Avalon ASIC users thread - page 140.

loshia

legendary

Activity: 1610

Merit: 1000

Quote from: TwinTurbo on July 03, 2013, 05:39:16 AM

I received my Avalon boxes today, but one of them has severe problems:

All of them hash just fine at 300 MHz, but one of them only works if I set it to 256 (but still problems).

When it starts hashing, it will climb to about 66 GH, but then resets every 1-5 minutes, sometimes resets even quicker.

This box had an older firmware on it than my other ones. It had 3.1.1 instead of 3.2.1 (I think).

The temperatures seem to be pretty normal. But it just resets a lot. Sometimes it will not start again even if I stop and start the service, and I have to power off and on again.

What are some things I can do to troubleshoot the issues with this box?

Check FPGA controller board USB fuse and power supply to be at least 750W. Then upgrade software and try

TwinTurbo

member

Activity: 91

Merit: 10

I received my Avalon boxes today, but one of them has severe problems:

All of them hash just fine at 300 MHz, but one of them only works if I set it to 256 (but still problems).

When it starts hashing, it will climb to about 66 GH, but then resets every 1-5 minutes, sometimes resets even quicker.

This box had an older firmware on it than my other ones. It had 3.1.1 instead of 3.2.1 (I think).

The temperatures seem to be pretty normal. But it just resets a lot. Sometimes it will not start again even if I stop and start the service, and I have to power off and on again.

What are some things I can do to troubleshoot the issues with this box?

candoo

hero member

Activity: 602

Merit: 500

Vertrau in Gott

Quote from: -ck on July 03, 2013, 12:31:24 AM

Quote from: cypherdoc on July 03, 2013, 12:26:02 AM

anything else you see that i could improve upon in the settings?

All looks pretty good. My only concern is more about your Avalon itself. It's obviously hot where you are at the moment, as your fans have ramped up to almost the maximum. Due to them including 2 high pressure 38mm fans and 1 regular pressure 25mm fan, when the 38mm fans run at maximum speed, the air pressure is so high in the case that the 25mm fan is blowing but air is actually moving backwards through it. I don't know what the solution there is...

I removed the stock fans and bought 3 Scythe fans with 5400 RPM. I power them with 12V and 100% via a molex connector. Its very loud but works fine and provides good temperatures

stfnnn

member

Activity: 75

Merit: 10

Quote from: -ck on July 01, 2013, 06:24:23 PM

Slightly updated firmware:

http://ck.kolivas.org/apps/cgminer/avalon/20130702-1/openwrt-ar71xx-generic-tl-wr703n-v1-squashfs-factory.bin

This rolls back to the previous hardware error target of <2% when using avalon-auto which is worth about 1.5GH more on average.

EDIT: Adds a check for maximum fanspeed before throttling speed at higher temps as well.

Thank you for your work! Donating some hashes as we speak!

SolarSilver

legendary

Activity: 1112

Merit: 1000

Quote from: -ck on July 03, 2013, 01:31:25 AM

Accepted/Rejected is changed in the current master version for precisely this reason. The web interface, on the other hand, may lag. It's the legacy of constantly shifting changing landscape of bitcoin mining.

I am not blaming you for legacy cruft, don't get me wrong...

By the way, for those using the 0519 firmware from Strombom, don't forget that old cgminer pre 3.2 counted errors as hashrate, which affects the calculations. Another thing that confuses people when comparing performance ;-)

spiccioli

legendary

Activity: 1379

Merit: 1003

nec sine labore

Quote from: cypherdoc on July 03, 2013, 12:17:42 AM

Quote from: jddebug on July 03, 2013, 12:10:37 AM

Is it stratum vardiif ramping up so you are submitting far less shares but same hashrate?

is there a way to tell?

