as scientific as "likely" ....cheers  If we are to be scientists we should perhaps quote guy who said that epoxy is not necessary to re-secure the heatsinks as the glue was there to hold them in place during shipping but I digress...just trying to give SAFE advice to someone who clearly has no idea what they are doing (OP) SAFE advice could be just an opinion without proper scientific evidence backing it up. The evidence points that the voltage for some chips are not at their desired levels. It is more likely that it's a SP20 voltage configuration settings or a power supply issue. 0: DC2DC/1/:[vlt1:625 vlt2:629(DCl:794 Tl:629 Ul:688) 54W 87A 73c] ASIC:[110c (125c) 720hz(BL: 720) 91 (E:193) F:0 L:0] 1: DC2DC/1/:[ vlt1:572 vlt2:576(DCl:794 Tl:576 Ul:688) 32W 56A 75c] ASIC:[ 85c (125c) 490hz(BL: 490) 76 (E:193) F:0 L:0] LOOP[1] ON TO:0 (w:139) 2: DC2DC/1/:[ vlt1:562 vlt2:565(DCl:794 Tl:565 Ul:688) 26W 47A 66c] ASIC:[ 85c (125c) 440hz(BL: 440) 58 (E:193) F:0 L:0] 3: DC2DC/1/:[vlt1:603 vlt2:608(DCl:794 Tl:608 Ul:688) 34W 57A 61c] ASIC:[ 85c (125c) 490hz(BL: 490) 81 (E:191) F:0 L:0] LOOP[2] ON TO:0 (w:226) 4: DC2DC/1/:[vlt1:666 vlt2:672(DCl:794 Tl:672 Ul:688) 80W 119A 88c] ASIC:[120c (125c) 900hz(BL: 900) 116 (E:193) F:0 L:0] 5: DC2DC/1/:[vlt1:625 vlt2:629(DCl:794 Tl:629 Ul:688) 59W 94A 90c] ASIC:[115c (125c) 770hz(BL: 770) 110 (E:193) F:0 L:0] LOOP[3] ON TO:0 (w:195) 6: DC2DC/1/:[ vlt1:587 vlt2:592(DCl:794 Tl:592 Ul:688) 32W 55A 73c] ASIC:[100c (125c) 480hz(BL: 480) 94 (E:193) F:0 L:0] 7: DC2DC/1/:[vlt1:646 vlt2:650(DCl:794 Tl:650 Ul:688) 66W 102A 76c] ASIC:[ 85c (125c) 820hz(BL: 820) 101 (E:192) F:0 L:0] Again here you go with your science of LIKELY which is by no means evidence lol as you can see from his post his start volts are .67 which is much higher than the ones doing .5x hmmmm wonder why they aren't getting the proper volts...you should know that the start volts are within .005 of the setting in the gui... could there be a short somewhere  He could have also damaged the boards taking the heatsinks off...so many variables but back on topic AS5 is not really the best choice to use based on price and composition Science is the art of Likelyhood. Well the start voltage is a good observation. It's possible that OP did not apply the epoxy adhesive to close the thermal gap between the chip, copper plate and the heatsink. Thanks for introducing another irrelevant angle, one that of price. Lets compare their thermal conductivity. I think thermal conductivity of AS5 is ~8X that of MG grease. |
|
|
|
as scientific as "likely" ....cheers If we are to be scientists we should perhaps quote guy who said that epoxy is not necessary to re-secure the heatsinks as the glue was there to hold them in place during shipping but I digress...just trying to give SAFE advice to someone who clearly has no idea what they are doing (OP) SAFE advice could be just an opinion without proper scientific evidence backing it up. Now you are just talking shit. u took 1 heatsink off and reapplied grease to it and call it safe without having the board schematics...try doing what the OP did and post your results ...have you re and re an entire machine to base your results on? I have... and the OP has fucked up his machine as you can clearly see from his results so please put your ego aside...or paste up an entire SP20 with AS5 and post your results like the OP The board schematics are not necessary since I am NOT pasting up the whole board, just the die.  |
|
|
|
as scientific as "likely" ....cheers If we are to be scientists we should perhaps quote guy who said that epoxy is not necessary to re-secure the heatsinks as the glue was there to hold them in place during shipping but I digress...just trying to give SAFE advice to someone who clearly has no idea what they are doing (OP) SAFE advice could be just an opinion without proper scientific evidence backing it up. Now you are just talking shit. u took 1 heatsink off and reapplied grease to it and call it safe without having the board schematics...try doing what the OP did and post your results ...have you re and re an entire machine to base your results on? I have... and the OP has fucked up his machine as you can clearly see from his results so please put your ego aside...or paste up an entire SP20 with AS5 and post your results like the OP I am trying to make sense of evidence provided without being overtly opinionated. It's your ego that is getting bruised without a specific reason. I donot see a plausible explanation as how AS5 can screw a SP20. |
