BFG and Antminer U2

[hardhouseinc]

Seem to have found the happy medium

--set-device antminer:clock=x0981


 ST:2  F:0  NB:1  AS:0  BW:[ 67/ 36 B/s]  E:7.38  I: 7.90uBTC/hr  BS:33
 1            |  2.17/ 1.97/ 2.00Gh/s | A:22 R:0+0(none) HW:0/none
-------------------------------------------------------------------------------
 AMU 0:       |  2.08/ 1.96/ 1.94Gh/s | A:22 R:0+0(none) HW:0/none
-------------------------------------------------------------------------------
 [2014-03-17 12:05:05] Accepted 51952aaf AMU 0  Diff 3/2
 [2014-03-17 12:05:06] Accepted 29f65084 AMU 0  Diff 6/2
 [2014-03-17 12:05:11] Accepted 72dbf9e8 AMU 0  Diff 2/2

Tried --set-device antminer:clock=x0B81 and got ALOT of errors immediately.

Gonna try this for a day and see.

Then try --set-device antminer:clock=x0A81 and see if it hashes better with a
USB powered hub and runs stable or if the HW errors come back.

This seems to be alot of trial and error with these things.

[hurricandave]

   If your trying to run more than one Asic at a time, I would highly recomend using a Powered Hub. Now, if you don't know, keep in mind. Some Powered Hubs will send power thru the USB connector back to the computer if they don't have a diode on the wire inside the hub. Its not a big deal if you remember to turn the hub off before shutting down the computer. It can do weird things to a computer trying to get into sleep mode and interfeer with restarts as the motherboard can't properly power down to the correct S-mode before it tries to start back up. If the hub doesn't have a power switch just use a surge protector that does. 

[hardhouseinc]

Thanks for the advice.

So I plugged it into a powered USB hub at home.

--set-device antminer:clock=x0A81

Got steady 2.2Gh/s all night after that.  0 HW errors so far.

I put a small fan in front of the heatsink and the thing runs
way cooler than the OLD 333Mh/s USB block erupters.
So far these things seem pretty legit for the size and price.

[hurricandave]

That's good stuff.

[hardhouseinc]

Yah not bad.
At the 2.2Gh/s clock setting there are maybe 5 or 10 max errors in a 24 hour period so next to nothing.
Im now getting more than 0 errors on my 333Mh/s USB erupters after starting the newer version
of BFG miner.  The super old version I started with always showed 0 errors but it was always crashing
and had a hard time to start up again so using the new version is better and doesnt constantly crash
if there is even the slightest internet hiccup.

Im getting 1 or 2 hour bursts up to 2.3Gh/s per my BTC dashboard.  Not bad.


XXXX_1   2.32 GH/s   48976 (99.68%)   158 / 10 / 0   0:00:15        = sexy stable
   
XXXX_2   0.685 GH/s   15722 (99.73%)   42 / 0 / 0   0:00:15

---------

Totals   3.00 GH/s   64698 (99.69%)   200 / 10 / 0   

[hurricandave]

     I am still using my GPU as a hedge against the rejects. Everyone tries to say GPU mining is a waste but if you already have one in place I think its a waste not to at least let it run at default speed and cover your stales and dupes.

[Cassey]

Just received (3) Antminer U2s to augment my (6) Antiminer U1s.

I originally tried running them all together on an Anker 10 port hub with a 5amp 12V power supply, and they started to work, but one was freaking out with hardware errors.  All 9 were launched with a x0981 speed parameter.

Move the three news ones to a cheap powered hub with a 2amp 5V power supply and a different miner machine so I could isolate and watch them.  At x0981 the problem child still complained, but runs steady at x0881.  Alas, after several hours the performance dropped to <0.50Gh/s, even though the short term counter indicated 2Gh/s.  No hardware errors.

Have just dropped down to x0681 (the default) to see what the multi-hour impact is.  If it runs stable, I'll move the two good ones back to the Anker hub and leave the problem child on the cheapie hub.

