|
Barn,
I appreciate your being open about any issues. It is good that this issue was detected before the boards were made, as defect rectification later would have been much more difficult. You have my support on this. Probably need to put this episode into the 'lessons learnt' for any future batches.
Cheers
Kabopar
|
|
|
|
How about to replace the restistor with an Ω:.85 instead of the pencil mod ?? Would that work to ??
Or what's the "best/savest" restistor to use and get the fastest speed ??
It would not be a ".85" but more likely a 10k instead of the present 22K, as per Beastlymac's earlier comment "from what the EE told me you want to replace the 22K resistor with a 10K one". (or maybe an added 18k in parallel with the present 22k to achieve the same 10k). The 0.85K reading during the 'pencil mod' is the observed 'in-circuit' resistance. I suspect that de-soldering the current SMD 22k resistor and replacing it with 10k is a high risk task that most of us do not have the tools and skills to perform. At the same time a few pencil lines are low risk and reversible... Cheers Or a 5K to 15K variable? set at nominal 10K for fine tuning? Probably a good idea for the next batch, if the cost impact is not too high. The 'pencil mod' is simple and low risk, so far seems to be quite stable after about 14 hours. |
|
|
|
How about to replace the restistor with an Ω:.85 instead of the pencil mod ?? Would that work to ??
Or what's the "best/savest" restistor to use and get the fastest speed ??
It would not be a ".85" but more likely a 10k instead of the present 22K, as per Beastlymac's earlier comment "from what the EE told me you want to replace the 22K resistor with a 10K one". (or maybe an added 18k in parallel with the present 22k to achieve the same 10k). The 0.85K reading during the 'pencil mod' is the observed 'in-circuit' resistance. I suspect that de-soldering the current SMD 22k resistor and replacing it with 10k is a high risk task that most of us do not have the tools and skills to perform. At the same time a few pencil lines are low risk and reversible... Cheers |
|
|
|
Check the posts in the support thread for BF ( https://bitcointalk.org/index.php?topic=319419.440), there is a simple 'pencil mod' that can increase the hash rate. I'm now running at abut 2.51 GH/sec (accepted shares by the pool) and others got higher. Cheers |
|
|
|
now there is a series of quick ones, so don't miss out....
Payout is stuck again, probably because of Block 269017 (#20797 on Slush's count) which seems to have no transaction fees. I know that it will eventually happen, but it is annoying still....
True but man, if we had 2 - 12 hour blocks that's one day. I can only see a single 12 hour block in the last 24 hours (the previous extra long one of 13 hours was on 8 Nov). It's not the largest pool, and luck or randomness or whatever you want to call it has an impact, but it averages out on the long run. |
|
|
|
|
now there is a series of quick ones, so don't miss out....
Payout is stuck again, probably because of Block 269017 (#20797 on Slush's count) which seems to have no transaction fees. I know that it will eventually happen, but it is annoying still....
|
|
|
|
2 of these are pencil modded at different resistances . Been running for 5.5 hrs now with a fan set to low blowing on them - cool to the touch USING BFGminer 3.2 because it shows the serial # Ill mod the other 3 and check back later so you can see the differances  Here are my results, with R15 resistance at 0.89K (in-circuit) after about 9 hours  The pool confirms these results (the 3rd number) The heatsink is a bit warm, but nothing to worry about Thanks to all those that worked this out  |
|
|
|
its going to be summer over here soon, I expect mine to fry  Thanks again for your suggestion. Here is the result after about 15 minutes  Clearly better than the 2.26 GH/sec that it was getting originally. Cheers |
|
|
|
Kabopar,
I found drawing a line along the side was easiest, one to two lines did it for me.
my initial scribbling over the top was not necessary :-)
i'd measure the resistance again after, I'd not let it go below 0.84.
you might need the eraser and a couple of attempts to get the right result.
Thanks for this suggestion, I just tried it and it does work, finally the resistance moved from 0.97K to 0.94K. after 4 minutes I'm getting 2.40/2.38/2.33 (bfgminer 3.5.1), which is better than anything this BF produced before. So far there is no noticeable heating, but it could be too early to tell. (It's going to be summer here soon too, hopefully the bushfires will not be ignited by overheating BlueFuries... Cheers |
|
|
|
|
FrontLineFodder,
I also measured 0.97K in circuit. The surface of R15 on my BF is very smooth, and the HB lead simply doesn't stick to it, so it the resistance stays unchanged... I'll keep trying...
Your results are encouraging, even if you went a bit 'over', causing the heatsink to dislocate
Cheers
|
|
|
|
Anyone know how to clear my com ports out. Every time I switch from Zadig to Alpha driver it keeps ahold of the old com port and uses a new one. They don't show up in the Device manager not even when I unhide  You could try the device manager cleanup instructions in https://bitcointalk.org/index.php?topic=300638.0It helped me with multiple Com port settings for BE's. Cheers |
|
|
|
FrontLineFodder - nice work. It looks like it is hotter after the mod, as you'd expect...
Can you please clarify which resistor was 'pencil modded' and any other hints/pics?
Cheers
As Before: All warranties void perform at your own risk.My view: these are a toy, I dont care much if I break them, do you ? R15 was modded, I used what I had around, looked like HB lead in a pacer automatic pencil. Not sure if it means anything, I threw the multimeter to measure the resistance across the resistor 'in circuit'. measured 0.97K ohms, after the mod it measures about 0.86Kohms. It's been running for over 13 hours with the same speeds and error rate as I posted last night. As my pics above show, make sure you have adequate cooling. I'd like to get a USB current meter (like: http://gizmodo.com/usb-power-meter-tells-you-which-devices-will-kill-your-514285397 ) to see if there is an increase in power draw from the hub. anybody got reccommendations ? Thanks for the info. I tried but the in-circuit resistance stays at 0.97K, no matter how many HB lines I make. So there was no difference in hashing either.... Did you need to remove any coating from the resistor? (I don't think that they would apply conformal coating to a USB miner). Cheers |
|
|
|
|
FrontLineFodder - nice work. It looks like it is hotter after the mod, as you'd expect...
Can you please clarify which resistor was 'pencil modded' and any other hints/pics?
Cheers
|
|
|
|
Pencil mod, in case anybody else is stupid enough to do it. Before Pencil Mod: after 22 hours of hashing: BPM 0: | 1.99/ 1.99/ 1.83Gh/s | A: 6507 R: 56+0(.81%) HW:1182/3.3% after Pencil Mod: 3 hours of hashing BPM 0: | 2.59/ 2.59/ 2.49Gh/s | A:1242 R: 6+0(.48%) HW: 371/4.5% My view: these are a toy, I dont care much if I break them, do you ? Nice work! Promising results how did the cooling hold up? You pencil moded r22? I tried on one of mine and no change, did you have a special pencil? I just used a #2 like i have on other things before. Per Beastlymac's and others earlier posts it's probably R15 which is supposed to be modified from 22k (says 223 on the device). R15 is next to IC2, close to the PRG switch. @FrontLineFodder - can you confirm that this is the component that you 'penciled'? https://bitcointalk.org/index.php?topic=319419.340Cheers |
|
|
|
I have an unconfirmed payment from 11/7 and a negative balance with no double payment visible.  On the Slush web site https://mining.bitcoin.cz/accounts/profile/ it says " Quick updates 08.11.2013 Some recent payouts didn't go thru bitcoin network because of unsatisfied transaction fees. We're finding a soluton and all payouts will be processed soon" Hopefully they will fix this soon, or try to contact Slush support email or the Live Support. In my case the double payments were not consecutive. looking at my wallet transactions, stats (I copy and keep my Slush stats) and the payout history I managed to figure it out more or less... |
|
|
|
Compensation from Beastly can go into the address in my sig (same as the one in our emails from other orders)
The original compensation is for the delay in shipping. There should be some other compo due to performance being lower than specified (most people had 2.1-2.3 GH/Sec, compared to the specified 2.6+/-10%) see https://bitcointalk.org/index.php?topic=299295.0 |
|
|
|
My "confirmed reward" has gone negative.... Slush, do I owe you any bitcoins?
Confirmed reward: -0.01298504 BTC Your confirmed balance from valid blocks
I didn't even think this was possible and am myself curious as to whats up. Edit: Anyone able to help with: https://bitcointalk.org/index.php?topic=327410.0The mystery was partially solved, as my account page now contains this message: Notice: The last payment to you was sent twice by a mistake. This has caused that your confirmed reward is temporarily below 0 BTC. We're sorry for the inconvenience. Strange, but I hope that it is correct. Partially solved ? Did you confirm a two times payment to your wallet ? I found 2 identical payments, and tracing back the block confirmation to payments over the last few days, Slush's claim looks correct. Still strange that it could happen (aren't these calculations supposed to happen automatically and not 'handrolically'?!!!). So Slush is honest... but more order is needed.... Cheers I don't have 2 identical payments. with these long blocks it would be quite noticeable with only 1 payment per day. so I dont see why slush is taking btc back You might contact Slush support at support@bitcoin.cz (per the FAQ). Tracing the actual confirmed blocks to payments is tricky. |
|
|
|
My "confirmed reward" has gone negative.... Slush, do I owe you any bitcoins?
Confirmed reward: -0.01298504 BTC Your confirmed balance from valid blocks
I didn't even think this was possible and am myself curious as to whats up. Edit: Anyone able to help with: https://bitcointalk.org/index.php?topic=327410.0The mystery was partially solved, as my account page now contains this message: Notice: The last payment to you was sent twice by a mistake. This has caused that your confirmed reward is temporarily below 0 BTC. We're sorry for the inconvenience. Strange, but I hope that it is correct. Partially solved ? Did you confirm a two times payment to your wallet ? I found 2 identical payments, and tracing back the block confirmation to payments over the last few days, Slush's claim looks correct. Still strange that it could happen (aren't these calculations supposed to happen automatically and not 'handrolically'?!!!). So Slush is honest... but more order is needed.... Cheers |
|
|
|
|
Show Posts
