ANTMINER S2 Discussion and Support Thread - page 59.

Tigggger

legendary

Activity: 1098

Merit: 1000

Quote from: nottm28 on June 04, 2014, 05:07:06 PM

Quote from: s1gs3gv on June 04, 2014, 05:03:42 PM

The SP10 is such a crappy deal compared to the Antminer S2

Yep, their reluctance (or greed) to reduce the price while everyone was pissed off with BM has cost them greatly. What's better, sell 1 miner for $500 profit or 10 miners for $200 profit... spondoolies definitlely missed the boat here...

Curious how you both think the S2 is a better deal

Price
S2: $2359, 1000GH = $2.36 per GH
SP10: $2795, 1400GH = $2.00 per GH

Power
S2: 1000 GH, 1000W = 1w/GH
SP10: 1400 GH, 1200W = 0.85w/GH

PS Both are beaten on price by the A1 clones, which can be had for around $1800, or $1.80 per GH (if only all the dealers weren't sketchy)

nottm28

hero member

Activity: 574

Merit: 500

Quote from: s1gs3gv on June 04, 2014, 05:03:42 PM

The SP10 is such a crappy deal compared to the Antminer S2

Yep, their reluctance (or greed) to reduce the price while everyone was pissed off with BM has cost them greatly. What's better, sell 1 miner for $500 profit or 10 miners for $200 profit... spondoolies definitlely missed the boat here...

s1gs3gv

legendary

Activity: 1316

Merit: 1014

ex uno plures

The SP10 is such a crappy deal compared to the Antminer S2

mdude77

legendary

Activity: 1540

Merit: 1001

Quote from: smn68 on June 04, 2014, 04:32:49 PM

I have 4 batch 4 S2's that max out at 1100 then drop down to 800-900 then back up to 1100 constantly, never a stable rate.

I tried spreading the load out over 2 power supplies today but it made no difference [1 seasonic 1000w and the S2 enermax 1000w]

They seem to be picky about pools.

My S2s run like champs on Eligius.

On BTCGuild, one had 36% "bad", mostly "unknown". The other was 1% bad.

On p2pool, they run around 920-930gh/s, and more than once I've had errors appear (x's or dashes) as soon as I point it to p2pool.

M

smn68

newbie

Activity: 19

Merit: 0

I have 4 batch 4 S2's that max out at 1100 then drop down to 800-900 then back up to 1100 constantly, never a stable rate.

I tried spreading the load out over 2 power supplies today but it made no difference [1 seasonic 1000w and the S2 enermax 1000w]

samsonn25

hero member

Activity: 882

Merit: 1003

Two high quality 660 platinum are enough to push 1400 watts easily. Sea sonics can he had after rebate for $80 every few months.

bspurloc

hero member

Activity: 569

Merit: 500

How are the 10 boards powered from the power supply? 10 atx wires? (oh 6 atx to power board) 3 per 750 power supply then!
Wouldn't it be cheaper and more efficient using 2 750watt supplies? of course one would be outside the case.

I would think as we phase out the S1's those left over 750's would power these easily.

anyhow that is what I would do instead of running a power supply at 100% 24/7/3.
3 being for 3 months when it fries itself.. haha.
even 2 600's or 650's.

xzempt

hero member

Activity: 504

Merit: 500

guess its about time to bite the bullet and buy an s2...

nottm28

hero member

Activity: 574

Merit: 500

Quote from: TracerX on June 04, 2014, 02:51:45 PM

Quote from: Biffa on June 04, 2014, 01:00:52 PM

I replaced my Batch 2 PSU with a 1300W EVGA and its made zero difference to the hash rate.

Thank you for the reply, I'll hold off for now. Appreciate it.

+1 me too - nearly ordered one before I seen your post...

TracerX

hero member

Activity: 918

Merit: 1002

Quote from: Biffa on June 04, 2014, 01:00:52 PM

I replaced my Batch 2 PSU with a 1300W EVGA and its made zero difference to the hash rate.

Thank you for the reply, I'll hold off for now. Appreciate it.

