I did some testing with comments you made.
One Antminer had hw error around 20% with cgminer and the other barely nothing.
I tried removing and replugging the one with high hw rate and it did not change.
Then, switch to a powered usb hub, same result after 12H.
This morning switch to bcgminer, up to now (for 20 minutes) I have overall 4.02 - 3.91 - 3.17GH/s HW : 2/ .23%.
It seems the OC does not work.
Here is my command line in win7 64bit :
C:\Bfgminer\bfgminer.exe --set-device antminer:freq=0981 -o http://stratum.bitcoin.cz:3333 -O User:Pass -S antminer:all
@anw : Which link did you use to get the bfgminer? And what is your command line?
Thanks!
Topic: Bitmain AntMiner U1 Tips & Tricks - page 26. (Read 106767 times)
I have had trouble with a pi, basically high usb traffic can cause the ethernet controler to drop out. I have a 20 port usb 2 hub plugged into the pi, and it will run for a while then randomly the ethernet driver starts getting errors and drops out.
This is a known issue with the pi, something to do with voltage drops. In any event i ended up using a spare old pc and throwing centos on it. USB issues gone
Ive tried doing firmware updates on the pi, and there is a kernel boot option for the nic but these make no difference.
Sounds like that could very well be the issue I'm having... The appeal to me of using the Pi is the ridiculously low power consumption, compared to triple figure wattage for even a basic desktop PC. I've got an N54L which could be another option if I get tired of messing about with my Pi...This is a known issue with the pi, something to do with voltage drops. In any event i ended up using a spare old pc and throwing centos on it. USB issues gone
Ive tried doing firmware updates on the pi, and there is a kernel boot option for the nic but these make no difference.
I might just set up another cron job on the Pi to ping Google or something like that, then reboot if it fails...
No need for hotair station, just fine soldering iron and skills 
I also noticed that i got less HW if I increase Difficulty.
I also noticed that i got less HW if I increase Difficulty.
My results after running for several hours 12 U1 @ 0981 + few erupters together with bfgminer/linux, on a 49-ports hub. Very stable now.
Thanks all for sharing experience on this topic.
@chromosoma @philipma1957 thanks for sharing your reports, I may try hardware mod but have to find hot air station
Thanks all for sharing experience on this topic.
@chromosoma @philipma1957 thanks for sharing your reports, I may try hardware mod but have to find hot air station
Code:
bfgminer version 3.9.0 - Started: [2014-01-11 00:28:09] - [ 1 day 13:23:43]
Block: ...d1f5f606 #280090 Diff:1.42G (10.15Ph/s) Started: [13:50:24]
ST:14 F:53 NB:1586 AS:0 BW:[500/ 38 B/s] E:22.46 I: 375uBTC/hr BS:1.07M
17 | 27.43/26.23/24.98Gh/s | A:33851 R:183+316(2.0%) HW:275/.03%
--------------------------------------------------------------------------------
ANT 0: | 2.09/ 2.05/ 1.93Gh/s | A: 2639 R: 18+ 22(2.8%) HW: 5/.01%
ANT 1: | 2.06/ 2.05/ 1.95Gh/s | A: 2654 R: 11+ 28(1.6%) HW: 5/.01%
ANT 2: | 2.06/ 2.05/ 1.96Gh/s | A: 2673 R: 12+ 25(1.7%) HW: 5/.01%
ANT 3: | 2.05/ 2.05/ 1.94Gh/s | A: 2569 R: 17+ 20(2.8%) HW: 6/.01%
ANT 4: | 2.05/ 2.05/ 1.93Gh/s | A: 2696 R: 17+ 26(3.0%) HW: 2/.00%
ANT 5: | 2.09/ 2.05/ 1.94Gh/s | A: 2687 R: 9+ 18(1.4%) HW: 1/.00%
ANT 6: | 2.03/ 2.05/ 1.94Gh/s | A: 2602 R: 15+ 21(2.3%) HW: 6/.01%
ANT 7: | 2.06/ 2.05/ 1.96Gh/s | A: 2587 R: 15+ 14(1.6%) HW: 3/.00%
ANT 8: | 2.06/ 2.05/ 1.96Gh/s | A: 2610 R: 12+ 20(1.4%) HW: 5/.01%
ANT 9: | 2.04/ 2.05/ 1.96Gh/s | A: 2610 R: 16+ 22(1.8%) HW: 5/.01%
ANT10: | 2.07/ 2.05/ 1.97Gh/s | A: 2665 R: 11+ 22(1.3%) HW: 3/.00%
ANT11: | 2.07/ 2.05/ 1.93Gh/s | A: 2675 R: 14+ 22(2.6%) HW: 2/.00%
BEE 0: | 341.5/333.6/326.6Mh/s | A: 448 R: 0+ 16(.75%) HW: 49/.47%
BEE 1: | 341.9/333.6/322.0Mh/s | A: 418 R: 4+ 9(3.1%) HW: 47/.45%
BEE 2: | 341.7/333.7/322.0Mh/s | A: 426 R: 4+ 11(2.9%) HW: 56/.54%
BEE 3: | 341.4/333.7/325.2Mh/s | A: 442 R: 4+ 8(2.4%) HW: 38/.36%
BEE 4: | 341.1/333.6/328.8Mh/s | A: 450 R: 4+ 12(1.5%) HW: 37/.35%
--------------------------------------------------------------------------------
I have had trouble with a pi, basically high usb traffic can cause the ethernet controler to drop out. I have a 20 port usb 2 hub plugged into the pi, and it will run for a while then randomly the ethernet driver starts getting errors and drops out.
