Avalon ASIC users thread - page 193.

Tinua

hero member

Activity: 871

Merit: 1000

Quote from: luffy on March 19, 2013, 11:09:16 AM

Quote from: johnyj on March 18, 2013, 10:02:48 AM

Anyone have such high HW errors? What does this indicate?

   [Getworks] => 1963
   [Accepted] => 27062
   [Rejected] => 963
   [Hardware Errors] => 866
   [Utility] => 30.57
   [Discarded] => 3649

your utility is very low!
assuming that you are in a stratum pool try to set variable difficulty at 64 to 66, i am not sure about the exact number though :p

Hi luffy
how big should it be (Utility)?
I have also about 30.5. (vardiff 32)
Edit: With vardiff 64 i get 15.5 Utility

luffy

hero member

Activity: 607

Merit: 500

Quote from: johnyj on March 18, 2013, 10:02:48 AM

Anyone have such high HW errors? What does this indicate?

   [Getworks] => 1963
   [Accepted] => 27062
   [Rejected] => 963
   [Hardware Errors] => 866
   [Utility] => 30.57
   [Discarded] => 3649

your utility is very low!
assuming that you are in a stratum pool try to set variable difficulty at 64 to 66, i am not sure about the exact number though :p

ProfMac

legendary

Activity: 1246

Merit: 1002

Quote from: johnyj on March 19, 2013, 09:06:11 AM

PSU cold air intake is a little bit insufficient, I'm planning to add a fan out side of PSU exhaust to help cool the PSU. Currently at -1c degree there is no problem, but when summer arrives, ambient temperature could reach 25+, it will become an issue

Summer ambient for me can reach 43°C.

johnyj

legendary

Activity: 1988

Merit: 1012

Beyond Imagination

PSU cold air intake is a little bit insufficient, I'm planning to add a fan out side of PSU exhaust to help cool the PSU. Currently at -1c degree there is no problem, but when summer arrives, ambient temperature could reach 25+, it will become an issue

jddebug

sr. member

Activity: 446

Merit: 250

Quote from: opentoe on March 18, 2013, 10:07:06 PM

I'm in the US. Will one of these work on my 15-20amp circuit? I have all 110V lines of course, regular house. Do I need to install a 240 or something?

It only uses 600 watts. Should be fine on any 110 outlet unless you have a coffee maker and a microwave on it too.

opentoe

legendary

Activity: 1274

Merit: 1000

Personal text my ass....

I'm in the US. Will one of these work on my 15-20amp circuit? I have all 110V lines of course, regular house. Do I need to install a 240 or something?

SZD

newbie

Activity: 33

Merit: 0

How loud are these things? Also looks like if you have 20amp circuits you can only have 3 units per circuit assuming you don't have a ton of other stuff already plugged in! Then maybe just 2. So much power use...

johnyj

legendary

Activity: 1988

Merit: 1012

Beyond Imagination

Quote from: Bogart on March 18, 2013, 10:37:35 AM

Quote from: johnyj on March 18, 2013, 10:02:48 AM

Anyone have such high HW errors? What does this indicate?

   [Getworks] => 1963
   [Accepted] => 27062
   [Rejected] => 963
   [Hardware Errors] => 866
   [Utility] => 30.57
   [Discarded] => 3649

I did when I was overclocking with --avalon-options 115200:24:10:45:300

Changing to --avalon-options 115200:24:10:42:300 cleared it up for me (thanks to luffy for suggesting it).

Supposedly the fourth parameter is a "controller timeout". I'm not sure why reducing it helps.

I tried different parameters, including the default 45/282 value, the result is almost the same, hardware error rate about 3% of accepted, and my ambient temp is -3 degree now
   [fan1] => 0
   [fan2] => 720
   [fan3] => 720
   [temp1] => -3
   [temp2] => -1
   [temp3] => 28
   [temp_max] => 28

Bogart

legendary

Activity: 966

Merit: 1000

Quote from: kano on March 18, 2013, 10:57:04 AM

Quote from: Bogart on March 18, 2013, 10:49:48 AM

...
Probably. I imagine the restart will be the easy part. Detecting the failure condition programmatically will be the harder part.

Of course not ... my API has all sorts of information that will show the condition ... and if the API isn't working, you take that to mean the condition exists also.
Of course there are updates to the API but ...

Cool. I'm not too familiar with it. Is it JSONRPC just like bitcoind? Care to give a cookbook query string to get the current hashrate?

invader

sr. member

Activity: 266

Merit: 250

Someone asked if i moved all modules, or just one - the answer is
MOVE ALL THE MODULES Grin

kano

legendary

Activity: 4634

Merit: 1851

Linux since 1997 RedHat 4

Quote from: Bogart on March 18, 2013, 10:49:48 AM

...
Probably. I imagine the restart will be the easy part. Detecting the failure condition programmatically will be the harder part.

