My 3* boards have just arrived. My power supply has no PCIe connector, just single positive and negative wires. My question is; what is the best way to wire this to my board?
Should I wire the power supply wires to some sort of adapter and plug into the PCIe connection or would it be better to figure out a way to wire this to the barrel jack connection?
I sell pci-e connectors with thick 16awg wires in my signature. might take a few days to arrive though but are an excellent solution that will not melt like cheap 18awg or smaller wires can |
|
|
|
Well, mine definitely broke somehow. All of a sudden I just started getting massive rejected shares, and my hashrate dropped to 500-600GH/s
Anyone else experienced anything similar?
pool issue? eligius is all over the place the last 2 days. otherwise does the webUI indicate it is certain boards faulting or an overall drop in speed? |
|
|
|
Weekly Update 4/23/2014.
Datacenter:
We are up to 64.34 Th/s and expect by the end of the month to be around ~170 TH/s. Everything is in place to expand our hash rate to this level.
Ken could you clear up the meaning of the text above that I've bolded. It suggests that we will be at 170 TH by the end of the month which is only one week away. Did you actually mean we will be at 170 TH one month from now? No, we have everything in place and my Project Manager has told me to expect ~170 TH by the end of the month. It may take a week or so longer as there are only a few more days left in this month, but we do have everything is place to put THs on line quick. Weekly Update 4/30/2014
Datacenter:
We are continuing to add TH; however, at this point our 2 units of air conditioning are maxed out. We have been repairing the third unit and it should be up and running next week when parts arrive. After the third unit starts running we will be adding more THs to the datacenter.
In the course of one week you planned to install >100TH, but only installed <10TH. Assuming your third broken AC is the same as the first two, its unlikely that you will be able to get above a total hashrate of 100TH total, let alone 170TH or more |
|
|
|
200 --------------------------1.6------------------------- 0781
225 ------------------------- 1.8 ------------------------- 0881
I found some more values on the site here
206.25------------------------------ 1.65------------------------ 5002
212.50 ----------------------------- 1.7 ------------------------ 5082
215.625 ---------------------------- 1.725 ------------------------ 6206
218.75 ------------------------------ 1.75 ------------------------ 5102
the above codes were from my hand calculations, GH speeds added. 215.625 gives me 1100Gh and less than 1.1% error rate |
|
|
|
1015 Watts, from the wall @120V
Ouch! That's more than 100%  Maybe I'll install my Seasonic 1050W in this thing tomorrow. the psu is ~88% efficienct near full load. that would be about a 900w draw from the miner |
|
|
|
Aww man, just as it looks like things are picking up someone goes and files a lawsuit, damn Americans and their "sue everyone" tendencies. Just kidding, but really I hope this is resolved soon so we can get back on track, although this will require Ken to step up and be completely open about whatever's gone down, regardless of how bad it is.
Sadly its true, you have over-invested panicky Americans who think the courts will save them from their own financial mistakes when the only person that can be blamed is themselves for committing too much money to this security. Ken hasn't done a great job but this lawsuit doesn't make sense at this time. A month ago we had nothing so it had logic, now there actually have been shipped products(albeit poorly) and we are actually hashing so that seems like an ideal time to try to ruin your investment prospects by filing more lawsuits. The complaints are about much more than a few poorly packaged and shipped hashfast boards. |
|
|
|
How can the unit draw 1Kw and only have a 1Kw PSU ?
If you buy an s-2 I advise you to pull 1 blade and list it on ebay for a high price. say 350 usd. Slowly drop the price if it does not sell. As long as you get more then 250 the price is not bad. having 9 blades will allow the stock psu to not fry and die. getting 250 or more for the blade will make up for the lost hash. do the math what does 100gh earn. and the math shows the 100gh lost does not pay off 250 usd so selling it for 250 usd works out as the best move. bottom line is the s-2 is not a great buy. it needs to be priced lower Bitmain is a shrewd businessman - that is clear. He knows at this price the S2 will not ROI - watch for the price drop coming soon which will blow the S1 out of the water. This is typical businessmanship. I like this guy. keep in mind the cost of chips - there are 10x as many chips in an S2 as there is in an S1. If the chips still cost bitmain >$1.50 each (~$0.50/GH in an antminer S1) then they are a significant portion of the S2 price. add in a nice LCD, quality case (still tweaking the card support of course), and a PSU - a price drop to $2000 after coupon with delivery before the 14th of may would be fair, and a price I could afford a small premium over the S1 to buy. (fewer cords and better stacking for bigger mining farms) |
|
|
|
So I got one client with an extrange case with his S1 miner. He connects the miner with 2 pci connectors and it tries to start working, but after few seconds the power supply simply powers off. If he simply powers up the blade with that holds the ethernet connector it works fine, but if he just tries to power up the other the power supply powers off again. The power supplly is a nox urano vs 750W. He also tried to put something on the 5V rail just instead the power supply was not balanced, but it doesn't work neither. What can it be ? a) The power supply b) Broken board c) other ? Thanks in advance for any help  his power supply might have multiple 12V rails. (lets assume 2 rails, each capable of ~320W). If he uses both pci-e connectors that run off a single rail, the 360W+ draw of the antminer pops the PSU current protection. |
|
|
|
One of the recent batch (ordered at the first recent sales day) S1 works for 5 min, then disconnects, then in a few seconds beeps, then reconnects.
This cycle repeats all day long. What could be the cause?
i tried to reboot multiple times, switched off power supply, then back on-no cure. I have a feeling that I have a rare faulty unit. Thanks for any suggestion.
70% likely its your power supply. after 3min or so the antminer is at full speed and after a minute or two of this your PSU is finally overloaded and switches off. a lot of power supplies will attempt to restart themselves when this happens |
|
|
|
LOL "the garbage 18AWG connectors found on ebay" You better not be referring to my pristine cables fool.
not sure which those are. I just know ive had an '8pin ATX to 6pin PCIe' cable burst into flames within 5 minutes of turning on the power supply, and only 200W of load through the connector (which actually had 8 18awg wires). I also cut open a molex->pcie wire that was labelled as 18awg and the inside wire is much thinner than a sewing needle. These 16AWG wires are good quality, and are fully capable of carrying a high current without heating up or failing yikes it is troubling that people would sell things and lie about them! No thicker than a sewing needle might be 22 gauge or less!  I'm the guy with pterodactyl teeth cables. http://r.ebay.com/vnn6Hq16 gauge and they work with your cables i think....  these cables would definitely be compatible! LOTS of cables still for sale! They work great and the tinned ends make soldering as easy as it gets |
|
|
|
Well everyone has their own standard for "freaking LOUD". I agree that, at full fan speed - heck even at half speed - they're pretty annoying. But the fans don't need to be up that fast to keep the supplies cool outside of a highly-restrictive-server installation. Mine are running at probably less than quarter speed, about a tenth of the full-speed noise, at 85-90% load for months with no problem. I know some people are touchy about being able to hear fans but honestly I don't care much. I am less annoyed by three of those server supplies with properly adjusted fans than a high-strung GPU fan though, if that helps. I have some running GPUs, and the GPU fans are definitely louder.
And actually, the most annoying server fan I've run into so far was on the DPS-800. I had one running two S1 for a couple weeks to see how it behaved, and those fans are internally thermostat controlled so there's no manual adjustments. Asking 800W from it really pushed the fan up into "Holy crap that's irritating" territory, far past the volume level of a manually adjusted and comfortable-steady-state-cooling 750W Dell.
If I had a dB meter I could give you comparisons, because it's pretty clear my fan noise tolerance is atypically high. I know a guy that can only sleep with earplugs, with less fan noise going on.The last time fans kept me up it was actually the lack of fan noise - a PSU died and my equipment turned off, which the sudden silence woke me up before the fans had even finished spinning down.
