Merit: 501

Quote from: retek on January 14, 2014, 05:10:58 AM

Hi,
is there anyway to overclock Usb block erupter (icarus), like is possibile with cgminer [1]?

No, because it's not possible unless you change a component on the board.

retek

full member

Activity: 238

Merit: 100

Hi,
is there anyway to overclock Usb block erupter (icarus), like is possibile with cgminer [1]?

[1] https://bitcointalksearch.org/topic/block-eruptor-usb-overclocked-357684

Thanks for any information =)

nwoolls

hero member

Activity: 840

Merit: 1002

Quote from: chiefstank on January 13, 2014, 07:25:24 PM

just an annoyance for me, and i don't fault the programmers for this, is that it seems the detection of the devices is random so for example, COM5 may be ANT0 when you run it one time, but another time COM5 may be ANT5. seems to be no rhyme or reason.

This is likely the cause. Note that ASIC manufacturers have the option to have their devices return a unique serial number. The serial number is returned along with the path when you use the -d? argument. This can also be used with the --scan and --set-device commands to target a specific device and is more reliable as it doesn't change with COM ports etc. However the AntMiner U1 does not return a serial number so it cannot be used in this case Sad

lajz99

hero member

Activity: 756

Merit: 500

Still no Drillbit support...it's only been about 2 months now...

Mudbankkeith

hero member

Activity: 868

Merit: 1000

Quote from: chiefstank on January 13, 2014, 07:25:24 PM

Quote from: cool77 on January 13, 2014, 06:00:19 PM

Quote from: nwoolls on January 13, 2014, 01:06:52 PM

Quote from: chiefstank on January 13, 2014, 12:55:52 PM

Greetings. I too have an antminer that is throwing a high rate of errors and I'm having problems with downclocking a single antminer. When I run the device list, I see all of them listed as \\.\COM(NUM). of course this doesn't help when you have more than one as you don't know which is which. but i deduced it was COM54. when I run bfgminer with the --set-device for that path, it seems a random antminer will be downclocked. never seems to be the same one nor the problem antminer.

The U1 may support sending it an identify command so that its LED lights up. I'll have to look into it.

On my side I was able to identify the com port by removing all except one. Then, run the command, so I know which it is. It seems Windows assign a port to a specific usb port.

Thanks for the reply. As I mentioned, I was able to find out the COM number for the one that was causing all the errors and it was by doing the exact same thing. Unfortunately, when I ran the command specified to downclock the misbehaving antminer, one antminer would be downclocked and one would still be throwing errors. haven't had time to go through the entire process again to see if it's the exact same one or a different one throwing the errors.

just an annoyance for me, and i don't fault the programmers for this, is that it seems the detection of the devices is random so for example, COM5 may be ANT0 when you run it one time, but another time COM5 may be ANT5. seems to be no rhyme or reason.

BFG allocates the slots on first started, first in, so it is different each time you start.

one of my miners was throwing a lot of errors, I re arranged them into different slots, I found one of the slots on my hub was weak.(left that one empty)
moved it to another hub and now, zero errors. (not hashing any faster tho.)

chiefstank

member

Activity: 81

Merit: 10

Quote from: cool77 on January 13, 2014, 06:00:19 PM

Quote from: nwoolls on January 13, 2014, 01:06:52 PM

Quote from: chiefstank on January 13, 2014, 12:55:52 PM

Greetings. I too have an antminer that is throwing a high rate of errors and I'm having problems with downclocking a single antminer. When I run the device list, I see all of them listed as \\.\COM(NUM). of course this doesn't help when you have more than one as you don't know which is which. but i deduced it was COM54. when I run bfgminer with the --set-device for that path, it seems a random antminer will be downclocked. never seems to be the same one nor the problem antminer.

The U1 may support sending it an identify command so that its LED lights up. I'll have to look into it.

On my side I was able to identify the com port by removing all except one. Then, run the command, so I know which it is. It seems Windows assign a port to a specific usb port.

