Topic: ANTMINER S5: 1155GH(+OverClock Potential), In Stock $0.319/GH & 0.51W/GH - page 54. (Read 451039 times)

its not the pci-e connector you should worry about..
i would NEVER run a miner with only 1 pci-e on each side... even tho its "in spec" this will probably happen.

all theoretical.

I have I PCi-e power meter they start to get hot over 150 watts at your specs they will be charcoal so your specs are theoretical no disrespect.  

all four connectors do not draw the same power level.

Here is the name of the power meter "Zalman ZM-VPM1"

Here is a link or just google it to find the cheapest price.

http://www.directron.com/zmvpm1.html

The actual connectors (per MOLEX datasheets) commonly used on PCi-E power connections are specified for 8 amps per contact (the contacts themselve are seperately rated for 11 amps but that is an OPEN AIR rating, you have to derate for using them in a multi-contact connector due to heat retention). This works out to the +12 connectors on either a 6-pin OR an 8-pin PCi-E connector maxxing out at 288 watts (which is where that maximum "Max PSU Power Unit" figure per connector/loop on the SP20 comes FROM).

 You're have to be overclocking a S5 to exceed that rating on one connector - but it's STILL a bad idea, as the boards aren't designed to be run from a single connector and you're likely to be overstressing some of the +12V traces ON the boards a lot by misconnecting the power leads.

 This is the reason Titans have so MANY power connector issues - they DID exceed the rating of the connector even at their STOCK clock, pulling over 300 watts per connector (the extra 2 pins on a PCi-E 8-pin connector are both GROUND leads, IMO a rather dumb thing for the PCi-E spec to have done - sure, the ground leads run cooler, but that doesn't appreciably help the +12 leads).


 Keeping the bottom of an S5 on a flat surface helps cooling overall a LOT - raising it up lets a LOT of the airflow escape without cooling anything. It's even a good idea to try to close off the open parts of the top to help cooling some, but that's less importans as the controller board blocks quite a bit of the top airflow excape route.

I've read somewhere that for the cooling system to work as it is intended you should lay the S5 on a flat surface, and indeed, i made the test of raising the S5 from the table and the thing went up 4ºC in an instant.

Also, those fans are out of spec, and you're running with only one PCI-e cable per board... that's a no-no, each PCI-e connector has a maximum rating that the S5 would exceed easily.

...  linoleum almost caught fire... This happened two months ago



Noctua
IndustrialPPC NF-F12  and Super Flower
Leadex Gold SF-1300F14MG

Abbot miracle. after a few weeks of testing to wake S5 ,!right now! awakened.(not first time happened) repeatedly reboot worked.
is not the first time that this series S5 collapse and then again from the fluke is started again. but as long as the internet again collapses or pool..  soo i recomended try repetat reboot or swich power on off.. 

But the problems are only from the new series S5 as shown in figure. there's probably something more problems but the longer I do not understand.
 older series are ok
I use second-hand. so I have no problem with warranty..

Can you provide a screenshot?

How long have you had it, what are mining conditions, psu, etc?   

Chain#   ASIC#   Frequency   Temp   ASIC status
1
30
350
115
xxxxxxxx xxxxxxxx xxxxxxxx xxxxxx
2
28
350
127
oxxooxxx xxxxxxxx xxxxxxxx xxxx

     
just  another .  I tried to reset it, attach another fan and nothing helps.

I recommend not to buy the whole series S5..

In all fairness, they might drop the price.  However, the trend seems to be, quick to raise the price, but slow to lower it.

Trust me they are not going to bring it down.

That rate has been that since the price re-increase ... and honestly, it is them that lose out if they value BTC more than it is (i.e if you are in the market for some, just convert $ to BTC and you are in the money!). Still, BTC value cooling SHOULD have meant a price drop ....

Bitmain... ಠ_ಠ

Come on guys... At least pretend to care at little bit better.





Edit: I mean, I suppose one is getting a "deal" with Bitmain erroneously valuing coin so highly, but where is the consistency ? Idonteven...

Go back to school I think you missed something.

Yawn boring you nimrod troll.

 If that transistor shorted out collector-to-emitter, the voltage on the output would be essencially zero.
 Also, ONE CIRCUIT does not prove anything - both a PNP and a NPN transistor can be used to pull an open-collector type output to ground, you just have to set up the INPUT to the transistor correctly to do so.

 I'm not "the idiot who said open collector", that's from the SPECIFICATION that you obviously still haven't bothered to read.



 BTW, where is the 555 timer in the circuit you posted? It seems to have gone missing....

your the idiot who said open collector not me. and yes transistors also can be used to change the pull up state state. were did you get your knowledge from a cracker jack box.



If that transistor shorted out the voltage would be equal to the voltage across R1 depending if the 555 timer IC fries or not.

Say no more very simple basic circuit but proves my point case closed end of conversation.

I really do think it is hit and miss on what one's do it.  It is not all, but obviously enough to cause a lot of talk.

Instead of waiting for fix I would get molex to fan connector and just not worry about it, if you have an effected unit.

You don't pull up an open collector line with a transistor, you pull it up with a resistor - which has NO polarity.

 Try reading the specification sometime, instead of raising strawman arguments with nothing to do with the spec then ASSuming that I have no knowlage of electronics.

Will Bitmain fix an issue with fans slowing down when S5 loses an Internet access?

They wont lower it they are price hogs like I said they are milking the cow.

Okay pull up huh , If its a PNP or NPN transistor what is the polarity of the voltage on the Collector when Each type is pulled up think about it.


If the collector was open On each type what would the polarity be.

Learn how the current flows through a bipolar transistor then you will know what I mean.




 

 Any fan that does not run 100% with the PWM line open is BROKEN (either by design or by failure), and does not meet the SPECIFICATION for PWM fans (the spec specifically calls for the fan to have an internal pull-up on that line and the DRIVER to the line to be open-collector).

 But yes, I did test it to be sure, having seen way too many items fail to meet specs over the years....