Topic: BITMAIN Antminer S3 support and OverClocking thread - page 26. (Read 158132 times)

getting errors, hashing stops thread here if anyone can help

https://bitcointalksearch.org/topic/antminer-s3-error-bitmain-1-1110-usb-bitmain-asic-0-now-disconnected-750439

Multipool Mining:

I got this from Bitmain.


--Balance: In balanced mode, the amount of diff1 solutions per pool is monitored as a rolling average per 10 minutes and if pools start getting more, it biases away from them to distribute work evenly. The share count is reset to the rolling average every 10 minutes to not send all work to one pool after it has been disabled/out for an extended period.

--Load Balance: Change multipool strategy from failover to even load balance.

--Failover only: Don't leak work to backup pools when primary pool is lagging.


When I use Load Balance I have 3 pool slots used.  2 are the same pool, and one is for a different pool.

I get about 80-gig on the single-slot pool, and the rest ( 360 gig ) on the double-slot pool.

I have it set to:

Load Balance, with...

Slot 1 - EclipseMC
Slot 2 - Eligius
Slot 3 - EclipseMC

Eclipse gets the 360 gig.


It does not like pools that update their hashrates quickly.  So you have to ignore the 5s and 15min figure.  It's the 1 hour figure you have to look at.

And it hates GHash when mining 2 pools ( or more ) at once.  It really messes up the stats.   Loads of low difficulty stuff, and it tricks GHash into thinking it's mining at 1TH or more.  A real mess.


Apart from that it's fine.



Clock-wise I use 2 cheap-ass supplies in a case mod ( because of the 12v rail issues ).  Each supply can put 26A minimum down the 12v.  I use 440w at the wall ( as they are cheap, so easy to replace if they blow; and there are 2 of them ) to get 440 GH/s.

If I dial back the clock to 200m I get 1.05 GH/w at an average of 400 GH/s.

If I dial the clock back to 175m I get 1.17 GH/w at an average of 340 GH/s.


Which is the same as if I dial down my Bitmine.ch Coincraft Desk.  Certainly interesting stuff, and it means that as difficulty rises you can adjust the power factor and squeeze a few more months out of your miners before you need to upgrade.

Power to hash ratio is adjustable, and that is certainly good news in this climate ( what the hell happened to BTC price of late?  Is this the Silk Road sale knock-on effect happening? )

i did mine, my hwe was in the 1000s on one machine

there was a few chips that didnt have much paste on it..
i used thermaltake tg-3

its what the store by me had in stock..

they dont really run any cooler but i get less hw errors..

i guess if you have problems it wouldnt hurt, but if not i wouldnt worry about it.


about the pool, ghash.io is the largest one i think.

but id try more then one out and find the one that works best for you.

Just got my S3 today its batch 7
I am running a corsair 1050x PSU
So I overclocked this for about 5 mins
I hit like 514 GH/s
The fans were running like 2080 rpm and the temperature started to climb to 44c
So I backed it off to stock settings

I live in florida And i keep the house at 78 degress F in my office its quite a bit warmer mostly due to the fact of this butterfly labs single 50 gh/s that runs at around 82 deg C  
Then my PC and these power supplies
I think i need a seperate AC for this room before I attempt anymore over clocking Just got it I think I will wait till its turned a profit before i chance something going horribly wrong.

I am drawing out some ideas for a heat removal system thinking i will take some dryer vent hoses and attach them to the back of the butterfly and antminer and port them out the window Wife is just going to love that haha.

I am going to buy a couple more of these units if i see any real ROI, imagine the heat

But for future knowledge
Does removing the case improve the cooling on these s3's?
Would it be a good idea to go ahead and replace the thermal paste with some high dollar stuff? anyone got any recommendations? What kind of paste is on the chips now?

And lastly What is the best pool to mine in? I know that one will have a different anwser from each person I like the interface of https://www.minep.it but they dont seem to be a very popular pool. I am new to this mining stuff so any and all input would be greatly appreciated.

Thanks

all you have to do AFTER you flash the new FW is go into miner config, advanced pick a speed and hit save and apply.
it will come back up and be at full speed..

Has your speed come back up? I flashed mine to the newest v3 firmware with the advanced setting and I see the same thing. 100G at 235.7 frequency that was set on flash. Did you end up setting it to default and getting the speed back before the flash? I'd hope that they'd have something in there to say if your not getting "x" that the miner tone itself back.

I hope installing this firmware doesn't count as overclocking since that was not my intent.

Just FYI manually setting back to default seems to get the speeds back where they are supposed to be.

yeah i think im going to need another power supply.
mines 70a so 840w

with 2 miners, one is a 225 and one is at 218 the psu is pretty much maxed out..

I went into the miner config>advance settings tab, to make the change.  Putty not needed.

post your asic-freq file from \etc\config.  if you dont have one (since you said you renamed the file to .old) then that's your 1st problem.

many people dont understand the power supply watt rating.  just because a power supply is 650 watts doesnt mean it can provide that much power on the 12v (which is what the miners use).  in order for you to know how much 12v power your power supply is capable of providing.  you can either look at it in terms of amps or watts as most power supplies show both.  if it only shows the amps available to the 12v rail you can calculate the watts it will provide by multiplying that number by 12v (amps x volts = watts).

take the Corsair 850RM for example - http://www.newegg.com/Product/Product.aspx?Item=N82E16817139011.  if you click on the link and click the specifications tab it tells you it has a max of 70A available for 12v.  70 x 12 = 840 watts total.  the rosewill 850 watt power supply cannot provide that much wattage for the 12v rail - http://www.newegg.com/Product/Product.aspx?Item=N82E16817182072.  note that they show 64A max for 12v which only comes out to 768 watts total.

as some have suggested just because the power supply CAN handle 840 watts you wouldnt want to have it running at 100% load (kinda similar to always running your car at max RPM) as that will cause it to fail more quickly, especially if you are using a cheap-o brand PS.

so bottom line - do your homework before thinking an 850w PS can run 2 overclocked S3's because the power usage is supposed to only be 375w per miner.

I rebooted and no change at all.  Do I need to upgrade the firmware?  It was working fine until I updated the firmware.  Then I didn't get any speeds over 100G on both of them.  I reset the miner back to factory and started over.  Now it's mining at 218 on the GUI, but 250 when I ssh into the miner. 

this is interesting

I had problems with the asic-frequency file being used, even though I was using the new advanced frequency setting in the GUI. I just renamed asic-frequency to .old then rebooted.

Did you remember to power cycle or type reboot in putty? I am not sure if these settings can be picked up simply by restarting cgminer, but you can try that too, else, after changing the values in the asic-freq file, you have to restart the rig.

ill try the 600w for a few and see if the hw errors go down.

i dont mind them at 237 they run good at that speed.

SSH into the S3 and un-quote to your required frequency. If you are not familiar with this process, read earlier in this thread, it is well explained.
I am not conversant with changing frequencies in the new firmware as I see no need to update mine, they are not broken so why fix anything?

Your PSU will run both S3's confortably at 225. If you want to overclock above that, then you'd ideally put each S3 on a separate PSU that can supply the required power. If I was in your place, I'd be happy to run both off the single 850 or even use the spare one and crank up the speeds on both. Getting another 850 PSU would NOT be on my radar.

On a side note, I run my S3 overclocked to 256.25 (approx 420 watts) at the moment off a 550 watt server PSU and it is running OK for a day now with just 140 HW errors. I am planning on cranking it up to higher speeds since my PSU can pump out approx 440 - 460 watts, so in your case, the 850 for each will be a bit of an overkill.

hmm it seems i am right at the edge then.

i have 850 gold psu.
so if i run them at 225 that would be better..

how do i add 225 to the menu?

a few pages back it says what file to edit, but how do i get to that file?


i do have an old 600w psu laying around, think that would be better to use for one of the until i can get another 850?
its not the best psu tho, its a ocz brand

General rule of thumb: Make sure the power draw exerted on the PSU is within 80% of its rated output. If the PSU is a gold or platinum, I'd stretch that to 85%. This is for both the longevity of the PSU and the ability to efficiently power-up your rig.

In your case, the PSU is rated @ 850 watts * 0.8 = 680 watts
Thus should have each of your S3's setup at speeds that will consume a maximum of 340 watts each.
According to info on this thread, @ chip_freq '225', power draw is approx 346 watts and at chip_freq '237.5', power draw is approx 375 watts.

So now you know why you are getting a high HW error rate. You gain NOTHING from HW errors, but in the medium to long run, you could lose some of your chips.

i have a question, i have 2 s3s. i think batch 6.

anyways, i updated the FW of them to the newest one and they are happily running away at 237Mhz @ 478Ghs.

im running them with a single Seasonic x850 PSU.. each one has 4 PCI-E plugged in.. 4x2 pci-e plugs. no adapters or anything..

i was wondering if i should get another psu and run each on their own psu or if 1 850 is ok.. its been about a week with no problems.


also, i was wondering if i should edit the new menu and add 225?

i dont know how to do that but if 1 850w psu with 2 s3s running at 237 is too much maybe running them at 225 is better?

i get a few hw errors.. about 50 per hour.. one gets more then the other.. i know its not a lot, like .005% but could the psu be causing it?

I had made one video hope to help you!

https://www.youtube.com/watch?v=90aaFWLIdaw

Noticed that new firmware also resets the "queue" back to 4096, so if you lowered that to reduce discards you will need to mod that again in:

vi /etc/init.d/cgminer

been running a B3 @ 256.25 for 30 minutes will update after a few hours..