kaerf
|
|
November 17, 2013, 10:35:29 AM |
|
my results are for an 8 chip h-card that i made (just a modified 16 chip card). for my standard h-cards I used http://www.newark.com/abl-heatsinks/325ab1000b/heat-sink/dp/07WX4339 (needed to drill 2 holes), but I've actually ended up not heatsinking the majority of my hcards...didn't see all that much benefit (standard hcards would probably benefit most from good heatsinks on the reg...not an issue for your mod).
|
|
|
|
goxed
Legendary
Offline
Activity: 1946
Merit: 1006
Bitcoin / Crypto mining Hardware.
|
|
November 17, 2013, 10:47:30 AM |
|
my results are for an 8 chip h-card that i made (just a modified 16 chip card). for my standard h-cards I used http://www.newark.com/abl-heatsinks/325ab1000b/heat-sink/dp/07WX4339 (needed to drill 2 holes), but I've actually ended up not heatsinking the majority of my hcards...didn't see all that much benefit (standard hcards would probably benefit most from good heatsinks on the reg...not an issue for your mod). Aah okay the famous mini-H Actually I have also not seen major benefits with heatsinking h-card except for making the on-board regulator a bit stable. I am still trying to wrap my head around the issue of slower hash-rate at 0.9V. I was expecting somewhere in the ballpark of 40GH/s from 16chips. Cool so newark sells heatskinks as well. This is useful info. Thanks Btw did you make the boards through Oshpark? How soon was delivery?
|
Revewing Bitcoin / Crypto mining Hardware.
|
|
|
Asinine
Member
Offline
Activity: 62
Merit: 10
|
|
November 17, 2013, 03:50:04 PM |
|
Updating, I went out this morning so I could get an 8 GB card to put the V2.5 Image on: Noncerate: 459.189GH/s - 441 W at the wall. So, I think I am at the point where it is just tweaking now. IE: upgrading to the v3 chainminer, tinkering with the voltages, and shuffling the weaker cards to the last slots of the groups.
can you post the link to the V2.5 image? is that the latest version for v3 m boards and v2.2 h boards? Sorry, looking back at the page, I don't know where I got the V2.5 from... It may be the same as the second link on Buzz's page. https://bitcointalk.org/index.php?topic=287590.0Direct link to file: https://dl.dropboxusercontent.com/u/43738391/2013-Sep-27_BFSB_GM2.rar
|
|
|
|
cypherdoc
Legendary
Offline
Activity: 1764
Merit: 1002
|
|
November 17, 2013, 04:37:04 PM |
|
Updating, I went out this morning so I could get an 8 GB card to put the V2.5 Image on: Noncerate: 459.189GH/s - 441 W at the wall. So, I think I am at the point where it is just tweaking now. IE: upgrading to the v3 chainminer, tinkering with the voltages, and shuffling the weaker cards to the last slots of the groups.
can you post the link to the V2.5 image? is that the latest version for v3 m boards and v2.2 h boards? Sorry, looking back at the page, I don't know where I got the V2.5 from... It may be the same as the second link on Buzz's page. https://bitcointalk.org/index.php?topic=287590.0Direct link to file: https://dl.dropboxusercontent.com/u/43738391/2013-Sep-27_BFSB_GM2.rarthat's the version i've been running for a couple of days on my v3 rig but its been suspiciously slow. according to Dave's post up a few, it sounds like it's not the right one. i look forward to him putting the right one up today.
|
|
|
|
klondike_bar
Legendary
Offline
Activity: 2128
Merit: 1005
ASIC Wannabe
|
|
November 17, 2013, 04:48:22 PM |
|
1 AIfDSo 56 2.777 2.558 194 0 0 0 242 [0:0] 1 12 12 12 12 13 13 12 12 12 12 12 12 12 12 12 12 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 2 AIfDSo 55 2.477 2.494 173 1 0 0 236 [0:1] 37 11 11 11 11 11 10 10 11 11 11 11 11 11 10 11 11 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 3 AIfDSo 54 2.334 2.452 163 3 0 0 232 [0:2] 38 11 11 11 11 11 10 10 10 10 10 10 9 10 10 9 10 0 0 0 0 0 0 0 0 0 0 0 1 0 0 1 1 4 AIfDSo 56 2.620 2.336 183 1 0 0 221 [0:3] 2 12 11 11 11 11 11 11 11 11 12 11 12 12 12 12 12 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 5 AIfDSo 54 2.520 2.632 176 12 0 0 249 [0:4] 1 10 11 11 12 11 11 9 12 12 11 11 11 11 12 11 10 1 0 0 0 1 1 3 0 0 1 1 1 1 0 1 1 6 AIfDSo 54 2.362 2.706 165 14 0 0 256 [0:5] 5 9 10 11 11 10 11 11 11 10 10 11 11 10 10 9 10 2 1 1 1 2 0 0 0 1 1 0 0 1 1 2 1 7 AIfDSo 55 2.033 2.759 142 26 0 0 261 [0:6] 36 8 9 9 8 8 8 9 10 10 8 10 9 10 9 11 6 2 1 1 2 2 2 1 1 1 3 1 2 1 2 0 4 speed down 8 AIfDSo 54 2.849 2.706 199 1 0 0 256 [0:7] 2 12 12 12 12 12 12 12 13 13 13 13 12 13 13 13 12 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 9 AIfDSo 56 2.377 2.399 166 0 0 0 227 [0:8] 4 10 11 11 11 11 11 11 10 10 10 10 10 10 10 10 10 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 10 AIfDSo 56 2.448 2.399 171 0 0 0 227 [0:9] 6 10 10 10 11 11 11 11 11 11 11 11 11 11 11 10 10 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 11 AIfDSo 54 2.462 2.357 172 0 0 0 223 [0:A] 5 11 11 11 11 11 11 11 11 11 11 10 10 10 10 11 11 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 speed up 12 AIfDSo 54 2.248 2.378 157 0 0 0 225 [0:B] 26 10 10 10 10 10 10 10 10 10 10 10 10 10 9 9 9 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 13 AIfDSo 56 2.419 2.537 169 2 0 0 240 [0:C] 3 11 11 11 11 11 11 11 10 10 10 9 10 11 10 11 11 0 0 0 0 0 0 0 0 0 0 1 0 0 1 0 0 14 AIfDSo 56 2.591 2.431 181 2 0 0 230 [0:D] 0 12 12 11 11 11 11 11 10 11 10 11 12 12 12 12 12 0 0 0 0 0 0 0 1 0 1 0 0 0 0 0 0 15 AIfDSo 54 1.890 2.960 132 102 0 0 280 [0:E] 7 11 9 10 9 9 9 10 7 6 10 9 8 8 4 5 8 4 6 5 6 6 6 5 8 8 4 5 6 6 10 10 7 speed:824 noncerate[GH/s]:36.407 (2.427/chip) hashrate[GH/s]:38.104 good:2543 errors:164 spi-errors:0 miso-errors:0 jobs:375 (record[GH/s]:37.037) 0: 824 36.407 38.104 2543 164 0 0 Used a modded EOL card with 15 chips for this test. At 0.9Vlooks good! Push it to 0.94V and see what happens. IME, when the chips are between 53-55 tuning at higher voltage you get the best hashrate. ps: how is the temperature in the capacitor clusters? in the second revision they were taken out of the deisgn since they often were creating a lot of heat relative to the chips, and with little or no benefit
|
|
|
|
Asinine
Member
Offline
Activity: 62
Merit: 10
|
|
November 17, 2013, 05:33:38 PM |
|
that's the version i've been running for a couple of days on my v3 rig but its been suspiciously slow.
according to Dave's post up a few, it sounds like it's not the right one. i look forward to him putting the right one up today.
Going back through Dave's old posts, it looks like the major pieces are the v3 tuned Chainminer with improved timings and a button on the web page that allows you to shut the Pi down cleanly without having to SSH or jack into it. So, it might help with the speeds. When I first started up the V2, I had to enable all three of the stratum accounts on the settings page and gave them all the same account information. I remembered that vaguely from posts back from the August release. Before I did this, I was only getting about 130 GHash. I also had issues with the V2 starting up the rig at about half it's actual speed and it's full power draw even after configuring it. Giving the rig a couple of minutes to warm up, stopping, and starting the miner from the webpage seems to help. Straying off the topic now: For comparison, I was getting 360-400 GHash on BFG and am now getting 450-470 GHash on the legacy chain miner prior to tinkering with the individual card/chip settings. The pool was following the miner's output closely enough to not be a concern. One plus with BFG seems to be prioritized pool fail-over.
|
|
|
|
Doff
|
|
November 17, 2013, 05:33:49 PM |
|
https://i.imgur.com/puIkEZC.jpg1 AIfDSo 56 2.777 2.558 194 0 0 0 242 [0:0] 1 12 12 12 12 13 13 12 12 12 12 12 12 12 12 12 12 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 2 AIfDSo 55 2.477 2.494 173 1 0 0 236 [0:1] 37 11 11 11 11 11 10 10 11 11 11 11 11 11 10 11 11 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 3 AIfDSo 54 2.334 2.452 163 3 0 0 232 [0:2] 38 11 11 11 11 11 10 10 10 10 10 10 9 10 10 9 10 0 0 0 0 0 0 0 0 0 0 0 1 0 0 1 1 4 AIfDSo 56 2.620 2.336 183 1 0 0 221 [0:3] 2 12 11 11 11 11 11 11 11 11 12 11 12 12 12 12 12 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 5 AIfDSo 54 2.520 2.632 176 12 0 0 249 [0:4] 1 10 11 11 12 11 11 9 12 12 11 11 11 11 12 11 10 1 0 0 0 1 1 3 0 0 1 1 1 1 0 1 1 6 AIfDSo 54 2.362 2.706 165 14 0 0 256 [0:5] 5 9 10 11 11 10 11 11 11 10 10 11 11 10 10 9 10 2 1 1 1 2 0 0 0 1 1 0 0 1 1 2 1 7 AIfDSo 55 2.033 2.759 142 26 0 0 261 [0:6] 36 8 9 9 8 8 8 9 10 10 8 10 9 10 9 11 6 2 1 1 2 2 2 1 1 1 3 1 2 1 2 0 4 speed down 8 AIfDSo 54 2.849 2.706 199 1 0 0 256 [0:7] 2 12 12 12 12 12 12 12 13 13 13 13 12 13 13 13 12 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 9 AIfDSo 56 2.377 2.399 166 0 0 0 227 [0:8] 4 10 11 11 11 11 11 11 10 10 10 10 10 10 10 10 10 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 10 AIfDSo 56 2.448 2.399 171 0 0 0 227 [0:9] 6 10 10 10 11 11 11 11 11 11 11 11 11 11 11 10 10 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 11 AIfDSo 54 2.462 2.357 172 0 0 0 223 [0:A] 5 11 11 11 11 11 11 11 11 11 11 10 10 10 10 11 11 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 speed up 12 AIfDSo 54 2.248 2.378 157 0 0 0 225 [0:B] 26 10 10 10 10 10 10 10 10 10 10 10 10 10 9 9 9 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 13 AIfDSo 56 2.419 2.537 169 2 0 0 240 [0:C] 3 11 11 11 11 11 11 11 10 10 10 9 10 11 10 11 11 0 0 0 0 0 0 0 0 0 0 1 0 0 1 0 0 14 AIfDSo 56 2.591 2.431 181 2 0 0 230 [0:D] 0 12 12 11 11 11 11 11 10 11 10 11 12 12 12 12 12 0 0 0 0 0 0 0 1 0 1 0 0 0 0 0 0 15 AIfDSo 54 1.890 2.960 132 102 0 0 280 [0:E] 7 11 9 10 9 9 9 10 7 6 10 9 8 8 4 5 8 4 6 5 6 6 6 5 8 8 4 5 6 6 10 10 7 speed:824 noncerate[GH/s]:36.407 (2.427/chip) hashrate[GH/s]:38.104 good:2543 errors:164 spi-errors:0 miso-errors:0 jobs:375 (record[GH/s]:37.037) 0: 824 36.407 38.104 2543 164 0 0 Used a modded EOL card with 15 chips for this test. At 0.9V sexy... i'm a little surprised at your chip speeds, though. At 0.9v, I'd found that dropping to 53 was most stable. i think these numbers are for voltage around 0.9v (can't remember for sure and too lazy to test right now) 1 AIfDSo 53 2.634 2.854 184 0 0 0 270 [0:0] 0 11 11 11 12 12 12 12 12 12 12 12 11 11 11 11 11 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 2 AIfDSo 53 2.491 2.759 174 5 0 0 261 [0:1] 0 11 10 11 11 11 11 11 10 11 11 11 11 11 11 11 11 0 1 0 0 0 0 0 1 0 0 0 0 1 1 1 0 3 AIfDSo 57 2.763 2.558 193 0 0 0 242 [0:2] 0 12 12 12 12 12 12 12 12 12 12 12 12 12 13 12 12 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 4 AIfDSo 54 2.777 2.685 194 0 0 0 254 [0:3] 0 12 12 12 12 12 12 12 12 12 12 12 12 12 12 13 13 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 speed up 5 AIfDSo 53 2.921 2.907 204 1 0 0 275 [0:4] 0 12 12 13 13 13 12 13 13 13 13 13 13 13 13 13 12 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 6 AIfDSo 54 2.663 2.907 186 15 0 0 275 [0:5] 0 12 12 13 12 12 11 11 10 11 11 13 13 12 12 11 10 1 1 0 0 0 1 1 2 1 1 0 0 1 1 2 3 speed down 7 AIfDSo 53 2.849 2.854 199 2 0 0 270 [0:6] 0 12 13 13 13 12 12 12 12 12 12 12 13 13 13 13 12 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 8 AIfDSo 52 2.606 2.621 182 0 0 0 248 [0:7] 0 11 11 11 11 11 11 11 11 11 12 12 12 12 12 12 11 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 speed:429 noncerate[GH/s]:21.704 (2.713/chip) hashrate[GH/s]:22.143 good:1516 errors:23 spi-errors:0 miso-errors:0 jobs:296 (record[GH/s]:23.422) 0: 429 21.704 22.143 1516 23 0 0
(of course, chainminer autotuned a few chips, but initially 53 gave me the least errors) I am very surprised as well. I has hoping for aforementioned per chip speeds. Are these results from a 16chip H-card? Any beefy heatsinks? Is the only way to notch your speed up by making a best.cnf file? Also whats unlcear to me is when I make that best cnf file what all should be in it? Example: 1 AIfDSo 53 <-- is that the only line you need for each chip or should it be everything that the .stat.log has minus the very bottom readout?
|
|
|
|
kaerf
|
|
November 17, 2013, 05:55:42 PM |
|
Aah okay the famous mini-H Actually I have also not seen major benefits with heatsinking h-card except for making the on-board regulator a bit stable. I am still trying to wrap my head around the issue of slower hash-rate at 0.9V. I was expecting somewhere in the ballpark of 40GH/s from 16chips. Cool so newark sells heatskinks as well. This is useful info. Thanks Btw did you make the boards through Oshpark? How soon was delivery? yeah, i went through oshpark. i think my 3 boards were delivered in about 2-2.5 weeks.
|
|
|
|
goxed
Legendary
Offline
Activity: 1946
Merit: 1006
Bitcoin / Crypto mining Hardware.
|
|
November 17, 2013, 08:39:39 PM Last edit: November 17, 2013, 09:30:03 PM by goxed |
|
Aah okay the famous mini-H Actually I have also not seen major benefits with heatsinking h-card except for making the on-board regulator a bit stable. I am still trying to wrap my head around the issue of slower hash-rate at 0.9V. I was expecting somewhere in the ballpark of 40GH/s from 16chips. Cool so newark sells heatskinks as well. This is useful info. Thanks Btw did you make the boards through Oshpark? How soon was delivery? yeah, i went through oshpark. i think my 3 boards were delivered in about 2-2.5 weeks. Thanks for the reply man. BTW did you solder the chips by hand? If yes, which tool did you use?
|
Revewing Bitcoin / Crypto mining Hardware.
|
|
|
Keefe
|
|
November 17, 2013, 08:54:45 PM |
|
Is the only way to notch your speed up by making a best.cnf file? Also whats unlcear to me is when I make that best cnf file what all should be in it? Example: 1 AIfDSo 53 <-- is that the only line you need for each chip or should it be everything that the .stat.log has minus the very bottom readout?
If the autotuner isn't giving you the results you expect, you can use best.cnf to turn it off (lowercase a) and manually set speeds. You don't need to put all the stuff from .stat.log into best.cnf, but it's not a problem either. It only reads the first 3 tab-separated columns and ignores the rest.
|
|
|
|
spiccioli
Legendary
Offline
Activity: 1379
Merit: 1003
nec sine labore
|
|
November 17, 2013, 09:15:43 PM |
|
Is the only way to notch your speed up by making a best.cnf file? Also whats unlcear to me is when I make that best cnf file what all should be in it? Example: 1 AIfDSo 53 <-- is that the only line you need for each chip or should it be everything that the .stat.log has minus the very bottom readout?
If the autotuner isn't giving you the results you expect, you can use best.cnf to turn it off (lowercase a) and manually set speeds. You don't need to put all the stuff from .stat.log into best.cnf, but it's not a problem either. It only reads the first 3 tab-separated columns and ignores the rest. Here there is a good way to create a starting best.conf https://bitcointalk.org/index.php?topic=251966.msg3565601#msg3565601I've not tested it, but it seem ok just reading it spiccioli ps. btw, it's yours
|
|
|
|
Doff
|
|
November 17, 2013, 09:21:26 PM |
|
Yeah, I should use that and just change it to turn A on and the speed to grab 53 instead of 52, I may even attempt 54 since my cooling is pretty good. Thanks Spiccioli, and Keefe.
|
|
|
|
kaerf
|
|
November 17, 2013, 11:03:41 PM |
|
Aah okay the famous mini-H Actually I have also not seen major benefits with heatsinking h-card except for making the on-board regulator a bit stable. I am still trying to wrap my head around the issue of slower hash-rate at 0.9V. I was expecting somewhere in the ballpark of 40GH/s from 16chips. Cool so newark sells heatskinks as well. This is useful info. Thanks Btw did you make the boards through Oshpark? How soon was delivery? yeah, i went through oshpark. i think my 3 boards were delivered in about 2-2.5 weeks. Thanks for the reply man. BTW did you solder the chips by hand? If yes, which tool did you use? There's a pic of the tool I made here: https://bitcointalk.org/index.php?topic=228677.msg3259455#msg3259455I basically copied http://vpapanik.blogspot.com/2012/11/low-budget-manual-pick-place.html. Then I used a toaster oven to reflow...out of 3 cards, I only managed to get one to work...one was overheated, the other probably had some bad connections.
|
|
|
|
cypherdoc
Legendary
Offline
Activity: 1764
Merit: 1002
|
|
November 17, 2013, 11:20:54 PM |
|
anyone hear from Dave about the latest v3 image?
|
|
|
|
Asinine
Member
Offline
Activity: 62
Merit: 10
|
|
November 18, 2013, 12:58:26 AM |
|
anyone hear from Dave about the latest v3 image?
...not since he last said he would have it up tomorrow. Does anyone know what would count as an RMA'able card when the card partially works?
|
|
|
|
cypherdoc
Legendary
Offline
Activity: 1764
Merit: 1002
|
|
November 18, 2013, 04:40:48 PM |
|
come on Dave!
i need that v3 image!
anyone got a phone # for them?
|
|
|
|
shade88
Member
Offline
Activity: 119
Merit: 10
|
|
November 18, 2013, 11:53:43 PM |
|
come on Dave!
i need that v3 image!
anyone got a phone # for them?
Someone with v3 board please image it for him. My upload stream is so slow it would take days to upload it.
|
|
|
|
goxed
Legendary
Offline
Activity: 1946
Merit: 1006
Bitcoin / Crypto mining Hardware.
|
|
November 19, 2013, 12:08:08 AM |
|
come on Dave!
i need that v3 image!
anyone got a phone # for them?
Someone with v3 board please image it for him. My upload stream is so slow it would take days to upload it. Can v2 image work as a stopgap with v3 boards? Or is there some fundamental change ?
|
Revewing Bitcoin / Crypto mining Hardware.
|
|
|
tacotime
Legendary
Offline
Activity: 1484
Merit: 1005
|
|
November 19, 2013, 12:15:35 AM |
|
come on Dave!
i need that v3 image!
anyone got a phone # for them?
Someone with v3 board please image it for him. My upload stream is so slow it would take days to upload it. am upload now
|
XMR: 44GBHzv6ZyQdJkjqZje6KLZ3xSyN1hBSFAnLP6EAqJtCRVzMzZmeXTC2AHKDS9aEDTRKmo6a6o9r9j86pYfhCWDkKjbtcns
|
|
|
tacotime
Legendary
Offline
Activity: 1484
Merit: 1005
|
|
November 19, 2013, 12:32:06 AM |
|
Write with imageusb V3 image file12HWFAsv1ojTuw5FzoP9T3SnyjZew5hFDL
|
XMR: 44GBHzv6ZyQdJkjqZje6KLZ3xSyN1hBSFAnLP6EAqJtCRVzMzZmeXTC2AHKDS9aEDTRKmo6a6o9r9j86pYfhCWDkKjbtcns
|
|
|
|