The DPS-800 does get loud. but IMO the worst is the artesyn 835w supplies. They sound okay up to about a 550W draw, then start to really get loud. at 800W+ they are basically srieking due to the tiny singluar 60mm fan |
|
|
|
|
^or the sketchy internet cafe (on a top floor?)?
Name & Registered Office:
ALLMINERS LTD
196 HIGH ROAD
LONDON
UNITED KINGDOM
N22 8HH
Company No. 08844830
Status: Active
Date of Incorporation: 14/01/2014
Country of Origin: United Kingdom
Company Type: Private Limited Company
Nature of Business (SIC):
None Supplied
Accounting Reference Date: 31/01
Last Accounts Made Up To: (NO ACCOUNTS FILED)
Next Accounts Due: 14/10/2015
Last Return Made Up To:
Next Return Due: 11/02/2015
Mortgage: Number of charges: ( 0 outstanding / 0 satisfied / 0 part satisfied )
|
|
|
|
yeah, this is an obscene scam. at the very best, they are using coupons to buy antminers for $2,200 then turning around to sell than for $2,500. at the worst its a total scam that chose the worng dry cleaners address to use :p |
|
|
|
Just out of curiosity, has anyone purchased the S2's at the $2199 price even? I'm tempted to if BTC reaches $480+ in the next few days.... thoughts?
$2199? They are currently listed at $2599. Alot of people who previously purchased an antminer s1 or s2 got a $400 coupon for a future order. Ah, sure enough. I initially saw a $100 coupon for an S2. Looking again, they are now $400 coupons. Yep...$2,599 - $400 = $2,199 ...or 4.93 BTC at the moment... But we can buy 5 x S1 for 2.5 BTC...or almost 10 x S1 for the price of one S2 right now...  I should wait for S2 price to come down ...like $1.599 ...by the end of May maybe... Cheers, ZiG If Bitmain would release the spec sheets of the boards someone would probably figure out some overclocking pencil mod, which would improve sales a little, I'm sure 1100GH is possible without pencil modding and 1.1% error rate. any frequency settings higher, and the error rate is over 2% and the unit can even fail offline and need a power cycle (not sure if the unit or the psu caused this). I think with the correct resistors identified you could get up to 1400GH/2100W, but probably not without pci entenders and much more cooling. 1280GH (~250MHz, ~1600W) is probably the most you could achieve with pencil modding while keeping all the components in the case with the lid on. $2200 with coupon is not a bad price. for those who have limited space or larger farms it is better than most of the competition products. For those with shelving units and more space though, the S1 makes a lot of sense and can be undervolted/clocked when the time comes in 1-2months to improve the efficiency and keep them profitable for a few extra difficulty jumps |
|
|
|
Terraminers at $3799 now!https://cointerra.com/product/terraminer-iv-1-6-ths-bitcoin-miner/STIN, Texas – 04/28/2014 CoinTerra, the market leader in ASIC Bitcoin mining solutions today announced that it has adopted a variable pricing model for retail sales of their TerraMiner IV Bitcoin mining hardware. Having delivered the April and May/June batches significantly ahead of schedule, CoinTerra has decided to move to a model of daily pricing for its from-stock sales in order to reflect the ever-changing variables which go to make up the Bitcoin ecosystem. “As the leading high-performance Bitcoin mining hardware manufacturer shipping from stock, we can now offer a pricing model that allows Bitcoin miners to make purchasing decisions based on current factors. This makes CoinTerra’s spot pricing exceptionally attractive to members of the Bitcoin mining community wishing to buy mining hardware for immediate delivery and deployment.” said Ravi Iyengar, CEO – CoinTerra, Inc. Every 24 hours market analysts at CoinTerra will determine the following day’s spot price, and the first price point offered on Monday April 28th is $3,799 per TerraMiner IV with a volume discount of $200 per unit when purchasing 10 or more. Depending on the market, pricing may increase or decrease, or remain unchanged. Additionally, CoinTerra has announced that it has begun to offer the TerraMiner’s GoldStrike I ASIC for sale to third-parties to integrate into their own system designs, with a minimum order of 500 ASICs. To order a TerraMiner IV from stock for immediate shipping visit: https://cointerra.com/shop 2 days later and no price change... |
|
|
|
Why do people still make or buy USB ASICS? I just cant see any value at all in this?
