Author

Topic: Swedish ASIC miner company kncminer.com - page 1270. (Read 3049528 times)

legendary
Activity: 4634
Merit: 1851
Linux since 1997 RedHat 4
October 19, 2013, 06:18:43 PM
What does BBB stand for?
BeagleBone Black - sorta like an RPi, for the uninitiated
http://beagleboard.org/Products/BeagleBone+Black/

I have one but I've never switched it on Smiley
sr. member
Activity: 378
Merit: 250
October 19, 2013, 06:13:35 PM
What does BBB stand for?
soy
legendary
Activity: 1428
Merit: 1013
October 19, 2013, 05:49:10 PM
back to all 3 sats running 0.96, but they all hang around 275 now still  Smiley
same procedure on all 3 saturns...
1. flash 0.90 on the miner, reboot
2. flash enablecores.bin to the miner, POWER CYCLE the miner.
3. putty in and watch cgminer until it crashes, when it does, go to the gui page again, and flash 0.96 back onto the miner, apply, then POWER CYCLE the miner(turn off for 15 seconds via the power supply switch)
thats it.
i had 3 sats going at 255, 265, and 275....  now all at 270 - 275ish


By crashes do you mean the hashrate falls badly?

I noticed the BBB temperature effected hashrate.  My front fans aren't really directed at the module any more.  I have a piece of cardboard down from the front top edge to the front of the module board and air goes out the sides.  If I remove the cardboard, at least yesterday when I found this and haven't retested it since, if I remove the cardboard the hashrate in the first number crashes very quickly.  I put the cardboard back and it comes up pretty quickly.  I removed the BBB and put stick-on heatsinks on the middle chip as that's where my IR detector says is heating.  I had two 12 v fans I was going to put on a Blade but changed my mind.  I have one of these each on 2 sides of the one module aiming at the vrm's.  I also have a 3" dia 120vac fan, very deep, pointed in from the side passing air over the BBB.  Keeping the BBB cool seems critical.

With all the extra room I have in the Mercury I'm thinking of getting ribbon cables and double pins so as to move the BBB somewhere else in the box where it can be cooled better.

Since you have the big fan cooling the board, try putting cardboard to force the 2 intake fans air down onto the BBB and see if the miner crashes under 0.90 then.

root@Mercury-3DF:~# uptime
 22:51:08 up  5:17,  load average: 1.83, 1.95, 1.94


Darn, I was just going to put the cgminer stats line here but when I hit "q" and quit the screen, unlike with minepeon, cgminer quit.  So, I've just restarted it from the command line but it had been cooking along at 134GH/s avg for 5hours 17 minutes.

Just adjusted my cardboard.  Better leave ½ fan clear on each side of the center and the cardboard in that center forcing air down onto the center of the BBB. 

The stats came back up pretty quickly.
KnC 0: | 162.6G/131.5Gh/s | A:39300 R:1695 HW:3638 WU:2150.2/m

Having changed the cardboard slightly, the first figure seems higher which bodes for an overall higher avg if it keeps up.
donator
Activity: 2352
Merit: 1060
between a rock and a block!
October 19, 2013, 05:33:12 PM
I looked at your website - the cost of your single blade is a half a bitcoin less than the cost of an Avalon mini...only difference is roughly 5.7-6X the hashing power.
in stock and shipping now:

https://bitcointalksearch.org/topic/closed-batch-32-065-btc-per-blade-49-port-hubs-30-or-less-303640
legendary
Activity: 1456
Merit: 1018
HoneybadgerOfMoney.com Weed4bitcoin.com
October 19, 2013, 05:28:32 PM
I looked at your website - the cost of your single blade is a half a bitcoin less than the cost of an Avalon mini...only difference is roughly 5.7-6X the hashing power.
donator
Activity: 2352
Merit: 1060
between a rock and a block!
October 19, 2013, 05:24:12 PM
I wonder how my miner is enjoying Ohio, now that it has spent 2 days there.

Maybe it is seeing the sights and visiting a few museums
did you call DHL?

Yes it cleared customs and it will be here Monday or Tuesday, whenever they get around to it.

What he said applies to me too...Canary is acting like I DIDNT make the phone call.
not acting... I'm very familiar with DHL, and from the comments it's hard to tell if the folks with miners sitting in OH called DHL or not, so just trying to help... I want you all to get them ASAP!!!
and promptly point them at the Eligius pool  Grin
legendary
Activity: 1456
Merit: 1018
HoneybadgerOfMoney.com Weed4bitcoin.com
October 19, 2013, 05:17:36 PM
I wonder how my miner is enjoying Ohio, now that it has spent 2 days there.

Maybe it is seeing the sights and visiting a few museums
did you call DHL?

Yes it cleared customs and it will be here Monday or Tuesday, whenever they get around to it.