Thanks for the reply. As I mentioned, I was able to find out the COM number for the one that was causing all the errors and it was by doing the exact same thing. Unfortunately, when I ran the command specified to downclock the misbehaving antminer, one antminer would be downclocked and one would still be throwing errors. haven't had time to go through the entire process again to see if it's the exact same one or a different one throwing the errors.

just an annoyance for me, and i don't fault the programmers for this, is that it seems the detection of the devices is random so for example, COM5 may be ANT0 when you run it one time, but another time COM5 may be ANT5. seems to be no rhyme or reason.

cool77

newbie

Activity: 13

Merit: 0

Quote from: nwoolls on January 13, 2014, 01:06:52 PM

Quote from: chiefstank on January 13, 2014, 12:55:52 PM

Greetings. I too have an antminer that is throwing a high rate of errors and I'm having problems with downclocking a single antminer. When I run the device list, I see all of them listed as \\.\COM(NUM). of course this doesn't help when you have more than one as you don't know which is which. but i deduced it was COM54. when I run bfgminer with the --set-device for that path, it seems a random antminer will be downclocked. never seems to be the same one nor the problem antminer.

The U1 may support sending it an identify command so that its LED lights up. I'll have to look into it.

On my side I was able to identify the com port by removing all except one. Then, run the command, so I know which it is. It seems Windows assign a port to a specific usb port.

IYFTech

hero member

Activity: 686

Merit: 500

WANTED: Active dev to fix & re-write p2pool in C

Quote from: HellDiverUK on January 13, 2014, 02:59:59 PM

Quote from: IYFTech on January 13, 2014, 02:24:18 PM

HellDiver, I don't suppose you remember the slush startup command line for p2pool can you? I can't find an example anywhere my man. I seem to remember having to change something on the cube webface too - but my brain fade has got the better of me today......

If you do a mining_proxy.py --help that'll bring up the commands available.

I'm using screen -dmS proxy mining_proxy.py -o -p -gp -cu -cp x -q

That spawns the proxy on it's own screen, and allows your miners to connect with any username, and the proxy will submit the shares to your p2pool node with your payout address. You'll want to change the -gp port to something different to p2pool's port (say 8330).

Yeah, that's what I tried earlier, plus a number of variations too - I just can't get it to click, sure it's something stupid I'm missing. Did you make any changes to the cube webface at all?

Thanks for the ideas by the way, it's appreciated - I hate seeing a third of my work going to waste with this proxy..... Tongue

HellDiverUK

hero member

Activity: 1246

Merit: 501

Quote from: IYFTech on January 13, 2014, 02:24:18 PM

HellDiver, I don't suppose you remember the slush startup command line for p2pool can you? I can't find an example anywhere my man. I seem to remember having to change something on the cube webface too - but my brain fade has got the better of me today......

If you do a mining_proxy.py --help that'll bring up the commands available.

I'm using screen -dmS proxy mining_proxy.py -o -p -gp -cu -cp x -q

That spawns the proxy on it's own screen, and allows your miners to connect with any username, and the proxy will submit the shares to your p2pool node with your payout address. You'll want to change the -gp port to something different to p2pool's port (say 8330).

IYFTech

hero member

Activity: 686

Merit: 500

WANTED: Active dev to fix & re-write p2pool in C

I had slush proxy running previously on 12.04 & p2pool before I switched my node off - but after switching it back on again I can't for the life of me remember how I done it.... Tongue

This bfg proxy is losing me 30+%!

HellDiver, I don't suppose you remember the slush startup command line for p2pool can you? I can't find an example anywhere my man. I seem to remember having to change something on the cube webface too - but my brain fade has got the better of me today......

HellDiverUK

hero member

Activity: 1246

Merit: 501

Quote from: IYFTech on January 13, 2014, 12:36:52 PM

Well bugger me sideways, that's several hours wasted then, thanks HelldiverUK. I'm running Xubuntu 12.04 but was unable to get slush's proxy working with my cube & p2pool for some reason, it just wouldn't connect, but works fine on other pools - any pointers or ideas my man?

Peace.

I'd try Debian or Arch.