Of course not ... my API has all sorts of information that will show the condition ... and if the API isn't working, you take that to mean the condition exists also.
Of course there are updates to the API but ...

Bogart

legendary

Activity: 966

Merit: 1000

Quote from: cypherdoc on March 18, 2013, 10:41:16 AM

Quote from: Bogart on March 18, 2013, 09:23:53 AM

I encountered the fan/temp bug again this morning. I noticed it cycled the fans on maximum speed for about 200ms, and when I checked the cgminer status page again the temp and fanspeed readings showed non-sane values that varied rapidly and were sometimes negative.

It also stopped hashing. I presume as a safety measure.

I let it go like this for about 3 minutes this time to see if it would recover. It cycled the fans onto maximum for 200ms a few more times, but never did.

When I restarted cgminer via System --> Startup in the web UI, it then mined normally again and showed normal sensor values.

is there a way to automate the cgminer restart?

Probably. I imagine the restart will be the easy part. Detecting the failure condition programmatically will be the harder part.

cypherdoc

legendary

Activity: 1764

Merit: 1002

Quote from: Bogart on March 18, 2013, 09:23:53 AM

I encountered the fan/temp bug again this morning. I noticed it cycled the fans on maximum speed for about 200ms, and when I checked the cgminer status page again the temp and fanspeed readings showed non-sane values that varied rapidly and were sometimes negative.

It also stopped hashing. I presume as a safety measure.

I let it go like this for about 3 minutes this time to see if it would recover. It cycled the fans onto maximum for 200ms a few more times, but never did.

When I restarted cgminer via System --> Startup in the web UI, it then mined normally again and showed normal sensor values.

is there a way to automate the cgminer restart?

Bogart

legendary

Activity: 966

Merit: 1000

Quote from: johnyj on March 18, 2013, 10:02:48 AM

Anyone have such high HW errors? What does this indicate?

   [Getworks] => 1963
   [Accepted] => 27062
   [Rejected] => 963
   [Hardware Errors] => 866
   [Utility] => 30.57
   [Discarded] => 3649

I did when I was overclocking with --avalon-options 115200:24:10:45:300

Changing to --avalon-options 115200:24:10:42:300 cleared it up for me (thanks to luffy for suggesting it).

Supposedly the fourth parameter is a "controller timeout". I'm not sure why reducing it helps.

johnyj

legendary

Activity: 1988

Merit: 1012

Beyond Imagination

Anyone have such high HW errors? What does this indicate?

   [Getworks] => 1963
   [Accepted] => 27062
   [Rejected] => 963
   [Hardware Errors] => 866
   [Utility] => 30.57
   [Discarded] => 3649

Bogart

legendary

Activity: 966

Merit: 1000

I encountered the fan/temp bug again this morning. I noticed it cycled the fans on maximum speed for about 200ms, and when I checked the cgminer status page again the temp and fanspeed readings showed non-sane values that varied rapidly and were sometimes negative.

It also stopped hashing. I presume as a safety measure.

I let it go like this for about 3 minutes this time to see if it would recover. It cycled the fans onto maximum for 200ms a few more times, but never did.

When I restarted cgminer via System --> Startup in the web UI, it then mined normally again and showed normal sensor values.

invader

sr. member

Activity: 266

Merit: 250

So i moved the modules, and it seems to be better at first look.
WARNING: use your own mind when considering doing the same.

before

Quote

[2013-03-18 12:37:02] Avalon: Fan1: 0/m, Fan2: 2280/m, Fan3: 2280/m Temp1: 22C, Temp2: -1C, Temp3: 45C, TempMAX: 46C

after

Quote

[2013-03-18 16:19:23] Avalon: Fan1: 0/m, Fan2: 2040/m, Fan3: 2040/m Temp1: 19C, Temp2: -1C, Temp3: 44C, TempMAX: 44C

SolarSilver

legendary

Activity: 1112

Merit: 1000

Quote from: invader on March 18, 2013, 05:40:54 AM

I really smiled when i seen this picture, this is a strange solution.

I'm sure ngzhang will post that he puked when he saw that picture :-)

invader

sr. member

Activity: 266

Merit: 250

I really smiled when i seen this picture, this is a strange solution.

SolarSilver

legendary

Activity: 1112

Merit: 1000

Quote from: invader on March 18, 2013, 05:01:48 AM

Is there someone who tried to move hashmodules lower? Does this improove cooling or not?
As i see, bottom fan just mostly passing cool air thru case (without fourth hashmodule)...

One picture I've seen posted on a Chinese forum is this one:

but I'm not sure if this is much of an improvement, it would force all the air through the fins of the left module

Topic: Avalon ASIC users thread - page 193. (Read 438724 times)