1) novelty: its less than $10/usb, or a reasonable ~$4/GH. if someone wants to learn about bitcoin mining this is a nifty cheap way to do it. (of course, thats if buying from a reseller since the 500 MOQ is a bit much) 2) test-bench: I currently have my antminer S2 running at a frequency of 215.625, which I figured out myself correlates to '6206'. However, rather than test a potentially wrong asic-freq on the S2, I implement it on the U1 and let it run for 10min to see that the hashrates and temps are where I expect them, THEN i make the changes to the S2 |
|
|
|
First overnight report of my Batch 3 running on stock settings, stock firmware and stock power supply. Hashing at 1025 to 1080GHs solid, no beeps or loss of power. Batch 3 S2 replaced 5 BE Blades running on a backplane  and heat output is significantly less. Compare that to 50GHs on 5 BE Blades with just about 80% of a S2 power consumption and twice is much heat. I'm happy with my S2 so far, it is a little loud though but I can live with that.  thats really good speeds - my batch 2 ran at something like 980GH stock. did they change something for batch 3, such as a higher frequency or slightly increased voltage? |
|
|
|
Update [April 30 2014]
The refund check (Now cancelled)Sent in Feb turned up today.
postmark confirms it was sent in Feb and the check number confirms with what ken said it was on pm. So I can't blame him for this delay. The replacement check should arrive soon.
If ken could have just communicated this postage it would have avoided a whole rigmarole for the past 2 months.
In any case I unreservedly apologise to ken in light of this check arriving when all the evidence was pointing to a scam.
curious how or why it took almost two months for the cheque to arrive? he should really be sending via registered mail for tracking and verification of delivery |
|
|
|
0.55 btc aint that yesterdays price? Whats todays price? Daily price change init
you are correct but it is not quite 24 hours. Since the drop to .55btc. Come back in about 2 or 3 hours. Since those will be shipped only on May 4 I doubt they will drop the price until the new batch. Also does anyone have a solution for underclock beside the graphite pen trick? 140Gh for 180W is quite decent compared to S2. thats the only real way without soldering new resistors. you can underclock it at stock voltage, but it will either get lots of errors below a certain speed (lets say below 160GH?) OR it will just use way too much wattage unless you bring down the operating voltage via the pencil mod (ie: you might get 140GH/220W if you dont undervolt) |
|
|
|
I tried the image from 1l1l11ll1l but for my it did not work.
win32discimager writes for a few seconds but after that i can't read the sd card.
There was also a other image ( updated? ) and that one worked.
It took a few minutes to write that one.
I have just send a email to bitmaintech to ask about this booting problem.
When i have found the problem i will post it in here.
You will probably need to format the SD card properly before writing the image or it might not work (it's partitioned). Download a SDFormatter to format the card and make sure you see the full memory available before writing the image with win32diskimager. Sounds like you got it working though... if you did, you should make an img of the working card and write it to a spare or at least to disk, to make any future resetting go quicker. Yes i can see the files on the SD card, both on the card i made with the diskimager and on the original one. btw, the sd card i used came formatted, i will try SDFormatter tonight to be sure. Thank's for the advice. Make sure they are class 10 sd cards. That worked for me. I had the same problem on my batch 2. Three of them just did not boot when restarted. Even after replacing the sd cards with new ones and the image provided by 1l1l11l1l (thanks!) still no boot. But they were not class 10 cards. I don't know why but as soon as I put in a class 10 card in, with the same image, it worked without a problem and on all three. Good luck same for my bitfury system a few months ago. the original SD corrupted after a few weeks and i could simply re-image it. however, after a few more it spat errors at me and would not read, format, or reimage. I switched to a new card of the exact same brand/size/speed and the same thing happened after a half-dozen power cycles. I switched to an 8GB class 10 and no issues ever since |
|
|
|
|
Show Posts
