intron
Sr. Member
![*](https://bitcointalk.org/Themes/custom1/images/star.gif) ![*](https://bitcointalk.org/Themes/custom1/images/star.gif) ![*](https://bitcointalk.org/Themes/custom1/images/star.gif)
Offline
Activity: 427
Merit: 251
- electronics design|embedded software|verilog -
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
November 23, 2013, 07:05:55 PM |
|
BTW,
having some hardware way to isolate "broken" chips without having to cut traces on the board would be a good thing.
spiccioli
pretty sure it was demonstrated that shorting the jumpers by the faulty chip causes it to simply by skipped over the same was as cutting the traces would achieve *i think thats what i read* Also it might be needed to remove the defective chip. You can use an hot air gun for that. Be careful not to heat up the rest of the board to much though. intron
|
|
|
|
GH
Member
![*](https://bitcointalk.org/Themes/custom1/images/star.gif)
Offline
Activity: 117
Merit: 10
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
November 23, 2013, 10:20:02 PM Last edit: November 24, 2013, 08:41:29 AM by GH |
|
Hello all,
recently had my very first "oops" experience with my bitfury miner. Was mining absolutely(!) rock stable with cron miner restarts 2 hourly at 225-230 Gh/s with lately 6 pro-modded(*) h-board for two month now. Suddenly only the first 3 h-boards (0/4/5) had full useable nonces, 8/9/C were hashing at normal speed but did produce absolutely 0% useable nonces. No exceptional SPI nor MISO errors were detected. Miner restarts (which work even in cases of a "hot-plug") and even complete reboot did not help. Only a halt with power-cycle restored full function. Why? Any similar experience or explainations, anyone?
Thanks, Gerald
*) 20x precision spindle trimmers for overvolting, additional power cabling to compensate for voltage drop over PCBs and even voltage for all chips, no heatsinks, 7 fans.
|
|
|
|
spiccioli
Legendary
Offline
Activity: 1378
Merit: 1003
nec sine labore
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
November 23, 2013, 10:26:29 PM |
|
BTW,
having some hardware way to isolate "broken" chips without having to cut traces on the board would be a good thing.
spiccioli
pretty sure it was demonstrated that shorting the jumpers by the faulty chip causes it to simply by skipped over the same was as cutting the traces would achieve *i think thats what i read* Hi, all of them or a particular couple? spiccioli
|
|
|
|
klondike_bar
Legendary
Offline
Activity: 2128
Merit: 1005
ASIC Wannabe
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
November 23, 2013, 11:33:41 PM |
|
BTW, having some hardware way to isolate "broken" chips without having to cut traces on the board would be a good thing. spiccioli
pretty sure it was demonstrated that shorting the jumpers by the faulty chip causes it to simply by skipped over the same was as cutting the traces would achieve *i think thats what i read* all of them or a particular couple? spiccioli if you look at the chip's traces, there are two sets: 1 runs to the capacitors (or where capacitors were in rev1) and the other set runs to vias to another layer of the PCB. These have very small silver pads that pair them up, and i belive are responsible for the MISO commands and other chaining functions. if the small pairs of pads are shorted, the chip is no longer included in the chain https://bitcointalk.org/index.php?topic=250249.1040step 1: short the small jumper pads and see how it works step 2: if problem persists, either remove the chip or cut the traces (preferably a little better than how some have done it on here lol) i know there is more info elsewhere but cant find it offhand
|
|
|
|
GH
Member
![*](https://bitcointalk.org/Themes/custom1/images/star.gif)
Offline
Activity: 117
Merit: 10
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
November 23, 2013, 11:37:36 PM Last edit: November 24, 2013, 08:40:32 AM by GH |
|
Hello all,
recently had my very first "oops" experience with my bitfury miner. Was mining absolutely(!) rock stable with cron miner restarts 2 hourly at 225-230 Gh/s with lately 6 pro-modded(*) h-board for two month now. Suddenly only the first 3 h-boards (0/4/5) had full useable nonces, 8/9/C were hashing at normal speed but did produce absolutely 0% useable nonces. No exceptional SPI nor MISO errors were detected. Miner restarts (which work even in cases of a "hot-plug") and even complete reboot did not help. Only a halt with power-cycle restored full function. Why? Any similar experience or explainations, anyone?
Thanks, Gerald
*) 20x precision spindle trimmers for overvolting, additional power cabling to compensate for voltage drop over PCBs and even voltage for all chips, no heatsinks, 7 fans.
