Topic: ANTMINER S2 Discussion and Support Thread - page 150. (Read 355730 times)

Guys, we're the Bitmain community here on the forums, not the BFL, KNC or HashLast. Let's keep it civil, let's avoid name calling and negativity. I'm here trying to catch up and getting ready to respond and I see negative comments, it's distracting and just puts a bummer on things. Keep it cool? Answering questions next :-)

Taking out 1 blade makes you pull 10% less power, thus if your power supply is struggling at full load, removing 1 card could give it the margin it needs. When the enermax PSU goes into overload, to power it back on you'll need to have it plugged in and on and remove the jumper, count to 5 and put them jumper back in. This requires the case to be removed...again lol

Meanwhile, you can use your S2 as a door stopper ..... 

Hey sucker, pay up or forget mining if its too hard for you.

No.

If you read through this thread, you'll find i've asked many questions to which there have been many generous answers, and i've shared my experiences so far.   i want $50/hr to provide one-on-one support.  If i'm unable to make any progress, i won't charge you.  
I am a professional IT support consultant.  This is what I do and i am rather good at it.  If you've made it to page 70 of this thread and still have problems, I may be able to help.

You find this unreasonable?

Batch 2 have gold psu
And not platinum psu

i am allergic to my scroll button and therefore cannot check previous posts on this thread.
take my money.

Setup Support for AntMiner S2
----------------------------------


As the owner of 3 batch 1 S2's, i've had a couple of weeks to work through some of the issues involved in getting these setup and running.  Virtually everything I know about the S2 i've either learned from others here in this thread or i've described it myself in this thread.  I'm not BitMainTech nor do i work for them - I'm just an IT guy.

If you are struggling to get your S2 working and want a helping hand, I may be able to help.

I can share your screen with join.me and video conference with Skype.

0.1 BTC per hour, but only if I'm able to help.


Email me to setup a time.  I'm in Toronto, Ontario (eastern standard timezone)
[email protected]
-Dale

so we are still not that ready for prime time on these?

those plastic things did nothing for me, 60% of them were broken completely when mine arrived and all the cards were loose and some had even bent the pci slots, 50% of the pci slots needed bending back. did all this and then found one of the fans is missing a piece of a blade, broken in transport like almost everything else.

replaced fan with a spare (well one i had to pull out of my everyday computer), then had firmware problems, finally managed to get logged in and changed pool settings, starts and mines for a while but after less than an hour blade 1 got really hot, like oven hot, strange how this was on the side of the case that didn't have a broken fan, hashed at around 1000 for a bit but majorly unstable, at the moment the unit is only running with half the blades in and gets about 500, this is not what i paid for.

How does one go about an RMA? I want a replacement

Sorry, but newb question, i have gone out and purchased a class 10 spare, but i havent the slightest clue on how to image that SD card and have it ready in case the one from Bitmain croaks. Can  anyone give me the procedure ? 111 Thanks so much for making a new image !

I followed some advice earlier in the thread and removed one blade, it's now been mining for a couple of hours so far.

These batch 2 units are a lot quieter than batch 1, with larger effective fan diameters and case holes, they also seem subjectively cooler.

can you ping the device?

ps. clear cache?

no it is actually machine oil you are smelling, probably left over from machining all those aluminim fins for heat sinks. I had one machine that had some oil on the blades, case and mainboard. Nothing alittle alchohol couldn't handle, and look in the twisted cables of the fans, oil likes to hide there. And yes, push down all boards, or better yet pull them all out and inspect every cable as I did find some loose and/or outright disconnected. Happily mining with 29 of 30 blades/hashboards and 1 spare PSU. Already RMA'd the blade, PSU is next.

Got my batch 2 S2 a few hours ago - this is exactly what I did.  Remove screws by hand, don't use a drill, to avoid stripping screw heads.  For the stubborn screws, do as per above and pry the case gently to provide pressure while unscrewing.  All my stubborn screws came out this way, and my case is undamaged.

Visual inspection of the inside showed everything was ok.  Each blade has a connector to the motherboard at the base, and is supported on each end by plastic sleeves (I think the sleeves are new for batch 2).  Much better design - prevents the blades from becoming unseated during transport.

When I turned it on it wouldn't hash (0 GH/s shown on LCD screen).  I took out every card, and closely inspected everything.  What I found was that the ethernet cable which connects the BBB to the external network port was loose at the network port end.  I re-seated the n/w cable, put all the blades back in, and fired it up.  Voila!  It started hashing!  I know other people have had 1 TH/s (and above) miners for a while, but this is my first unit, and I'm quietly amazed to look at this box and consider that it has 1 TH/s of hashing power inside.  When I first heard about bitcoin (April 2013), the entire network hashrate was only 80 TH/s.

My S2 has been mining for an hour and a half.  The hash rate shown on the LCD fluctuates between roughly 950 to 1100 GH/s.  At the pool so far I'm seeing 968 GH/s (and slowly rising), I expect this to settle in closer to 1 TH/s when it has been running longer.  At the wall it is draws 1039 Watts max (usually 1 or 2 W less than that).  Temperature is 53 C (not sure what the room temp is, but I feel a bit hot).  In the Miner Stats, for some reason chain 4 shows a temperature of 0 (no sensor / faulty sensor?).

I had it mining with the case off originally, putting the case on caused the fans to reduce speed.  I guess the wind-tunnel affect of having the case on improves cooling, so the fans don't need to work as hard.  

After I took the case off I could smell a light machine oil / grease odour.  I don't find it unpleasant, just noting it.  I think it happened because I pulled all the cards out.  With the case back on now I don't smell anything.

tek eet ezay tek eet ezay.....no need for ripping nor drilling. Patience and breathe. Remove all screws that cooperate. Then gently pry with blade screw driver, butter knife or similar tool under cover near unfriendly screw with slight pressure while turning screw out, and proceed to next one. Most if not all uncooperative screws should come out this way.  

I've got four Batch 1 units running, uptime is steady and problems have been minimal so far,

I'm just unpacking two Batch 2 units, the first one started up, network cable internally needed reseating and one card was slightly out of its guides.

It started hashing to my pool when I changed the settings and I thought all was good.

After about 10 mins the whole unit powered off and has been unresponsive ever since.

I've unplugged all outputs from the PSU and it's still not coming back up, so it looks to be the PSU.

Any ideas on what to try to resurrect? I have multiple CX750M PSU's but nothing big enough to replace the 1000W Enermax.

Cheers

They are hashing on your S2, Dogie...sure, everything is fine...for them...

I've tried to  ask something about overclocking but I was ignored....

Don't care, try it again....

I've asked 111 that he tried to oc with the stock 225M clock setting in the asic-freq.conf? I think the clock numbers and timeouts are not for this machine.

If you have a look at the timeout at 196M I think you realized that the timeout is '40'. So if you use the same parameters as they left in the config, the timeout will be '70'.

196M   '40' timeout

225M   '70' timeout

 

Think it's no good, as you make higher clock you need more less timeout.

I've tried the stock '0781' with timeout 40-35-30-25. No succes. At 25 the cards stopped working so don't try....

I made a calculation it looks like '5f06' is 200M not 196 but this is not for sure, it was late night so I'll calculate again today just to be sure.

So in my opinion the values in asic-config are useless except the stock settings.

I'll try to calculate some clocks later today.

Sidenote: These are just my ideas, maybe a rocket engineer knows it better. 

I literally had to RIP the fucking case off.... was not cool, but I was able to push from below once it was open and now I rebent the case and rescrewed in (some) of the screws...