[hurricandave]

    I would isolate the bad one alone and verify against your pool that it is really underperforming by that much. My short time always ranges up and down 1.70 Gh/s to 2.6 Gh/s even though after 45 mins to an hour it stabilizes at my pool and slowly ranges between 1.44 Gh/s to 1.65 Gh/s. They usually spend the majority of the time at 1.54 Gh/s. You may try letting it launch without adding the speed argument and let it default on its own. I can't find where I read it but the U2 may draw up to 1.35 Amps when overclocked. The powered hub would be straining to maintain that many devices over an extended period of time. I bet the hubs power supply got smoking hot.

[Cassey]

Well, happy to report (2) of the (3) are merrily chunking along at ~2GHs right alongside the Antminer U1s in the Anker powered USB strip.  I did need to use short (4") USB extension cables since those heat-sinks are too large to allow the U2s to be placed in adjacent slots in the Antek.

#3 is running one step lower at 1.8GHs on a cheapie powered USB strip wtih no signs of slowing down after 24 hours. 

Having enough power appears to be key to stable run rates.

Cassey

[rca]

Does anybody know how many milliamps the Antiminer U2s are supposed to require to run stable?  It seems like the original Block Erupter ASIC miners took 500-510 ma, recommended.

[hurricandave]

I have been trying to save the money for an USB power meter but they cost as much as an Antminer U2, sooo....
  For now it seems that as long as they can get a solid 1/2 amp they are stable at default speeds. But if you plan to overclock look at your power adapter and figure 625-900 milliamps each device for stability and leave a little head room to prevent running the wall-wart at 100% 24-7.  The new heat-sinks are very efficient but a light breeze from a fan will help keep them happy and probably help reduce power consumption overcoming heat induced resistance.

[Cassey]

Just for the record...  I plugged in a cheap ($6) USB volt/amp meter between my hub and a Antminer U2 running overclocked at 0x0981.

Voltage = 5.14 - 5.16
Amperage = 0.32 - 0.33

This is with a "Charger Doctor".  Most of the reading we at the higher end, but I did see them dip a bit once in awhile.

Also happy to report the U2 works fine behind the Charger Doctor - I had some trouble with U1s being found.  Maybe I just was unlucky before.

Cassey

[tomwoodsnh]

I am eagerly awaiting my first U2. These responses will be of great assistance when I throw it in with my U1's and Bfgminer running on a Raspberry Pi system. I had all the same problems trying to get my U1's to work. Two things made a huge difference.
1. Purchased a large 5vdc switching power supply to run my USB hubs... allowing almost 1 amp capacity for each miner.
2. copied the following statements in the following order to my bfg commandline:
     --set-device antminer:clock=x0781  --icarus-options 115200:2:2   --icarus-timing long -S all

The icarus timing long statement causes the clock and hashrate to continually correct itself and is not appropriate in computer systems with a cpu load shared with other applications. You could experiment with "--icarus-timing short" which requires full CPU time without delays, and makes the self-corrections,  but only for a limited time (~1 hr). Observing this, I saw my hash rate jump all over the place but over the course of a couple days it smoothed out until now after a couple weeks of continuous running, I see 3.16 or 3.17 gh/s on each miner with total errors at 0% and total rejected at .08% Very pleasing.

[nybbler905]

I am eagerly awaiting my first U2. These responses will be of great assistance when I throw it in with my U1's and Bfgminer running on a Raspberry Pi system. I had all the same problems trying to get my U1's to work. Two things made a huge difference.
1. Purchased a large 5vdc switching power supply to run my USB hubs... allowing almost 1 amp capacity for each miner.
2. copied the following statements in the following order to my bfg commandline:
     --set-device antminer:clock=x0781  --icarus-options 115200:2:2   --icarus-timing long -S all

The icarus timing long statement causes the clock and hashrate to continually correct itself and is not appropriate in computer systems with a cpu load shared with other applications. You could experiment with "--icarus-timing short" which requires full CPU time without delays, and makes the self-corrections,  but only for a limited time (~1 hr). Observing this, I saw my hash rate jump all over the place but over the course of a couple days it smoothed out until now after a couple weeks of continuous running, I see 3.16 or 3.17 gh/s on each miner with total errors at 0% and total rejected at .08% Very pleasing.

I had my main system blow it's power supply so now I've got a dual Coppermine main board handling the connection.  In this case, the onboard USB can handle 1 ANT and 1 Erupter, from the outside, I have a 7 port hub ( bought the wrong kind, they are sideways ) and can only have it set as ant, erupter, ant, erupter and between the 2 closest to the power connection a D-Link USB extender ( 4 inch long USB male to female, now I know WHY they made these! ).  I cut and soldered an old D-Link wireless's power adapter to get the hub 5V@2.5A and it will do strange things when I don't have 1 ANT per hub IC including sudden and mysterious hashing issues on either the ANT or Erupters.  I've yet to use the --icarus-options as all 4 Erupterschug away around the rated 333 khash/sec and I can only go for =x0981 due to not adding the 10K SMD 0603 resistor ( yet ).  With what I've got mining, I know I'm going to have to get a new hub ( with proper orientation ) and another used wireless with adapter if/when I get more Ants!

Power issues show up as the weirdest errors and easier to troubble shoot if you can find what part of the hub it's plugged in to.

[ct1aic]

Greetings.

With the "BFGminer ... --set-device antminer:clock=x0A81 --icarus-options 115200:2:2  --icarus-timing long -S all ..." option, I'm getting more or less 4.5 GHash/s from my new Antminer U2. Is it a rist to the health of the miner? I'm cooling it with a Arctic fan...

bfgminer version 3.99.0 - Started: [2014-05-21 15:17:50] - [  0 days 00:17:12]
 [M]anage devices [P]ool management (S)ettings [D]isplay options  [H]elp [Q]uit
 Connected to nl1.ghash.io diff 1 with stratum as user ct1aic.RPI5
 Block: ...f87cdad1 #301922  Diff:8.85G (63.38Ph/s)  Started: [15:35:02]
 ST:7  F:0  NB:5  AS:0  BW:[ 62/ 97 B/s]  E:10.44  I: 8.30uBTC/hr  BS:2.57k
 5            |  5.75/ 5.95/ 3.50Gh/s | A:839 R:0+0(none) HW:7/.83%
--------------------------------------------------------------------------------
 BES 0:       | 335.9/338.1/391.1Mh/s | A: 94 R:0+0(none) HW:1/1.1%
 BES 1:       | 362.1/339.1/295.4Mh/s | A: 71 R:0+0(none) HW:4/5.3%
 BES 2:       | 336.0/338.1/316.2Mh/s | A: 76 R:0+0(none) HW:1/1.3%
 BES 3:       | 336.0/337.8/328.7Mh/s | A: 79 R:0+0(none) HW:1/1.2%
 AMU 0:      |  4.47/ 4.61/ 2.17Gh/s   | A:521 R:0+0(none) HW:0/none
--------------------------------------------------------------------------------
 [2014-05-21 15:34:21] Accepted 4c1b0ce8 AMU 0  pool 0 Diff 3/1
 [2014-05-21 15:34:22] Accepted d0347c53 AMU 0  pool 0 Diff 1/1
 [2014-05-21 15:34:24] Accepted dd4bfb94 AMU 0  pool 0 Diff 1/1
 [2014-05-21 15:34:25] Accepted a29ed1aa AMU 0  pool 0 Diff 1/1
 [2014-05-21 15:34:27] Accepted 935c4156 AMU 0  pool 0 Diff 1/1
 [2014-05-21 15:34:27] Accepted f64ef54b BES 0  pool 0 Diff 1/1
 [2014-05-21 15:34:30] Accepted ec87d112 AMU 0  pool 0 Diff 1/1
 [2014-05-21 15:34:32] Accepted 808e4bd3 AMU 0  pool 0 Diff 1/1
 [2014-05-21 15:34:34] Accepted acd5daec BES 2  pool 0 Diff 1/1
 [2014-05-21 15:34:35] Stratum from pool 0 requested work update
 [2014-05-21 15:34:38] Accepted 3ebb9c07 BES 2  pool 0 Diff 4/1
 [2014-05-21 15:34:39] Accepted 451e9491 BES 0  pool 0 Diff 3/1
 [2014-05-21 15:34:42] Accepted 1e9f909e BES 3  pool 0 Diff 8/1
 [2014-05-21 15:34:42] Accepted fa45b6cb AMU 0  pool 0 Diff 1/1
 [2014-05-21 15:34:43] Accepted b18a52b8 AMU 0  pool 0 Diff 1/1
 [2014-05-21 15:34:44] Accepted 5a070e85 AMU 0  pool 0 Diff 2/1
 [2014-05-21 15:34:44] Accepted 21367eb4 BES 2  pool 0 Diff 7/1
 [2014-05-21 15:34:44] Accepted 8e2a3fd2 AMU 0  pool 0 Diff 1/1
 [2014-05-21 15:34:45] Accepted 663de2fb AMU 0  pool 0 Diff 2/1
 [2014-05-21 15:34:46] Accepted 3f6039ca BES 2  pool 0 Diff 4/1
 [2014-05-21 15:34:47] Accepted 7b422759 AMU 0  pool 0 Diff 2/1
 [2014-05-21 15:34:48] Accepted 99a2464e BES 0  pool 0 Diff 1/1
 [2014-05-21 15:34:49] Accepted 88f57a4d BES 0  pool 0 Diff 1/1
 [2014-05-21 15:34:50] Accepted b46c08bb AMU 0  pool 0 Diff 1/1
 [2014-05-21 15:34:53] Accepted 2183a1a2 BES 2  pool 0 Diff 7/1
 [2014-05-21 15:34:53] Accepted 3a612b98 AMU 0  pool 0 Diff 4/1

[-ck]

Greetings.

With the "BFGminer ... --set-device antminer:clock=x0A81 --icarus-options 115200:2:2  --icarus-timing long -S all ..." option, I'm getting more or less 4.5 GHash/s from my new Antminer U2. Is it a rist to the health of the miner? I'm cooling it with a Arctic fan...

 AMU 0:      |  4.47/ 4.61/ 2.17Gh/s   | A:521 R:0+0(none) HW:0/none

You're misusing the icarus options and telling it to report double the hashrate. It's actually running at half that speed.

[ct1aic]

Greetings.

With the "BFGminer ... --set-device antminer:clock=x0A81 --icarus-options 115200:2:2  --icarus-timing long -S all ..." option, I'm getting more or less 4.5 GHash/s from my new Antminer U2. Is it a rist to the health of the miner? I'm cooling it with a Arctic fan...

 AMU 0:      |  4.47/ 4.61/ 2.17Gh/s   | A:521 R:0+0(none) HW:0/none

You're misusing the icarus options and telling it to report double the hashrate. It's actually running at half that speed.

Thanks for your help, ckolivas.

[long3000]

Hi all, I'm trying to run my Antminer U2 at 2.4GH/s but they seem to get stuck at 2.2GH/s no matter which clock hex I attempted.
Can someone give me some advice?

THanks.

https://lh3.googleusercontent.com/-A-PCv-V6te8/U5EQI3OkCNI/AAAAAAAAD7g/gsH2_nxyCWk/w1020-h851-no/CAM00034.jpg

[hurricandave]

You need to look at the U2 board and determine which variation of R1 and R2 resistors are physically present. It seems that as the U2 was developed these resistors were changed/upgraded and the values they have will determine ultimately what the OC can achieve. There may be a picture of the board on the Bitmain website under support in the U1 manual with the resistor values and top voltage supported, this will show in the chart what the top OC is for that exact device. Outside of that it may be an issue with the particular Miner program your running as each one behaves a bit different.

[nikkocz]

Hi, I am totally new with only one piece of Antminer U2. BFG shows 2.35 / 2.31 / 2.29Gh/s - but I dont know if it is OK... If I will put next one or two  into a powered USB hub, get I more power as from single used miners or equal?