Hmmm. Now one of my 96 friends is missing in action... 89 aIfDSo 54 2.792 2.600 195 1 0 0 246 [C:8] 461 13 13 12 12 12 12 12 12 12 12 12 12 11 12 13 13 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 90 aIfDSo 54 0.000 2.600 0 3375 0 0 246 [C:9] 756 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 338 338 337 337 0 0 337 337 337 0 0 338 338 338 0 0 91 aIfDSo 54 2.319 2.484 162 0 0 0 235 [C:A] 496 10 11 11 11 10 10 10 10 10 10 10 10 10 10 10 9 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
Edit: Hmmm 2: found his way home himself this morning, a little wounded... 89 aIfDSo 54 2.649 2.600 185 0 0 0 246 [C:8] 275 12 12 12 11 11 11 11 11 11 11 12 12 12 12 12 12 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 90 aIfDSo 54 2.534 2.600 177 8 0 0 246 [C:9] 301 10 11 11 11 12 12 10 11 12 12 11 11 10 11 11 11 1 0 0 0 0 0 2 1 0 0 1 1 2 0 0 0 91 aIfDSo 54 2.534 2.473 177 0 0 0 234 [C:A] 332 12 11 11 11 11 11 11 11 11 11 11 11 11 11 11 11 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
|
|
|
|
Trupik
Member
![*](https://bitcointalk.org/Themes/custom1/images/star.gif)
Offline
Activity: 136
Merit: 13
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
November 24, 2013, 08:33:32 AM |
|
Hello all,
recently had my very first "oops" experience with my bitfury miner. Was mining absolutely(!) rock stable with cron miner restarts 2 hourly at 225-230 Gh/s with lately 6 pro-modded(*) h-board for two month now. Suddenly only the first 3 h-boards (0/4/5) had full useable nonces, 8/9/C were hashing at normal speed but did produce absolutely 0% useable nonces. No exceptional SPI nor MISO errors were detected. Miner restarts (which work even in cases of a "hot-plug") and even complete reboot did not help. Only a halt with power-cycle restored full function. Why? Any similar experience or explainations, anyone?
Thanks, Gerald
*) 20x precision spindle trimmers for overvolting, additional power cabling to compensate for voltage drop over PCBs and even voltage for all chips, not heatsinks, 7 fans.
I do have similar experience. About once a week, the noncerate reported by the pool drops to 15-40% of former value. The bitfury itself reports full noncerate as if nothing happened. I have to log in and reboot the RaspberryPi ( sudo shutdown -r now), then it works for another week or so. I was thinking about a watchdog script. Any ideas on how to do it?
|
|
|
|
GH
Member
![*](https://bitcointalk.org/Themes/custom1/images/star.gif)
Offline
Activity: 117
Merit: 10
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
November 24, 2013, 09:23:03 AM Last edit: November 24, 2013, 05:05:22 PM by GH |
|
Hello all,
recently had my very first "oops" experience with my bitfury miner. Was mining absolutely(!) rock stable with cron miner restarts 2 hourly at 225-230 Gh/s with lately 6 pro-modded(*) h-board for two month now. Suddenly only the first 3 h-boards (0/4/5) had full useable nonces, 8/9/C were hashing at normal speed but did produce absolutely 0% useable nonces. No exceptional SPI nor MISO errors were detected. Miner restarts (which work even in cases of a "hot-plug") and even complete reboot did not help. Only a halt with power-cycle restored full function. Why? Any similar experience or explainations, anyone?
Thanks, Gerald
*) 20x precision spindle trimmers for overvolting, additional power cabling to compensate for voltage drop over PCBs and even voltage for all chips, not heatsinks, 7 fans.
I do have similar experience. About once a week, the noncerate reported by the pool drops to 15-40% of former value. The bitfury itself reports full noncerate as if nothing happened. I have to log in and reboot the RaspberryPi ( sudo shutdown -r now), then it works for another week or so. I was thinking about a watchdog script. Any ideas on how to do it? Thanks for the feedback, but I think this is a different problem, especially because a reboot without power cycle did not cure the problem for me. In your case I would more think of an issue with the stratum proxy and the pool. Did you try to restart the stratum proxy? I would check the queue length in /run/shm/.putstat.log. If it is much greater than 0, it could be a bandwitdh upstream issue or high CPU load of the RasPi if difficulty is set quite low.
|
|
|
|
spiccioli
Legendary
Offline
Activity: 1378
Merit: 1003
nec sine labore
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
November 24, 2013, 08:55:36 PM |
|
if you look at the chip's traces, there are two sets: 1 runs to the capacitors (or where capacitors were in rev1) and the other set runs to vias to another layer of the PCB. These have very small silver pads that pair them up, and i belive are responsible for the MISO commands and other chaining functions. if the small pairs of pads are shorted, the chip is no longer included in the chain https://bitcointalk.org/index.php?topic=250249.1040step 1: short the small jumper pads and see how it works step 2: if problem persists, either remove the chip or cut the traces (preferably a little better than how some have done it on here lol) i know there is more info elsewhere but cant find it offhand Hi klondike_bar, thanks, I'll try step 1 tomorrow and then I'll report back spiccioli
|
|
|
|
rammy2k2
Legendary
Offline
Activity: 1974
Merit: 1003
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
November 26, 2013, 10:05:17 AM |
|
hey punin, any news from those tests ?
|
|
|
|
fractal02
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
November 26, 2013, 08:31:20 PM |
|
hey punin, any news from those tests ?