This is a known issue with the pi, something to do with voltage drops. In any event i ended up using a spare old pc and throwing centos on it. USB issues gone
Ive tried doing firmware updates on the pi, and there is a kernel boot option for the nic but these make no difference.
This is a known issue with the pi, something to do with voltage drops. In any event i ended up using a spare old pc and throwing centos on it. USB issues gone
Ive tried doing firmware updates on the pi, and there is a kernel boot option for the nic but these make no difference.
Quote from: cs2000
Hoe did yours go, mine ran until about 4:30AM before crashing, have since restarted the pi and was running all day, still is. CGminer CPU usage was/is a lot lower still but not 100% solid yet.
Started at 11am yesterday and they stopped at roughly 8pm... I'm away from home so can't check on it yet though.
I set up a cron job to run every 5 minutes and restart the Pi if cgminer isn't running, but that doesn't seem to have kicked in so I assume it's just locked up...
I might try setting up watchdog to see if that helps, or might just ditch cgminer and the LCD and go with bfgminer instead...
What USB Hub are you guys running your u1's on? I'm currently running 8 U1's on a Rosewill but I'd like to go to 10 soon and I don't know if this hub can handle the power draw. Any suggestions?
Try and get hold of a killawatt device. Then test your power draw.
What USB Hub are you guys running your u1's on? I'm currently running 8 U1's on a Rosewill but I'd like to go to 10 soon and I don't know if this hub can handle the power draw. Any suggestions?
Small R values will cause more current through the divider to the ground=> wasted current from your usb-port.
I really doubt you will manage to solder through-hole resistors on the 0603 footprint
I really doubt you will manage to solder through-hole resistors on the 0603 footprint
Using the formula listed in the manual, can you simply leave the R1 resister and solder a 4.7k to R2 for .97v or a 3.3 for 1.04v?
Also, will the regular radioshack resisters work, or do you need to use the tiny SMD resistors?
Also, will the regular radioshack resisters work, or do you need to use the tiny SMD resistors?
Hey guys,
I've been getting intermittent errors (shown below) on CGMiner with 9 Antminers. Although I'm seeing the error message, the antminers still seem to be running. I'm running Windows 64 bit on cgminer 3.8.5. All 9 Antminers are connected to a Anker 9+1 USB hub... Any ideas on how to fix this? Any help would be greatly appreciated. Thanks!
AMU X SendWork usb write err:(-7) LIBUSB_ERROR_TIMEOUT
AMU X attempted reset got err:(0) LIBUSB_SUCCESS
AMUX: Comms error (werr=0 amt=0)
I've been getting intermittent errors (shown below) on CGMiner with 9 Antminers. Although I'm seeing the error message, the antminers still seem to be running. I'm running Windows 64 bit on cgminer 3.8.5. All 9 Antminers are connected to a Anker 9+1 USB hub... Any ideas on how to fix this? Any help would be greatly appreciated. Thanks!
AMU X SendWork usb write err:(-7) LIBUSB_ERROR_TIMEOUT
AMU X attempted reset got err:(0) LIBUSB_SUCCESS
AMUX: Comms error (werr=0 amt=0)
ok
New results from overclocking attempts:
I upgraded my 4 Anminers with 10K and 2.5K Resistors, and connected only one antminer first.