Biodom

legendary

Activity: 3892

Merit: 4331

Quote from: samsonn25 on June 04, 2014, 01:47:58 PM

Assuming Enermax cable is minimum 16 guage thats 10amp. Could be psu is struggling at full load to supply the necessary amps to miner, because even 1000 watt unit is most efficient at 50% load rating this drops at 95-100%. Maybe switch the power cables for another test that would narrow down.

Actually, 16 gauge wire is rated up to 13A (safely 12.5A), so it should be fine in US for S2 (no need for 14 gauge).
the confusion about 10A stems from the fact that it is 13A for a single phase (residential recepticle) or 10A for three phase
see
http://www.rvpowerprotection.com/Links_files/extension-cords.pdf
and
http://www.chromausa.com/pdf/app-notes/AN%20-%2061500_61000%20-%20Single%20&%203%20Phase%20Power%20-%2003122013.pdf

samsonn25

hero member

Activity: 882

Merit: 1003

Quote from: ?? on ??

Quote from: Biffa on June 04, 2014, 01:00:52 PM

Quote from: ?? on ??

Quote from: ak49er on June 04, 2014, 10:59:30 AM

Quote from: edgar on June 04, 2014, 01:58:54 AM

1000w is clearly NOT ENOUGH and anyone who hasnt upgraded is losing GH by the second.

......

anyway - ive learnt that they need more power for a fact and i will be lucky if the ever helpful bitmain can sort me out

My batch 4 S2 is running quite happily at 1005 GH/s on the original 1000W PSU, (1086W at the wall). Not sure where you're getting your facts.

and mine are fluctuating 'happily' around 900-1000 INCLUDING the one with ONLY 9 boards which in the short time its been running seems more stable.

you could probably get 850GH from a 850w PSU.

the point which seems obvious to me is the boards have the potential to do more..and the 1000w PSU prevents that.

not sure where you're getting your logic...

@Tracer X - excellent point about new blades..who knows if higher or lower, id guess higher.

but might be better to hold off unless u have the 1200-1300w psu handy... which u and i do not..

you might check your thermal grease. if its dried out or non-existent you may be getting increased errors - its something im looking into.

I replaced my Batch 2 PSU with a 1300W EVGA and its made zero difference to the hash rate.

cant say fairer than that.

perhaps my older batch 4 is just a lemon and my battered 'new' rig is a potential 1st place contender.

perhaps there are other limiting factors...

just because we paid for 1TH doesnt mean we should willfully ignore the possibility of a potential increase.

EVERY other rig ive ever ran has given more than it was touted as able to - eventually...

why should these be any different??

forgot to add - the rig powered by my OCZ1000w hasnt dipped below 1TH in days - while both the 10 blade & 9 blade rig are powered by enermax 1000w and are fluctuating between 900 -1000Gh/s

the power cable for the OCZ is almost 3 times as thick as the enermax cable... and relatively cool to touch - enermax cables are considerably warmer.

Assuming Enermax cable is minimum 16 guage thats 10amp. Could be psu is struggling at full load to supply the necessary amps to miner, because even 1000 watt unit is most efficient at 50% load rating this drops at 95-100%. Maybe switch the power cables for another test that would narrow down.

nottm28

hero member

Activity: 574

Merit: 500

Quote from: ?? on ??

perhaps there are other limiting factors...

Could be networking issue which I have found has been the cause of most of my unexplained pain.

Try swapping the network cable from the best miner to the worst miner and run for a few hours...

philipma1957

legendary

Activity: 4326

Merit: 8950

'The right to privacy matters'

Quote from: -droid- on June 04, 2014, 12:25:26 PM

Quote from: Biodom on June 04, 2014, 12:21:54 PM

minersource.net has a good deal on used EVGA1300 for $140.
I've got a new power board with black caps-anybody has this? Will take a picture later for comparison and will also try the new board with old psu, then try to change to EVGA1300 and report performance.

its sold out

buy one new for 195 and get back 2900 in point = 29usd to spend later.

read how here.

https://bitcointalksearch.org/topic/m.7073381

there are more discounts possible so a new one is about 160 with a 10 year warranty

Biffa

legendary

Activity: 3234

Merit: 1221

Quote from: ?? on ??

Quote from: ak49er on June 04, 2014, 10:59:30 AM

Quote from: edgar on June 04, 2014, 01:58:54 AM

1000w is clearly NOT ENOUGH and anyone who hasnt upgraded is losing GH by the second.