Yes,

look at the lines with "Checking retarget for...", you'll see that it grows slowing down share submission.

spiccioli

Tesla71

sr. member

Activity: 302

Merit: 252

3 Questions:

- my cgminer process seems to restart after every 4 hours or so.. must watch longer to see if it happens regulary. could it have something to do with the --avalon-auto, maybee the freq gets to high and it restarts? I am using --avalon-freq 300-340

- As I read I can forget the "discarded" shares, what about the number of "rejected", do I have to set them also in relation to diff1shares or to the accepted?

-Is it save to use --avalon-temp 55 ? the fans are running at about 2700 under that environment conditions atm.

-ck

legendary

Activity: 4088

Merit: 1631

Ruu \o/

Accepted/Rejected is changed in the current master version for precisely this reason. The web interface, on the other hand, may lag. It's the legacy of constantly shifting changing landscape of bitcoin mining.

SolarSilver

legendary

Activity: 1112

Merit: 1000

Quote from: -ck on July 03, 2013, 12:33:46 AM

Quote from: SolarSilver on July 03, 2013, 12:29:05 AM

I think the interface is confusing most of us: to the occasional observer your system has

Accepted: 1013
Not accepted: (Rejected 4 + HW 406 + discarded 64)= 474

so about 31% is lost? :-)

Sigh, also wrong.

The absolute values that count here are:
Accepted: 39126
Rejected: 789
HW: 406

Again, this proves my point that the way the values are displayed is confusing

The occasional observer does not see Accepted: 39126 (He/she sees 1013)

Where is that value? You calculate it?

Hands up for those that also misread the stats page (and are not afraid to admit it)?

cypherdoc

legendary

Activity: 1764

Merit: 1002

Quote from: BenTuras on July 03, 2013, 12:48:33 AM

Quote from: cypherdoc on July 02, 2013, 06:14:36 PM

however, initially it was streaming thru pages of hashes w/o problem but now has slowed to a crawl. some memory buffer or something must be filling up.

Is it possible that your disk is near its capacity because of the hugh logfile ?

this was answered above: vardiff.

cypherdoc

legendary

Activity: 1764

Merit: 1002

Quote from: Vagnavs on July 03, 2013, 12:34:28 AM

how long does it typically take to be delivered from China to the US? Any feedback would be greatly appreciated!
Thanks for time,
Brian

not long. 4 days max?

i got tracking # for my last avalon yesterday and i just checked DHL and it's already left Hong Kong for LA.

BenTuras

hero member

Activity: 826

Merit: 1001

Quote from: -ck on July 03, 2013, 12:11:00 AM

Hardware error percentage is:
100 * HW / (diff1shares + HW)

Friendly suggestion: Is it possible to add this percentage to the cgminer web page to make it easier for us ?

BenTuras

hero member

Activity: 826

Merit: 1001

Quote from: cypherdoc on July 02, 2013, 06:14:36 PM

however, initially it was streaming thru pages of hashes w/o problem but now has slowed to a crawl. some memory buffer or something must be filling up.

Is it possible that your disk is near its capacity because of the huge logfile ?

demkd

full member

Activity: 242

Merit: 100

Quote from: -ck on July 03, 2013, 12:42:14 AM

That's unusual. It should work out to 42:300 now. Did you start with a fresh configuration or have you always done "keep settings" from the old firmware?

Yes, I flashed with "keep settings", I will try reflash later today w/o keeping settings.

-ck

legendary

Activity: 4088

Merit: 1631

Ruu \o/

Quote from: demkd on July 03, 2013, 12:35:16 AM

Quote from: -ck on July 02, 2013, 11:48:59 PM

I'm not really sure why this happens, but it appears to happen when code outside of cgminer kills it, so I suspect it's because it's not allowing it time to shut down properly and perhaps the API port is still bound to the old zombie process and you're not seeing the new one mining, or the new cgminer process is fighting the old one for access to the device.

But I see only one cgminer process, and unit not mining at this time (I have watt meter ~34W at idle) and GetWorks slowly incresing over time so It's a same process.
I see difference in options & 07/02 frimware:
06/07 frimware: --avalon-options 115200:24:10:43:300 (mining)
07/02 frimware: --avalon-options 115200:24:10:38:300 (not mining & zero temps displayed)
What timeout value means?

That's unusual. It should work out to 42:300 now. Did you start with a fresh configuration or have you always done "keep settings" from the old firmware?

cypherdoc

legendary

Activity: 1764

Merit: 1002

Quote from: -ck on July 03, 2013, 12:31:24 AM

Quote from: cypherdoc on July 03, 2013, 12:26:02 AM

anything else you see that i could improve upon in the settings?

All looks pretty good. My only concern is more about your Avalon itself. It's obviously hot where you are at the moment, as your fans have ramped up to almost the maximum. Due to them including 2 high pressure 38mm fans and 1 regular pressure 25mm fan, when the 38mm fans run at maximum speed, the air pressure is so high in the case that the 25mm fan is blowing but air is actually moving backwards through it. I don't know what the solution there is...

what is the max fan speed? 4000? what's a good range for them to be in?

which of the 3 fans is the 25 mm? the one closest to the PSU?

edit: what's an optimal temp to run at? someone who knows should publish a set of normal or optimum values for all these parameters.

demkd

full member

Activity: 242

Merit: 100

Quote from: -ck on July 02, 2013, 11:48:59 PM

I'm not really sure why this happens, but it appears to happen when code outside of cgminer kills it, so I suspect it's because it's not allowing it time to shut down properly and perhaps the API port is still bound to the old zombie process and you're not seeing the new one mining, or the new cgminer process is fighting the old one for access to the device.

But I see only one cgminer process, and unit not mining at this time (I have watt meter ~34W at idle) and GetWorks slowly incresing over time so It's a same process.
I see difference in options & 07/02 frimware:
06/07 frimware: --avalon-options 115200:24:10:43:300 (mining)
07/02 frimware: --avalon-options 115200:24:10:38:300 (not mining & zero temps displayed)
What timeout value means?

Vagnavs

legendary

Activity: 1121

Merit: 1003

how long does it typically take to be delivered from China to the US? Any feedback would be greatly appreciated!
Thanks for time,
Brian

-ck

legendary

Activity: 4088

Merit: 1631

Ruu \o/

Quote from: SolarSilver on July 03, 2013, 12:29:05 AM

Quote from: cypherdoc on July 03, 2013, 12:00:03 AM

Quote from: ?? on ??

Looks fine to me? Your hardware error count is under 1.5% and your hashrate is 77GH.

i thought the error count was HW/Accepted shares?

actually no

https://bitcointalksearch.org/topic/m.2579903
https://bitcointalksearch.org/topic/m.2574423

Quote

as you can see, cgminer's hashing rate looks perfectly normal and now you're telling me the hardware error count is normal also. am i misinterpreting something here?

I think the interface is confusing most of us: to the occasional observer your system has

Accepted: 1013
Not accepted: (Rejected 4 + HW 406 + discarded 64)= 474

so about 31% is lost? :-)

Sigh, also wrong.

The absolute values that count here are:
Accepted: 39126
Rejected: 789
HW: 406

Discarded is totally meaningless, I wish it wasn't in the display at all. It's a number from when getwork was the protocol in use and means nothing in the local work generation era.

-ck

legendary

Activity: 4088

Merit: 1631

Ruu \o/

Quote from: cypherdoc on July 03, 2013, 12:26:02 AM

anything else you see that i could improve upon in the settings?

All looks pretty good. My only concern is more about your Avalon itself. It's obviously hot where you are at the moment, as your fans have ramped up to almost the maximum. Due to them including 2 high pressure 38mm fans and 1 regular pressure 25mm fan, when the 38mm fans run at maximum speed, the air pressure is so high in the case that the 25mm fan is blowing but air is actually moving backwards through it. I don't know what the solution there is...

Topic: Avalon ASIC users thread - page 140. (Read 438724 times)