Ubuntu seems to be bugged or just broken in a lot of stuff that isn't desktop related. I never had any luck getting mining stuff working properly on Ubuntu, yet everything worked perfectly first time on Debian. I've had p2pool working on Debian, I have slush's proxy as backup to BFG running, and I run bfgminer on one Debian machine and cgminer on the other - both compiled and worked perfectly first time, and I'm a bit of a linux noob.

Neil (mineforeman) said he'd recommend Arch over Ubuntu-based OS, too.

nwoolls

hero member

Activity: 840

Merit: 1002

Quote from: chiefstank on January 13, 2014, 12:55:52 PM

Greetings. I too have an antminer that is throwing a high rate of errors and I'm having problems with downclocking a single antminer. When I run the device list, I see all of them listed as \\.\COM(NUM). of course this doesn't help when you have more than one as you don't know which is which. but i deduced it was COM54. when I run bfgminer with the --set-device for that path, it seems a random antminer will be downclocked. never seems to be the same one nor the problem antminer.

The U1 may support sending it an identify command so that its LED lights up. I'll have to look into it.

chiefstank

member

Activity: 81

Merit: 10

Quote from: nwoolls on January 12, 2014, 01:17:24 PM

Quote from: cool77 on January 12, 2014, 12:51:07 PM

Is it possible to overclock antminers separately with different oc? I have one antminer I would like to oc to 0881 and the other one to 0981?

One of my antminer has 14% HW and the other one 0% HW with 0981 oc.

Yes you can. First run bfgminer with -d? -S antminer:all to list the devices and their paths:

Code:

bfgminer -d? -S antminer:all [2014-01-12 13:16:59] Started bfgminer 3.9.0
 [2014-01-12 13:17:00] Devices detected:
 [2014-01-12 13:17:00]  Device (driver=antminer; procs=1; path=/dev/cu.SLAB_USBtoUART)
1 devices listed

Then use the path displayed (above) in the --set-device argument:

Code:

bfgminer -S antminer:all --set-device antminer@/dev/cu.SLAB_USBtoUART:clock=x0881

Greetings. I too have an antminer that is throwing a high rate of errors and I'm having problems with downclocking a single antminer. When I run the device list, I see all of them listed as \\.\COM(NUM). of course this doesn't help when you have more than one as you don't know which is which. but i deduced it was COM54. when I run bfgminer with the --set-device for that path, it seems a random antminer will be downclocked. never seems to be the same one nor the problem antminer.

any suggestions?

IYFTech

hero member

Activity: 686

Merit: 500

WANTED: Active dev to fix & re-write p2pool in C

Quote from: HellDiverUK on January 13, 2014, 12:08:25 PM

Quote from: IYFTech on January 13, 2014, 12:04:07 PM

Greetings,

I'm running a cube using the bfgminer proxy but the longpoll is showing as inactive & I'm only getting 67% efficiency as well as 34% rejects - can you tell me the command to enable longpolling as I can't find it in the readme......thanks anyone.

Peace. Grin

I had this discussion with Luke-Jr about a month ago. There is no Longpoll support, but he said he'd look in to adding it.

I found that running bfgminer on Linux helped a lot - I'm getting a solid 38GH out of my Cube connected to bfgminer on a Celeron 847 PC running Debian Wheezy.
When I point the Cube to the Pi running MinePeon and bfgminer, the Pi's CPU is running flat out and efficiency goes to crap. It runs OK on the BeagleBone Black running Debian, too (dual 1GHz ARMv7). The same Celeron machine running a stripped down Windows 7 install is also inefficient.

Well bugger me sideways, that's several hours wasted then, thanks HelldiverUK. I'm running Xubuntu 12.04 but was unable to get slush's proxy working with my cube & p2pool for some reason, it just wouldn't connect, but works fine on other pools - any pointers or ideas my man?

Peace.

Topic: OLD: BFGMiner 3.10.0: modular ASIC+FPGA, GBT+Strtm, RPC, Mac/Lnx/W64, AntU1, DRB - page 64. (Read 1193413 times)