Finaly i managed to get very stable 2,4-2,5 GHS from one miner.
The overclocking procedure is not straight forward.
The antminer SilLab microcontroller seems to store some kind of calibration table , which he uses to control PWM of the DC/DC converter and adjust the voltage for TXO ( Frequency).
After you change the voltage output of the DC/DC converter those calibration values didn't match with reality, so you just cant start with high frequencies, the error is too high.
Instead start with lowers GHS possible: 1.4 GHS.
Wait until it stabilize, and increase the frequency by one value.
Step by step I managed to run Antimner at 0B81, which gave me about 2,5 GHS.
After i plugged the second calibrated miner into USB3.0 HUB (4A max), the speed of the 1st miner gone to 2.1 GHS, and the second one gave 1.9 GHS. The third miner just gone zombie...
Seems like there is not enough power? But funny enough I got the same 2.5 GHS result from USB2.0 PC port.... so current shouldn't be a problem here, probably....
I am, not quite sure WTF is wrong with AntimnerU1. Maybe some kind of connection/synchronisation errors with cgminer...
I tried to run BFG miner, but it couldnt find any device at all....even after i deinstall zadig drivers.
Moreover CGMINER cant find a device directly from USB3.0 (my USB3.0 hub is connected to the USB2.0 Port) port.
I will purchase a PCI USB2.0 extension card with separate powersupply from the main PSU. Costs only 10 Euro, so i give it a try, in case there is not enough current...
New results from overclocking attempts:
I upgraded my 4 Anminers with 10K and 2.5K Resistors, and connected only one antminer first.
Finaly i managed to get very stable 2,4-2,5 GHS from one miner.
The overclocking procedure is not straight forward.
The antminer SilLab microcontroller seems to store some kind of calibration table , which he uses to control PWM of the DC/DC converter and adjust the voltage for TXO ( Frequency).
After you change the voltage output of the DC/DC converter those calibration values didn't match with reality, so you just cant start with high frequencies, the error is too high.
Instead start with lowers GHS possible: 1.4 GHS.
Wait until it stabilize, and increase the frequency by one value.
Step by step I managed to run Antimner at 0B81, which gave me about 2,5 GHS.
After i plugged the second calibrated miner into USB3.0 HUB (4A max), the speed of the 1st miner gone to 2.1 GHS, and the second one gave 1.9 GHS. The third miner just gone zombie...
Seems like there is not enough power? But funny enough I got the same 2.5 GHS result from USB2.0 PC port.... so current shouldn't be a problem here, probably....
I am, not quite sure WTF is wrong with AntimnerU1. Maybe some kind of connection/synchronisation errors with cgminer...
I tried to run BFG miner, but it couldnt find any device at all....even after i deinstall zadig drivers.
Moreover CGMINER cant find a device directly from USB3.0 (my USB3.0 hub is connected to the USB2.0 Port) port.
I will purchase a PCI USB2.0 extension card with separate powersupply from the main PSU. Costs only 10 Euro, so i give it a try, in case there is not enough current...
Thanks guys, I have plenty to try. I'll go one by one in order to understand what's happening. Cheers!
That's an idea. Theorically, at 0881 instead of 0981, I would get 1.8GH/s so 3.6GH/s instead of the actual 3.7GH/s.
Is there a way to reduce HW w/o touching the actual overclocking?
Would it be good to try an external usb hub or bfgminer?
Cheers
Is there a way to reduce HW w/o touching the actual overclocking?
Would it be good to try an external usb hub or bfgminer?
Cheers
You need a powered hub, your pc slots do not always provide enough power.
Give BFG a try.
plenty of info here:- https://bitcointalksearch.org/topic/m.4453006
try going back a few days in the thread.
That's an idea. Theorically, at 0881 instead of 0981, I would get 1.8GH/s so 3.6GH/s instead of the actual 3.7GH/s.
Is there a way to reduce HW w/o touching the actual overclocking?
Would it be good to try an external usb hub or bfgminer?
Cheers
Is there a way to reduce HW w/o touching the actual overclocking?
Would it be good to try an external usb hub or bfgminer?
Cheers
Stick 0 get hardly any errors, while stick 1 is getting way too many...try taking out stick 1 for 2 seconds and put back in again.
That's an idea. Theorically, at 0881 instead of 0981, I would get 1.8GH/s so 3.6GH/s instead of the actual 3.7GH/s.
Is there a way to reduce HW w/o touching the actual overclocking?