|
|
|
as scientific as "likely" ....cheers If we are to be scientists we should perhaps quote guy who said that epoxy is not necessary to re-secure the heatsinks as the glue was there to hold them in place during shipping but I digress...just trying to give SAFE advice to someone who clearly has no idea what they are doing (OP) SAFE advice could be just an opinion without proper scientific evidence backing it up. The evidence points that the voltage for some chips are not at their desired levels. It is more likely that it's a SP20 voltage configuration settings or a power supply issue. 0: DC2DC/1/:[vlt1:625 vlt2:629(DCl:794 Tl:629 Ul:688) 54W 87A 73c] ASIC:[110c (125c) 720hz(BL: 720) 91 (E:193) F:0 L:0] 1: DC2DC/1/:[ vlt1:572 vlt2:576(DCl:794 Tl:576 Ul:688) 32W 56A 75c] ASIC:[ 85c (125c) 490hz(BL: 490) 76 (E:193) F:0 L:0] LOOP[1] ON TO:0 (w:139) 2: DC2DC/1/:[ vlt1:562 vlt2:565(DCl:794 Tl:565 Ul:688) 26W 47A 66c] ASIC:[ 85c (125c) 440hz(BL: 440) 58 (E:193) F:0 L:0] 3: DC2DC/1/:[vlt1:603 vlt2:608(DCl:794 Tl:608 Ul:688) 34W 57A 61c] ASIC:[ 85c (125c) 490hz(BL: 490) 81 (E:191) F:0 L:0] LOOP[2] ON TO:0 (w:226) 4: DC2DC/1/:[vlt1:666 vlt2:672(DCl:794 Tl:672 Ul:688) 80W 119A 88c] ASIC:[120c (125c) 900hz(BL: 900) 116 (E:193) F:0 L:0] 5: DC2DC/1/:[vlt1:625 vlt2:629(DCl:794 Tl:629 Ul:688) 59W 94A 90c] ASIC:[115c (125c) 770hz(BL: 770) 110 (E:193) F:0 L:0] LOOP[3] ON TO:0 (w:195) 6: DC2DC/1/:[ vlt1:587 vlt2:592(DCl:794 Tl:592 Ul:688) 32W 55A 73c] ASIC:[100c (125c) 480hz(BL: 480) 94 (E:193) F:0 L:0] 7: DC2DC/1/:[vlt1:646 vlt2:650(DCl:794 Tl:650 Ul:688) 66W 102A 76c] ASIC:[ 85c (125c) 820hz(BL: 820) 101 (E:192) F:0 L:0] |
|
|
|
I still don't think its a good idea to use AS5 on top of the ASIC...likely and maybe are just guessing that it wont screw anything up...better safe than sorry I always say...and that warning is there for a reason, keeping in mind that not everyone knows how to even apply grease we could assume its likely though  Nice scientific opinion here. thanks! |
|
|
|
I repasted one of the chips, because I had pried open the heatsink to take a picture.
Arctic Silver 5 is not a conductive thermal compound so you are good.
Not Electrically Conductive:
Arctic Silver 5 was formulated to conduct heat, not electricity.
(While much safer than electrically conductive silver and copper greases, Arctic Silver 5 should be kept away from electrical traces, pins, and leads. While it is not electrically conductive, the compound is very slightly capacitive and could potentially cause problems if it bridges two close-proximity electrical paths.)Ofcourse a thermal compound is thermally conductive, it's apparent in the definition. Second holds true for almost all thermal compounds. Likely SP20 does not have high speed (GHz) data buses on the PCB, where a drop of AS5 can close the eye. |
|
|
|
|
I repasted one of the chips, because I had pried open the heatsink to take a picture.
Arctic Silver 5 is not a conductive thermal compound so you are good.
But I wonder how did you stick the heatsink back? Since a good amount of pressure and a strong epoxy is needed to make sure thermal conductivity is restored between the chip, the copper plate and the aluminum heatsink.
|
|
|
|
When is the last time BFL sent a refund out?
My lawyer and I are also interested in this information. |
|
|
|
|
With the impending reward halving next year, when do you think will the price start ramping up?
|
|
|
|
|
Is developing your own ASIC, out of question?
|
|
|
|
|
It might be a touch and go.... $2000/BTC is unsustainable at current mining reward rates ($7.2 million rewards a day, ~$200 a day per 10TH/s at current difficulty), considering most large-scale miners are short-term profiteers.
|
|
|
|
And GPUs are still 28nm. What a joke
That's because of TSMCs failed FinFet attempts.  |
|
|
|
 Interested. |
|
|
|
God I loved those old Bitfury kits. I had 5 semi-full kits running at one point. None were ever to run all 16 cards though without random cards shutting down. There was always lots of wiggling/swapping cards and taping them together to try to keep as many cards running as possible, lol. And when I could get 14 or 15 cards running on one M board... I just tried not to walk by too fast and cause a wind disturbance or vibration or else I'd have to start wiggling and swapping cards again. Them were the good ol' days! I would have never imagined that I would own 100s of GPUs or have such an intimate relationships with circuit boards. |
|
|
|
|
3.0.