Pretty long tests... ![Undecided](https://bitcointalk.org/Smileys/default/undecided.gif) Any news about hardware Punin ? Give us some timeline. ![Wink](https://bitcointalk.org/Smileys/default/wink.gif)
|
|
|
|
Meizirkki
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
November 26, 2013, 08:40:21 PM |
|
Every passing day lowers the price people are willing to pay for them miners ![Wink](https://bitcointalk.org/Smileys/default/wink.gif)
|
|
|
|
rammy2k2
Legendary
Offline
Activity: 1974
Merit: 1003
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
November 28, 2013, 12:59:28 AM |
|
punin, still alive?
|
|
|
|
shade88
Member
![*](https://bitcointalk.org/Themes/custom1/images/star.gif)
Offline
Activity: 119
Merit: 10
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
November 28, 2013, 01:01:36 AM |
|
Must be busy cutting heatsink ![Cheesy](https://bitcointalk.org/Smileys/default/cheesy.gif)
|
|
|
|
xstr8guy
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
November 28, 2013, 06:31:55 AM |
|
Every passing day lowers the price people are willing to pay for them miners ![Wink](https://bitcointalk.org/Smileys/default/wink.gif) No, quite the opposite. Prices for mining hardware in the resale market have more than doubled in the past two weeks. We won't see another price drop unless bitcoin crashes or the other vendors actually start shipping preorders and the difficulty rises sharply again. Punin and Dave (MBP) are smart men. They know this and won't be selling H boards for less than the original price anytime soon. In fact, I suspect they may be self-mining right now and have no intention of selling. We probably won't get the opportunity to buy new boards until it's again marginally profitable (at best) after bitcoin crashes.
|
|
|
|
rupy
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
November 28, 2013, 10:12:29 AM |
|
Yes, I think you are right, ASIC technology will be very valuable and rare if the hyberbolic price continues since it's smarter to mine with it than sell it. All this talk about BTC ROI is just retarded, having a machine that generates a salary is not!
|
BANKBOOK GWT Wallet & no-FIAT Billing API
|
|
|
rammy2k2
Legendary
Offline
Activity: 1974
Merit: 1003
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
November 29, 2013, 06:16:45 PM |
|
hello ? anyone there ?
|
|
|
|
Micky25
Legendary
Offline
Activity: 974
Merit: 1000
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
November 29, 2013, 06:31:35 PM |
|
closed due to sudden wealth. Who are we to blame? We got our share ![Smiley](https://bitcointalk.org/Smileys/default/smiley.gif)
|
|
|
|
fractal02
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
November 29, 2013, 06:44:19 PM |
|
Strange... Punin last activity ==> Last Active: Today at 05:38:53 PM But no words for us. ![Huh](https://bitcointalk.org/Smileys/default/huh.gif)
|
|
|
|
philipma1957
Legendary
Online
Activity: 4228
Merit: 8394
'The right to privacy matters'
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
November 29, 2013, 06:50:59 PM |
|
Yes, I think you are right, ASIC technology will be very valuable and rare if the hyberbolic price continues since it's smarter to mine with it than sell it. All this talk about BTC ROI is just retarded, having a machine that generates a salary is not!
btc roi means nothing. fiat roi is the goal. My piece of a knc has returned 342 usd for a 78 usd share . it has returned .39btc for a .63btc investment. come tax time the IRS does not give 2 or 3 shits about btc return but they do care about usd return.
|
|
|
|
ricolein
Member
![*](https://bitcointalk.org/Themes/custom1/images/star.gif)
Offline
Activity: 85
Merit: 10
|
![](https://bitcointalk.org/Themes/custom1/images/post/xx.gif) |
November 30, 2013, 03:49:20 PM |
|
btc roi means nothing. fiat roi is the goal.
[...]
it has returned .39btc for a .63btc investment.
Fiat roi is the gaol? Really? For your invested .63btc you would get more fiat than for the earned .39btc.... Ciao Enrico
|
|
|
|
|