Would it be good to try an external usb hub or bfgminer?
Cheers
Is there a way to reduce HW w/o touching the actual overclocking?
Would it be good to try an external usb hub or bfgminer?
Cheers
Thanks for your fast reply!
The overclocking seems to be good but the real number at pool is 3.7GH/S. Which seems to correlate with my actual HW error. I'm wondering if it's too much...
Thank you again.
The overclocking seems to be good but the real number at pool is 3.7GH/S. Which seems to correlate with my actual HW error. I'm wondering if it's too much...
Thank you again.
You are getting about 10% hardware error...try to clock one notch lower to 1.8 gh/s and see if the 3.7gh/s at pool improves.
Thanks for your fast reply!
The overclocking seems to be good but the real number at pool is 3.7GH/S. Which seems to correlate with my actual HW error. I'm wondering if it's too much...
Thank you again.
The overclocking seems to be good but the real number at pool is 3.7GH/S. Which seems to correlate with my actual HW error. I'm wondering if it's too much...
Thank you again.
Hello!
I'm pretty new about bitcoin mining and would need some advice.
I started 2 X U1 yesterday evening and wondering if my setup seems good. I'm using a fan and my pc usb.
Here is my actual Cgminer screen (don't know how to attach an image, hoping it'll work...)
http://C:\cgminer.png
In case my pic does not show up here is a summary :
A : 54234 : Accepted
R : 162 : Rejected
HW 5007 Hardware error
AMU 0 A:29367 R:84 HW: 170 WU: 27.9/m
AMU 1 A:24873 R:78 HW: 4837 WU: 23.6/m
Started 2014-01-10 20:36 Time started
Now 2014-01-11 14:07 Time @ analysed
Running since : 0.73 Day
Real BTC from pool : 0.00095
Real BTC/day : 0.00130
Theorical BTC/day : 0.00140 according to a bitcoin calculator
Real BTC from pool / Theorical BTC : 93%
Adjusted BTC/day : 0.00130
A / (A+HW) : 92%
A / (A+R+HW) : 91%
HW / (A+HW) : 8.5%
HW / (A+R+HW) : 8.4%
Mean GH/s - Cgminer : 4.02600
Mean GH/s - Pool : 3.69806
% Cgminer / Pool : 92%
According to these numbers I have an efficacity of about 92-93%. Is it ok or should I be concerned? Are my math ok?
I'm running this line :
cgminer.exe --bmsc-options 115200:20 -o stratum.bitcoin.cz:3333 -u MYUSER -p MYPASS --bmsc-freq 0981
I read somewhere else we should expect around 1% of HW. Am I overclocking too much? The U1 are barely warm, around 30-35C.
Thank you guys!
Cheers,
Cool77
I'm pretty new about bitcoin mining and would need some advice.
I started 2 X U1 yesterday evening and wondering if my setup seems good. I'm using a fan and my pc usb.
Here is my actual Cgminer screen (don't know how to attach an image, hoping it'll work...)
http://C:\cgminer.png
In case my pic does not show up here is a summary :
A : 54234 : Accepted
R : 162 : Rejected
HW 5007 Hardware error
AMU 0 A:29367 R:84 HW: 170 WU: 27.9/m
AMU 1 A:24873 R:78 HW: 4837 WU: 23.6/m
Started 2014-01-10 20:36 Time started
Now 2014-01-11 14:07 Time @ analysed
Running since : 0.73 Day
Real BTC from pool : 0.00095
Real BTC/day : 0.00130
Theorical BTC/day : 0.00140 according to a bitcoin calculator
Real BTC from pool / Theorical BTC : 93%
Adjusted BTC/day : 0.00130
A / (A+HW) : 92%
A / (A+R+HW) : 91%
HW / (A+HW) : 8.5%
HW / (A+R+HW) : 8.4%
Mean GH/s - Cgminer : 4.02600
Mean GH/s - Pool : 3.69806
% Cgminer / Pool : 92%
According to these numbers I have an efficacity of about 92-93%. Is it ok or should I be concerned? Are my math ok?
I'm running this line :
cgminer.exe --bmsc-options 115200:20 -o stratum.bitcoin.cz:3333 -u MYUSER -p MYPASS --bmsc-freq 0981
I read somewhere else we should expect around 1% of HW. Am I overclocking too much? The U1 are barely warm, around 30-35C.
Thank you guys!
Cheers,
Cool77
Jump to: