50 GH/s BFL miner running at 15 GH. | Bitcointalksearch.org

lightfoot

legendary

Activity: 3164

Merit: 2258

I fix broken miners. And make holes in teeth :-)

Quote from: Photon939 on December 15, 2013, 09:52:54 PM

Unfortunately I doubt they care. Once they sell out the last of their 65nm chips that's the end of them, and the end of their "lifetime" warranty. Then we get to watch and see how many delays happen on the Monarch. I'm sitting back and watching this round Undecided

Probably not, but still it would be fun.

I'm not buying a Monarch, but I will bet dollars to donuts that the techniques I have figured out for these systems will work the same on Monarchs. And possibly other miners, maybe I should become a repair depot for companies or something.

You know, in retrospect that would have been a good idea: Have a troubleshooter or two identifying trends to minimize returns and such. RMA's are incredibly expensive and can tear down a company's profits in very short order.

But given that bitcoin is powered by delusion and optimism, I can see why that is not done. Interesting :-)

Photon939

sr. member

Activity: 452

Merit: 250

Quote from: lightfoot on December 15, 2013, 08:21:28 PM

I had an idea: Every Single owner should email BFL and ask that they send some wrecked units to Lightfoot for analysis. I'd be more than happy to figure this out once and for all.

I am working on the Jalapeno failure issue, but that only cost me .15btc+$50 shipping plus heck knows what else. Can't do that for single prices. :-)

C

Unfortunately I doubt they care. Once they sell out the last of their 65nm chips that's the end of them, and the end of their "lifetime" warranty. Then we get to watch and see how many delays happen on the Monarch. I'm sitting back and watching this round Undecided

lightfoot

legendary

Activity: 3164

Merit: 2258

I fix broken miners. And make holes in teeth :-)

I had an idea: Every Single owner should email BFL and ask that they send some wrecked units to Lightfoot for analysis. I'd be more than happy to figure this out once and for all.

I am working on the Jalapeno failure issue, but that only cost me .15btc+$50 shipping plus heck knows what else. Can't do that for single prices. :-)

C

BitCloud

newbie

Activity: 48

Merit: 0

had the same issue, even slower actually

put it in the freezer for 1h

full spec

notme

legendary

Activity: 1904

Merit: 1002

Got it open. All fans were working. Removed thermal pads and applied paste. Hashing at 37GH/s now, which is better, but still not what it should be. Software is now recognizing it as a BAS with 16 cores.

sameev29

full member

Activity: 238

Merit: 100

Its BFL what can we expect after all the bad experience people are having.

vayvanne

full member

Activity: 217

Merit: 101

I heard some noise when unpacking and a screw dropped out of miner. Lips sealed

One of five miners was DOA. Other one hashes at 4, 8 or sometimes 15 GH/s out of 25-30.
Still waiting for reply from [email protected].

notme

legendary

Activity: 1904

Merit: 1002

Code:

BFL 0: 59.0C | 26.84/25.47/21.17Gh/s | A:144 R:24+0( 12%) HW:6/.70%

Code:

Select processor to manage using up/down arrow keys
 BFL 0a: 60.0C |  3.16/ 3.02/ 2.56Gh/s | A: 20 R: 4+0( 15%) HW:0/none
  BitFORCE SHA256 SC from FTDI
Kernel: parallel queue
Temperatures: 30.0C 60.0C
Voltages: 3.264 / 0.915 / 18.620

Using the arrow keys I see BFL 0a - BFL 0h.

As far as I can tell, both top fans are running, but even with a flashlight I can't really see well until I get it open.

Thanks for all the help guys. I'm leaving tomorrow to visit family for Thanksgiving, but I'll update this when I get back and get a T10 bit.

lightfoot

legendary

Activity: 3164

Merit: 2258

I fix broken miners. And make holes in teeth :-)

Quote from: notme on November 25, 2013, 04:59:00 PM

Stock firmware. There is one usb port.

I will give bfgminer a try later. What specifically should I be looking for?

See if it hashes at 30 or 60. You can also select Display, "M" option to see all the engines it sees. If it only sees 8 then something is definately odd.

Can you tell if both top fans are running inside the unit? Any smell of smoke or burned at all when you stick your nose to the external fan openings?

C

notme

legendary

Activity: 1904

Merit: 1002

Stock firmware. There is one usb port.

I will give bfgminer a try later. What specifically should I be looking for?

lightfoot

legendary

Activity: 3164

Merit: 2258

I fix broken miners. And make holes in teeth :-)

CK, ok; thought so. I have my BFL call itself all sorts of things, WAIT A SEC I can load that with all sorts of diagnostic messages if I want to. Thanks for the thought, now I have a better way of getting internal state out to the world without screwing with the USB port....

Anyway, from my POV that looks like a totally normal little single. Running at about the same temps as my jally with 5 chips, I'm cruising at 55c or so with similar voltages. So it's sinking heat. Also that hashing level looks like the limit is enabled, as it's really close to 30gh+4%. This is a stock firmware, right (yes, I'm asking again, just verifying)

Just as a cross check can you try it with BFGMiner? But I'd guess half the unit is down and the half running also powers the USB interface (it does have only one USB interface, right?)

C

notme

legendary

Activity: 1904

Merit: 1002

Quote from: -ck on November 25, 2013, 03:24:22 PM

Quote from: notme on November 21, 2013, 03:01:19 PM

I left it off longer this time.... now about where it was when I unboxed it, but still half of expected. I'll let it run for now until I can get a smaller hex bit.

Code:

BAL  1:  max 56C 3.26V | 25.59G/25.49Gh/s | A:  453 R:  61 HW:  4 WU:339.7/m

BAL means it's looking like a little single to cgminer, not a single. So either you don't actually have a single, or one of the boards isn't working at all or some other software confusion on cgminer's part. Try 'java api stats' if you have the api enabled (--api-listen) and see what information is in the dump.

Excluding most of it since I have a bunch of AM hardware plugged in too, but here is the entry for BAL0:

Code:

[STATS25] =>
(
   [STATS] => 25
   [ID] => BAL0
   [Elapsed] => 25
   [Calls] => 0
   [Wait] => 0.000000
   [Max] => 0.000000
   [Min] => 99999999.000000
   [Temp1] => 30.05
   [Temp2] => 56.86
   [Vcc1] => 3.264
   [Vcc2] => 0.918
   [Vmain] => 19.638
   [Temp1 Max] => 31.00
   [Temp2 Max] => 58.00
   [Temp1 Max Time] => 1385413534
   [Temp2 Max Time] => 1385413546
   [Work Queued] => 19
   [Work Complete] => 174
   [Overheat] => false
   [Flush ID] => 105
   [Result ID] => 109
   [Flushed] => true
   [Scan Sleep] => 336
   [Results Sleep] => 100
   [Work ms] => 143
   [Result Size] => 65/64/31/10/4/0/0/0/0/0
   [Driver] => 2
   [Firmware] =>  1.2.9
   [Chips] =>  YES @ 16
   [Que Size] => 40
   [Que Full] => 36
   [Que Watermark] => 32
   [Que Low] => 16
   [GetInfo] => DEVICE: BitFORCE SC0x0a
FIRMWARE: 1.2.90x0a
IAR Executed: NO0x0a
CHIP PARALLELIZATION: YES @ 160x0a
QUEUE DEPTH:400x0a
PROCESSOR 0: 16 engines @ 267 MHz -- MAP: FFFF0x0a
PROCESSOR 1: 15 engines @ 244 MHz -- MAP: FFBF0x0a
PROCESSOR 2: 16 engines @ 252 MHz -- MAP: FFFF0x0a
PROCESSOR 3: 15 engines @ 244 MHz -- MAP: DFFF0x0a
PROCESSOR 4: 15 engines @ 255 MHz -- MAP: FFDF0x0a
PROCESSOR 5: 16 engines @ 249 MHz -- MAP: FFFF0x0a
PROCESSOR 6: 16 engines @ 252 MHz -- MAP: FFFF0x0a
PROCESSOR 7: 15 engines @ 244 MHz -- MAP: FFF70x0a
THEORETICAL MAX: 31125 MH/s0x0a
ENGINES: 1240x0a
FREQUENCY: 274 MHz0x0a
CRITICAL TEMPERATURE: 00x0a
TOTAL THERMAL CYCLES: 00x0a
XLINK MODE: MASTER0x0a
XLINK PRESENT: NO0x0aOK0x0a0x00
   [USB Pipe] => 0
   [USB Delay] => r0 0.000000 w0 0.000000
   [USB tmo] => 0 0
)

-ck

legendary

Activity: 4088

Merit: 1631

Ruu \o/

Quote from: lightfoot on November 25, 2013, 03:57:45 PM

Quote from: -ck on November 25, 2013, 03:24:22 PM

Quote from: notme on November 21, 2013, 03:01:19 PM

I left it off longer this time.... now about where it was when I unboxed it, but still half of expected. I'll let it run for now until I can get a smaller hex bit.

Code:

BAL  1:  max 56C 3.26V | 25.59G/25.49Gh/s | A:  453 R:  61 HW:  4 WU:339.7/m

BAL means it's looking like a little single to cgminer, not a single. So either you don't actually have a single, or one of the boards isn't working at all or some other software confusion on cgminer's part. Try 'java api stats' if you have the api enabled (--api-listen) and see what information is in the dump.

CK, I run BFG myself, but how do you know if something is a little single, single, or truck? Reason is if this person loaded a custom firmware and you base your "what it is" based on the string back then I could see how cgminer would fail to load all the chips.

It doesn't care, it just sends continuous work regardless of what it is, be it a BAJ BAL BAS or BAM. The only difference is the name comes up differently. It's not like cgminer goes "ooh you're a piece of shit, let me only hand you shit work instead of lots of work".

lightfoot

legendary

Activity: 3164

Merit: 2258

I fix broken miners. And make holes in teeth :-)

Quote from: -ck on November 25, 2013, 03:24:22 PM

Quote from: notme on November 21, 2013, 03:01:19 PM

I left it off longer this time.... now about where it was when I unboxed it, but still half of expected. I'll let it run for now until I can get a smaller hex bit.

Code:

BAL  1:  max 56C 3.26V | 25.59G/25.49Gh/s | A:  453 R:  61 HW:  4 WU:339.7/m

BAL means it's looking like a little single to cgminer, not a single. So either you don't actually have a single, or one of the boards isn't working at all or some other software confusion on cgminer's part. Try 'java api stats' if you have the api enabled (--api-listen) and see what information is in the dump.

CK, I run BFG myself, but how do you know if something is a little single, single, or truck? Reason is if this person loaded a custom firmware and you base your "what it is" based on the string back then I could see how cgminer would fail to load all the chips.

C

-ck

legendary

Activity: 4088

Merit: 1631

Ruu \o/

Quote from: notme on November 21, 2013, 03:01:19 PM

I left it off longer this time.... now about where it was when I unboxed it, but still half of expected. I'll let it run for now until I can get a smaller hex bit.

Code:

BAL  1:  max 56C 3.26V | 25.59G/25.49Gh/s | A:  453 R:  61 HW:  4 WU:339.7/m

BAL means it's looking like a little single to cgminer, not a single. So either you don't actually have a single, or one of the boards isn't working at all or some other software confusion on cgminer's part. Try 'java api stats' if you have the api enabled (--api-listen) and see what information is in the dump.

lightfoot

legendary

Activity: 3164

Merit: 2258

I fix broken miners. And make holes in teeth :-)

By the way, someone shared with me an interesting set of pictures of an SC30 that was having issues. The important thing from my understanding is that SC30's have only one side of chips, including only one side of power FETs.

That is very interesting. It implies that the two sides are actually independent, boosting this is the possiblity that there are two power feeds from the power supply. If that's the case, then it's possible that an SC60 could have a power failure either in the supply or the +1 rail that would shut down half the unit and leave the other half running (depending on which side also runs the USB logic).

So if your unit is only hashing at half power, see if you can check the power supplies. If one is down, do NOT try to backfeed with a big power supply unless you put a fuse in series, something like a 6-7a fast blow. For proper testing I would use a 3a fuse and not start hashing (just see if the board comes up). If the fuse blows, you have a short on that side of the board.

If you send unregulated power into a shorted FET, there will be... smoke....

C

lightfoot

legendary

Activity: 3164

Merit: 2258

I fix broken miners. And make holes in teeth :-)

Quote from: notme on November 24, 2013, 04:43:13 PM

Quote from: lightfoot on November 24, 2013, 01:33:26 PM

Odd. A co-worker had a 25gh unit and same thing happened. I recommended putting it in the freezer for 15 minutes, he did it and it came right up.

Given that I have been abusing the hell out of my 20gh jalapeno here by basically dumping a volcano's worth of heat on the board, I wonder what is up...

One way to find out, there is a utility over on the BFL forums, the handy dandy bfl com port scanner that works right with FTDI drivers and will dump out what the unit is thinking. Try grabbing it and giving it a shot.

Another thought on those lines, grab another computer, load up bfgminer, and see what it says. Do the D for display, then M to view all the CPUs and see what's up. Maybe it's conflicting with something else in your mining pile.

C

Do you have a link to this debugging tool, and does it work with Linux?

Google says:

https://forums.butterflylabs.com/jalapeno-single-sc-support/4564-diagnostic-tool-bfl-com-port-scanner.html

And I don't think it does.

C

notme

legendary

Activity: 1904

Merit: 1002

Quote from: lightfoot on November 24, 2013, 01:33:26 PM

Odd. A co-worker had a 25gh unit and same thing happened. I recommended putting it in the freezer for 15 minutes, he did it and it came right up.

Given that I have been abusing the hell out of my 20gh jalapeno here by basically dumping a volcano's worth of heat on the board, I wonder what is up...

One way to find out, there is a utility over on the BFL forums, the handy dandy bfl com port scanner that works right with FTDI drivers and will dump out what the unit is thinking. Try grabbing it and giving it a shot.

Another thought on those lines, grab another computer, load up bfgminer, and see what it says. Do the D for display, then M to view all the CPUs and see what's up. Maybe it's conflicting with something else in your mining pile.

C

Do you have a link to this debugging tool, and does it work with Linux?

lightfoot

legendary

Activity: 3164

Merit: 2258

I fix broken miners. And make holes in teeth :-)

Odd. A co-worker had a 25gh unit and same thing happened. I recommended putting it in the freezer for 15 minutes, he did it and it came right up.

Given that I have been abusing the hell out of my 20gh jalapeno here by basically dumping a volcano's worth of heat on the board, I wonder what is up...

One way to find out, there is a utility over on the BFL forums, the handy dandy bfl com port scanner that works right with FTDI drivers and will dump out what the unit is thinking. Try grabbing it and giving it a shot.

Another thought on those lines, grab another computer, load up bfgminer, and see what it says. Do the D for display, then M to view all the CPUs and see what's up. Maybe it's conflicting with something else in your mining pile.

C

bimblebob

newbie

Activity: 10

Merit: 0

Thats the real screwdriver T10. Change the paste and it should work better. Maybe you have there little single?

Topic: 50 GH/s BFL miner running at 15 GH. (Read 3883 times)