......

anyway - ive learnt that they need more power for a fact and i will be lucky if the ever helpful bitmain can sort me out

My batch 4 S2 is running quite happily at 1005 GH/s on the original 1000W PSU, (1086W at the wall). Not sure where you're getting your facts.

and mine are fluctuating 'happily' around 900-1000 INCLUDING the one with ONLY 9 boards which in the short time its been running seems more stable.

you could probably get 850GH from a 850w PSU.

the point which seems obvious to me is the boards have the potential to do more..and the 1000w PSU prevents that.

not sure where you're getting your logic...

@Tracer X - excellent point about new blades..who knows if higher or lower, id guess higher.

but might be better to hold off unless u have the 1200-1300w psu handy... which u and i do not..

you might check your thermal grease. if its dried out or non-existent you may be getting increased errors - its something im looking into.

I replaced my Batch 2 PSU with a 1300W EVGA and its made zero difference to the hash rate.

ak49er

sr. member

Activity: 672

Merit: 250

Buy, sell and store real cryptocurrencies

Quote from: ?? on ??

not sure where you're getting your logic...

My logic is quite simple. I paid for a 1TH/s miner, and that's exactly what I got.

Now undoubtedly the S2 pushes that Enermax PSU to its limits, and a much safer bet would be a 1200 or 1300W unit. But I'm not loosing GH by the second. I certainly don't have the overhead required to overclock, but then I'm not sure I want to do that anyway.

ak49er

sr. member

Activity: 672

Merit: 250

Buy, sell and store real cryptocurrencies

Quote from: TracerX on June 04, 2014, 11:32:38 AM

Quote from: ak49er on June 04, 2014, 10:59:30 AM

Quote from: edgar on June 04, 2014, 01:58:54 AM

1000w is clearly NOT ENOUGH and anyone who hasnt upgraded is losing GH by the second.

......

anyway - ive learnt that they need more power for a fact and i will be lucky if the ever helpful bitmain can sort me out

My batch 4 S2 is running quite happily at 1005 GH/s on the original 1000W PSU, (1086W at the wall). Not sure where you're getting your facts.

How long can that PSU run overspec? That it's running fine now is great, but how long will it do that? I'm not being confrontational, I'm an owner of an S2 and am curious.

Well I have to say that I'm curious too.

I have a spare 1000W for if and when this one goes tits up.

-droid-

legendary

Activity: 1050

Merit: 1001

Quote from: Biodom on June 04, 2014, 12:21:54 PM

minersource.net has a good deal on used EVGA1300 for $140.
I've got a new power board with black caps-anybody has this? Will take a picture later for comparison and will also try the new board with old psu, then try to change to EVGA1300 and report performance.

its sold out

Biodom

legendary

Activity: 3892

Merit: 4331

Quote from: TracerX on June 04, 2014, 11:50:51 AM

Quote from: samsonn25 on June 04, 2014, 11:44:55 AM

Most psu can run over spec 10-20% for awhile, I dont know about the build quality of Enermax, maybe thats why they dont ship as complete units anymore.

Corsair and Seasonic Plats are pretty good.

I'm thinking about a Corsair 1300, but would hate to drop the cash if it won't really "buy" me anything. That said, what are the odds the upgrade blades BM has referred to will come it at lesser wattage?

I'll probably just bite the bullet and replace the PSU, my S2 slowly works down to 970 GH/s, even though only 4 chips are showing as disabled. I'm guessing my PSU is acting up.

minersource.net has a good deal on used EVGA1300 for $140.
I've got a new power board with black caps-anybody has this? Will take a picture later for comparison and will also try the new board with old psu, then try to change to EVGA1300 and report performance.

TracerX

hero member

Activity: 918

Merit: 1002

Quote from: ?? on ??

you might check your thermal grease. if its dried out or non-existent you may be getting increased errors - its something im looking into.

Let me know if you find out anything, re: thermal grease. Do we know of a way to monitor individual chip hash rate via SSH? If I could narrow it down to a specific blade, I would happily reapply thermal paste.

Topic: ANTMINER S2 Discussion and Support Thread - page 59. (Read 355892 times)