What he said applies to me too...Canary is acting like I DIDNT make the phone call.
legendary
Activity: 938
Merit: 1000
LIR DEV
October 19, 2013, 05:14:04 PM
update...
eventually,each saturn freaked out on 0.90, so had to go back to 0.96...

How did they change?  Disabling and re-enabling cores?
the ol' 0 submitted on every line scrollling at 1000 lines a second.....lol
slowest saturn that was 255 ish is still holding on at 271
hero member
Activity: 854
Merit: 500
October 19, 2013, 05:06:11 PM
I wonder how my miner is enjoying Ohio, now that it has spent 2 days there.

Maybe it is seeing the sights and visiting a few museums
did you call DHL?

Yes it cleared customs and it will be here Monday or Tuesday, whenever they get around to it.
donator
Activity: 2352
Merit: 1060
between a rock and a block!
October 19, 2013, 05:04:05 PM
I wonder how my miner is enjoying Ohio, now that it has spent 2 days there.

Maybe it is seeing the sights and visiting a few museums

I compared your DHL info with mine and it may be that DHL has a LIFO buffer.
(Last In First Out) Just a few hours difference between customs.

Mine is hopefully enroute to Franklin Park Il. I wonder how many intermediate hubs lay in between?

I picture in my mind a tortoise loaded with miners crawling along the interstate right of way.

You may just be waiting for the tortoise that is heading to your locale to depart Cincy.
I hope so.



none between Franklin Park, IL and OH.

Anyone who's package hasn't left OH, needs to call DHL ASAP.  your package has customs/clearance issues.  don't wait, it ain't gonna start moving on it's own, you need to call DHL and see what else they need from you.
soy
legendary
Activity: 1428
Merit: 1013
October 19, 2013, 05:01:55 PM
update...
eventually,each saturn freaked out on 0.90, so had to go back to 0.96...

How did they change?  Disabling and re-enabling cores?
soy
legendary
Activity: 1428
Merit: 1013
October 19, 2013, 04:56:17 PM
BertMod 0.1 for 0.93 works well with firmware 0.90 and shows the Merc using 183 watts @ MHS av=134461.07 although  Slush now has the Merc at 143489.394MH/s.
sounds sweet soy..
the enablecore tweak is on 2 of 3 miners now, on 0.90 & the tweak
 the 3rd gets severe errors with 0.90, but it was the fastest one at 275 on 9.6, so there it stays.
the other two seem to be happily hashing at 275... so all 3 now at 275 is cool!

you have to power cycle the machine after applying the .bin file tho.

Didn't know that.  I'm going to let it run 24 hours as it is now then try 0.92 and see if the duplicate share issue is fixed.  So, BertMod works on all versions except 0.96?
legendary
Activity: 1066
Merit: 1098
October 19, 2013, 04:37:01 PM


Nice.  Guess I need to go read up on this cgi you speak of... Wink



I'll speed you up, but I am more of a hack than a polished programmer so rather just give tips to what I did





the fields that are already constantly being updated and outputted for use are:

STATUS=S,When=1382207241,Code=11,Msg=Summary,Description=cgminer 3.4.0|SUMMARY,
Elapsed=310,MHS av=450396.80,Found Blocks=0,Getworks=17,Accepted=102,Rejected=2,
Hardware Errors=7707,Utility=19.77,Discarded=26,Stale=0,Get Failures=0,Local Work=45860,
Remote Failures=0,Network Blocks=3,Total MH=139401753.5263,Work Utility=6997.99,
Difficulty Accepted=26112.00000000,Difficulty Rejected=512.00000000,Difficulty Stale=0.00000000,
Best Share=468225,Device Hardware%=17.5935,Device Rejected%=1.4183,Pool Rejected%=1.9231,Pool Stale%=0.0000|

You can use any of these but will have to change two files:


/www/tmpl/mining_stat.html_tmpl

that one will do the formatting and I make a new one and call it mining_stat.html_tmpl2



the one that does the real work is

/www/pages/cgi-bin/fetch_mining_stat.cgi


when you look inside these two files you will see how to add more fields and just stick with it's naming conventions to pass the values

since I changed the name of the other file, make sure to point to the new named one in this file like:

s!#%#ASIC_STATUS#%#!'"$asic_status"'!g' < /www/tmpl/mining_stat.html_tmpl2


so when you are all set with files in the two directories, you just do a simple
cp fetch_mining_stat.cgi.[newversion] fetch_mining_stat.cgi

then it doesn't disrupt anything since now it just goes to the other formatting file smoothly









Thanks! Smiley
sr. member
Activity: 462
Merit: 250
October 19, 2013, 04:33:34 PM


Nice.  Guess I need to go read up on this cgi you speak of... Wink



I'll speed you up, but I am more of a hack than a polished programmer so rather just give tips to what I did





the fields that are already constantly being updated and outputted for use are:

