ANTMINER S5: 1155GH(+OverClock Potential), In Stock $0.319/GH & 0.51W/GH

[RichBC]

It is a decoupling capacitor for that node in the ASIC chain. Unfortunately if it has de-soldered this is probably because it has overheated, so it may be damaged as may be the ASIC's on that stage? One of the problems with a String design is that many faults can cause a one down all down situation.

Rich

Thanks.
This is not cause of overheating. this was happen because of bad shipment.

OK well that is good news, as you can resolder and it should be ok. You will need a fairly watty iron because of the heatsing effect of the copper on the circuit board.


Rich

[blindminer]

It is a decoupling capacitor for that node in the ASIC chain. Unfortunately if it has de-soldered this is probably because it has overheated, so it may be damaged as may be the ASIC's on that stage? One of the problems with a String design is that many faults can cause a one down all down situation.

Rich

Thanks.
This is not cause of overheating. this was happen because of bad shipment.

OK well that is good news, as you can resolder and it should be ok. You will need a fairly watty iron because of the heatsing effect of the copper on the circuit board.


Rich

Thanks for help.

[Mikestang]

When I try to ssh into my S5 and it asks for the password I input it, but it says access denied.  I know the password is correct, I log into the gui just fine with it.  Any idea why I can't ssh in successfully?

[zOU]

The credentials are root/admin in ssh, you may want to try these as they're the default.

They differ from the UI.

Good luck

[Mikestang]

The credentials are root/admin in ssh, you may want to try these as they're the default.

That got it, thanks!  Is it the same on the S1/S3?  I can't remember from when I ssh to them last.



I'll be curious to see if/how this affects performance.  After the reboot the api is reporting 100GH/s less than before, but I know bitmain's fork was suspect as to the numbers it would report...  I'll let it run for a while and see where it stabilizes.

**Edit - hash crept its way back up and stabilized at 1.29TH, just like with 4.8.0, but the error rate is lower than with the older bitmain fork, so hurray for improvements!

[AJRGale]

i wondered, if i got a S5, laid the 2 boards down (open them up like a book) with a 200mm fan over the top of the sinks, would it run cooler? does the other side need some air movement to be happy? other then the LM75As are there..

or is that a good question?

im going to hunt out some PWM 200mm fans and test this theory of mine...

[aarons6]

The credentials are root/admin in ssh, you may want to try these as they're the default.

That got it, thanks!  Is it the same on the S1/S3?  I can't remember from when I ssh to them last.



I'll be curious to see if/how this affects performance.  After the reboot the api is reporting 100GH/s less than before, but I know bitmain's fork was suspect as to the numbers it would report...  I'll let it run for a while and see where it stabilizes.

**Edit - hash crept its way back up and stabilized at 1.29TH, just like with 4.8.0, but the error rate is lower than with the older bitmain fork, so hurray for improvements!

the s5 is read only file system. did you make a new july rom with 4.9.0 in it?

i know NH has one but its the 4/15 firmware not the 7/15

[Mikestang]

the s5 is read only file system. did you make a new july rom with 4.9.0 in it?

i know NH has one but its the 4/15 firmware not the 7/15

No, I would not know how to do that.

I do not want to use the NH fork, I like it straight from the source.  I don't mind having to install it after reboots, I won't be rebooting it very often.

[zOU]

I have received my 2nd S5 earlier this week.

it is located in the same room (my attic) as my 1st one. (and just beside it)

However I noticed and 6C/10F temperature difference, and they're both "stock" (no additionnal fan, no fan limit)

[VirosaGITS]

I have received my 2nd S5 earlier this week.

it is located in the same room (my attic) as my 1st one. (and just beside it)

However I noticed and 6C/10F temperature difference, and they're both "stock" (no additionnal fan, no fan limit)

https://i.imgur.com/rjKduwm.png

If you're asking if this is normal, the answer is yes. They do the paste job sloppy and the sensor is cheap.

Also your temps are really low, you could reduce the fans by quiet a bit to save power/noise.

[zOU]

Thank you.

They're in a large, empty and not insulated attic, so plenty of fresh air and no one to hear the noise.

One of the two is vibrating a bit but I installed a piece of foam under it.

[VirosaGITS]

Thank you.

They're in a large, empty and not insulated attic, so plenty of fresh air and no one to hear the noise.

One of the two is vibrating a bit but I installed a piece of foam under it.

There's no gain in running them to 30C however. My hashrate seem to go down a bit when its too cool. I guess so would electricity usage. Well is you, i'd tone down the fan a bit to reduce strain, unless you expect temps to raise by a lot at some point in the day. Maybe 70% instead of default's i think 80%. I run my down to the minimum though, stock fan goes down to 2400RPM.

[zOU]

I've tried 80/50/20 %

And whatever I set, the fan speed reported in the Ui is always 3840 or 4080 rpm...

i'm going to try this:

https://bitcointalk.org/index.php?topic=902305.msg13623813#msg13623813

[Mikestang]

I've tried 80/50/20 %

And whatever I set, the fan speed reported in the Ui is always 3840 or 4080 rpm...

I noticed (and posted in another thread) that when you update the cgminer on the S5 to 4.9.0 it appears to ignore the user selected manual fan setting and it just runs the fans at whatever speed it wants.

[zOU]

I've tried 80/50/20 %

And whatever I set, the fan speed reported in the Ui is always 3840 or 4080 rpm...

I noticed (and posted in another thread) that when you update the cgminer on the S5 to 4.9.0 it appears to ignore the user selected manual fan setting and it just runs the fans at whatever speed it wants.

Oh right, I must have missed that !

Thank you Mike

[Mikestang]

One other thing I noticed, you can restart cgminer after the 4.9.0 update and it won't reset to 4.8.0.  It's only when you do a full reboot of the machine that it reverts back to 4.8.0.

[zOU]

one of my 4 S5 is underperforming, it also has quite a lot of HW errors...





the other 3 are hashing fine at 1.15Th/s

I tried the update to 4.9.0 but it makes no differences.
I had a couple of power outages recently so I did not do the upgrade after...

[aarons6]

one of my 4 S5 is underperforming, it also has quite a lot of HW errors...





the other 3 are hashing fine at 1.15Th/s

I tried the update to 4.9.0 but it makes no differences.
I had a couple of power outages recently so I did not do the upgrade after...

that is alot of hwe.. id check the power supply..

[zOU]

thx you aaron, I'll try to swap with another PSU

So previous PS was a IBM X3650 835W Redundant Power Supply, I swapped it with another PS and I still have HW errors. (can't remember the name/type/power of the new one and too lazy to go back in the attic)

I'm expecting a 1kW PS this week that I intended to use for 2 S3+ (but they're now running on the IBM PS mentionned above)

So I can use the 1kW for the S5 when it arrives

[Mikestang]

It's not the number of HW errors that matters, but rather the %.  Looks like you have 0.06% HW error which is about what my S5 sees overclocked at 400.  I noticed a drop of about 0.03% when I update to cgminer 4.9.0.