3.1 was much better on my 386DX PC.
|
|
|
|
Here is the best strategy to earn big money with ASIC miners... Take your 250k and use it to invent a time machine. And now forget how that invention has already made you the wealthiest man in the world and travel back to the summer of 2012. Buy a couple of AMD GPUs just to familiarize yourself with BTC mining. Don't go crazy and build a massive GPU farm or buy those intriguing FPGAs you've been reading about. They will all be obsolete by the beginning of 2103. Sometime during the summer of 2012 you will be reading here on BTCtalk about 4 companies who are planning to release ASIC miners by the end of the year... Butterfly Labs, bASIC, ASICMiner and Avalon. Now here's the really important part; the first two companies will completely flake-out and rip-off their customers, the third will produce functional miners that are super high-priced and the last (Avalon) will pre-sell their first batch 66GHs ASIC miners for $1300. Order as many of the first batch Avalon miners as is humanly possible! And maybe slip those Chinese fellows some extra cash to skip ahead in line. Being first in line is VERY important. Order more than you can possibly host at home. You will soon be able to sell the extras for 20 times their original retail price. However, do not (under any circumstances) order from any later Avalon batches. The company will self-mine with them for many months before delivering the dusty miners to the customers. Your shiny new ASIC miners will arrive nearly on time around the very beginning of 2013. You will now be the king of BTC mining! You will be mining hundreds of BTC every day (depending on how many miners you bought). Save those BTCs in a secure offline wallet and don't spend any of them until you hear about another company developing much more efficient ASIC chips... Bitfury. Bitfury is important because soon everyone and their brother will have these loud, hot ASIC miners built by Avalon (and later, Butterfly Labs) that suck 6-8W/GHs and BTC mining difficulty will absolutely skyrocket. But Bitfury miners will use only ~1W/GHs. They will look like amateur crap, be finicky as heck, you'll have to buy tons of fans to keep them cool and probably quit your day job just to keep them running. Now open your very full BTC wallet and preorder every Bitfury August batch card you can afford. Towards the end of August, sell every Avalon miner you have to clear space for your Bitfury gear. And surprise! Your Bitfury miners will arrive on time while people are still waiting for their year-late Butterfly Labs miners. And now here's the really, really important part... sell every remaining BTC you have as soon as it reaches $1150 near the end of 2013. And keep selling every BTC as soon as it's generated by you quiet, cool Bitfury ASICs. Go outside and enjoy the weather, take that vacation you've been waiting to take, kiss a girl (or boy), walk the dog, etc. Do anything you want, you're rich! Oh and don't forget, you still have that kick-ass Time Machime to play with.  Don't worry about buying BTC again until January of 2015. And certainly don't bother trying to keep up with the rising BTC difficulty by buying miners from KNC, Bitmain or Spondoolies. Travel back to 2009, use some spare desktops, laptops and servers. Save yourself the trouble of a mining farm. Buy pizzas for 10K BTC a pop with the coins. Don't forget to Hoard 50% of your mining stash in an offline wallet. |
|
|
|
|
We have to also understand the mining dynamics.
If BTC were at $50k a pop, and mining rewards at 3600BTC / day, the total mining revenue will be $180million a day, $66Billion a year.
If we assume a 10% mining profitability rate, then BTC network will consume $60Billion worth of electric energy to mine.
Clearly BTC is unlikely to reach $50k a pop at the current reward rate.
|
|
|
|
I've actually plugged in my two SP20s at least part time for the past few days. I've had all of my gear shutdown since April or May when it got hot here in the desert. But I thought I'd at least run the SP20s again for shits-and-giggles since the BTC exchange rate is up ~$300.
So, what do you experts suggest for the most efficient settings? I've look through some of the other SP20 threads for suggestions but that exact info is difficult to find.
And any pool suggestions? I used to mine at BTCguild but I was a bit surprised to see that they've completely shut down.
My most efficient settings is running them at 0.6V ( 1000W @ wall off 2 SP20s ;Gold PSU). The hashrate per miner is 1000GH/s. Goxed, what is your max voltage setting? I'm currently at .61V with a max of .65V and still getting 1200GHs so I probably have room to come down a bit more for better efficiency (and lower fan speed/noise!). Cheers, thanks everyone for your SP20 settings and pool suggestions! The start voltage is 0.6V and Voltage Limit is 0.61V. Max PSU power is 125W for all. |
|
|
|
|
Show Posts