STATUS=S,When=1382207241,Code=11,Msg=Summary,Description=cgminer 3.4.0|SUMMARY,
Elapsed=310,MHS av=450396.80,Found Blocks=0,Getworks=17,Accepted=102,Rejected=2,
Hardware Errors=7707,Utility=19.77,Discarded=26,Stale=0,Get Failures=0,Local Work=45860,
Remote Failures=0,Network Blocks=3,Total MH=139401753.5263,Work Utility=6997.99,
Difficulty Accepted=26112.00000000,Difficulty Rejected=512.00000000,Difficulty Stale=0.00000000,
Best Share=468225,Device Hardware%=17.5935,Device Rejected%=1.4183,Pool Rejected%=1.9231,Pool Stale%=0.0000|

You can use any of these but will have to change two files:


/www/tmpl/mining_stat.html_tmpl

that one will do the formatting and I make a new one and call it mining_stat.html_tmpl2



the one that does the real work is

/www/pages/cgi-bin/fetch_mining_stat.cgi


when you look inside these two files you will see how to add more fields and just stick with it's naming conventions to pass the values

since I changed the name of the other file, make sure to point to the new named one in this file like:

s!#%#ASIC_STATUS#%#!'"$asic_status"'!g' < /www/tmpl/mining_stat.html_tmpl2


so when you are all set with files in the two directories, you just do a simple
cp fetch_mining_stat.cgi.[newversion] fetch_mining_stat.cgi

then it doesn't disrupt anything since now it just goes to the other formatting file smoothly


ps-save copies in /config since about everything else gets wiped on reboots






legendary
Activity: 938
Merit: 1000
LIR DEV
October 19, 2013, 04:19:31 PM
back to all 3 sats running 0.96, but they all hang around 275 now still  Smiley
same procedure on all 3 saturns...
1. flash 0.90 on the miner, reboot
2. flash enablecores.bin to the miner, POWER CYCLE the miner.
3. putty in and watch cgminer until it crashes, when it does, go to the gui page again, and flash 0.96 back onto the miner, apply, then POWER CYCLE the miner(turn off for 15 seconds via the power supply switch)
thats it.
i had 3 sats going at 255, 265, and 275....  now all at 270 - 275ish
sr. member
Activity: 251
Merit: 250
October 19, 2013, 04:18:45 PM
I hope it is best to use the latest firmware version available, as I dont have so much time to test all the firmware versions.

Wish it worked best with the first version, but hey at least I received as promised
hero member
Activity: 616
Merit: 500
October 19, 2013, 04:17:25 PM
Both running .96
One has a dead & a weak die

good miner:





bad miner:




I guess you've already tested it, but does it change anything if you put the slot configurarion of the faulty unit exactly like the working one?

Bitfury unit's performance may suffer if hashing boards are in the wrong slots. I'm wondering whether KnC units have similar issues.
legendary
Activity: 1456
Merit: 1018
HoneybadgerOfMoney.com Weed4bitcoin.com
October 19, 2013, 04:14:18 PM
Is there going to be a updated bertmod ?
The saturn i have running shows a nice 265 Gh on average so i do not want to mess with it too much
However it would be nice to have those extra stats showing some info about the errors and such
I read about bertmod and was mentioned you need the same version as you use on the miner
however mine is 0.96 and could not find a matching bertmod yet.

is your name Bronan as in = CONAN for Bros?  Its pretty catchy if so.
hero member
Activity: 774
Merit: 500
Lazy Lurker Reads Alot
October 19, 2013, 04:07:37 PM
Is there going to be a updated bertmod ?
The saturn i have running shows a nice 265 Gh on average so i do not want to mess with it too much
However it would be nice to have those extra stats showing some info about the errors and such
I read about bertmod and was mentioned you need the same version as you use on the miner
however mine is 0.96 and could not find a matching bertmod yet.
soy
legendary
Activity: 1428
Merit: 1013
October 19, 2013, 04:04:12 PM
so how to get the enablecores.bin to run on the slo sat?
i have it (the file) on my desktop...
using putty I imagine,...


Shut down, let sit a minute, unplug the ASIC modules at the modules (the PCIE power), restart and only BBB will be running, go to the Sat. web page, upgrade to 0.90, shutdown, let sit a minute, plug the ASIC modules back in, start, go to web page, run Upgrade with enablecores.bin.  At least that's how I did it.


Phoenix1969, I didn't mention that I generally don't just shut off the power to the miner.  Since I generally have putty open to cgminer, when I go to shut down I hit "q" in putty, quitting the screen of cgminer, then giving the command "halt" and since I'm logged in a root, the system shuts down in a more orderly manner than simply hitting a switch.  I do this because I've noticed with Raspberry Pi if I just kill power the program can get corrupted.  When you quit the screen and command halt, the first thing that's going to happen is putty will send a notice it has lost the connection.  That's when I kill the power.  Might not matter but why take a chance.
Jump to: