Pages:
Author

Topic: Pair of s-9s now running photos are up. - page 5. (Read 7751 times)

legendary
Activity: 1096
Merit: 1021
I'm confused.  Now they are giving us two temp readings.  In the past, for the S7 and others have we just been seeing the pcb temp (labeled: Temp).
Then why now they showing us the chip temp.  Why is that important now?  Poor design?
My pcb temps are running 56, 56, 59 at 625M.  On any other Antminer I would say there is plenty of room for some over-clocking.  Or in this case to move to the default 650M. (EDIT)
Chip temps are 87, 86, 91.  Are these temps higher than an S7 with the above pcb temps?  In other words is there more of variance on this miner than the older ones.  Where we can burn out a chip before the pcb gets to 80C.  That's still the safety shutdown temp. I believe.
Inquiring minds want to know.

PCB temps are useless...just a cheap way to guesstimate what the die temps would be. My guess would be that with more real-estate afforded by 14/16nm it became worth it it put temps sensors on the dies, and the fact that chips at these node levels are more sensitive to high temps. Anything over 90c continuous on die temps will shorten the life of the chip. I would be concerned if they are running over 80c.

if 80c on the chips is bad then the s-9 should be clocked at 575 vs 650 as to go to 79c will not be easy in rooms without ac.

i went to 625 but may indeed go to 600.

mean while i fixed the mad whistle


https://bitcointalksearch.org/topic/m.15223731
https://bitcointalksearch.org/topic/m.15223315


I take it that the filter doesn't block that much air flow?  I'm going to have to pick some of this up for my S7 that still whistles once in a while.
legendary
Activity: 4256
Merit: 8551
'The right to privacy matters'
afaik, chips max temp is 125 deg C then it starts to fail.

well thats for s7 chip though but i'd assume should be the same.

i think other components will fail before asic chip.

like i said, below 75C pcb temp should be pretty safe.

keep it cool Wink

4.2 Recommended Operation Conditions

TOPT Operation Temperature  125 ℃ (MAX)

https://cnshop.bitmain.com/files/download/BM1385_Datasheet_v2.0.pdf

bm1384 & bm 1385 pretty close, same value.

i noticed, pcb temp is roughly 30 deg C lower than chip temp so that can be used as a guesstimate imo.

so the built in 80 deg c cut off (pcb) +30 c (guesstimate) = 110 c on chip so should still be ok.

just my 2 cents.

philip, some screenies pls.


  sorry real world is in the way.

doing a full kitchen reno gutting to the studs.

won't post as much for next 3 weeks.
legendary
Activity: 1500
Merit: 1002
Mine Mine Mine
afaik, chips max temp is 125 deg C then it starts to fail.

well thats for s7 chip though but i'd assume should be the same.

i think other components will fail before asic chip.

like i said, below 75C pcb temp should be pretty safe.

keep it cool Wink

4.2 Recommended Operation Conditions

TOPT Operation Temperature  125 ℃ (MAX)

https://cnshop.bitmain.com/files/download/BM1385_Datasheet_v2.0.pdf

bm1384 & bm 1385 pretty close, same value.

i noticed, pcb temp is roughly 30 deg C lower than chip temp so that can be used as a guesstimate imo.

so the built in 80 deg c cut off (pcb) +30 c (guesstimate) = 110 c on chip so should still be ok.

just my 2 cents.

philip, some screenies pls.

legendary
Activity: 1161
Merit: 1001
Don`t invest more than you can afford to lose
Really nice Smiley put them on a colder place.
legendary
Activity: 2408
Merit: 1102
Leading Crypto Sports Betting & Casino Platform
I'll leave other to post pics. Got mine today and so far am - happy - with it but...
A. is so far averaging just over 13THs
B. A thermal design issue raises it head with these. The middle card and the one with long finned heatsinks facing the case wall run well over 4.3-.5 THs each. They report chip temps of around 90C. The one with the short fin topsinks facing the case is giving a bit over 3THs and reports the chip temp as banging around 99-100C. The air coming out of that card is also substantially warmer than the other 2. Will measure temps tomorrow sometime.

The root problem is that the short sinks are presenting a much high airflow resistance because the flow channel is restricted by them and the case wall. If somehow move the cards over a smig, hell even 1/8" might do it, then airflow should remarkably improve.

yes short term I down clocked from 650 to 625.

But UPS was late.
I had to dissemble the avalon6s from the array.
My sprained ankle from 65 days ago still hurts. ( bitch to be 59 it would have healed in under 10 days when I was 25-30)
The mining gear is in a loft with a ladder.

First time I turned them on I climbed down the ladder knowing they were mining for bitmaintech. So we went to the pc to check them and  setup kano.is

temps were very high   chips in the 90's






only 2 machines and a large loft  I was supposed to point miners in the direction in this photo
about 70 feet and an up draft fan in the center


Baller alert , you have one of those EGO eclectic mowers how do you like it ?
legendary
Activity: 1593
Merit: 1004
I'm confused.  Now they are giving us two temp readings.  In the past, for the S7 and others have we just been seeing the pcb temp (labeled: Temp).
Then why now they showing us the chip temp.  Why is that important now?  Poor design?
My pcb temps are running 56, 56, 59 at 625M.  On any other Antminer I would say there is plenty of room for some over-clocking.  Or in this case to move to the default 650M. (EDIT)
Chip temps are 87, 86, 91.  Are these temps higher than an S7 with the above pcb temps?  In other words is there more of variance on this miner than the older ones.  Where we can burn out a chip before the pcb gets to 80C.  That's still the safety shutdown temp. I believe.
Inquiring minds want to know.

PCB temps are useless...just a cheap way to guesstimate what the die temps would be. My guess would be that with more real-estate afforded by 14/16nm it became worth it it put temps sensors on the dies, and the fact that chips at these node levels are more sensitive to high temps. Anything over 90c continuous on die temps will shorten the life of the chip. I would be concerned if they are running over 80c.

if 80c on the chips is bad then the s-9 should be clocked at 575 vs 650 as to go to 79c will not be easy in rooms without ac.

i went to 625 but may indeed go to 600.

mean while i fixed the mad whistle


https://bitcointalksearch.org/topic/m.15223731
https://bitcointalksearch.org/topic/m.15223315


I let mine run overnight at 650M when it's especially cool in the mine.  Even then my die temps were in the upper 80's, with PCB temps in the mid to low 50's.

But when you say anything over 90 will shorten the die's life, are you speaking generally about asic chips?  Or specifically about asics designed to mine crypto?  The reason I ask is that these chips do not need to last 3-4 years.  Even with the S9 being by far the leader right now, I can't imagine it being profitable 18 months from now.
I would not want to keep a miner throttled back 10% below "manufacturer's default" speed just so the chips will last 3-4 years.  I just need them to last about 18 months at the most.  Even that would be the longest life cycle for a miner since asics.  And I want to suck the life out of them for those 18 months.  No sense leaving coin on the table.
legendary
Activity: 4256
Merit: 8551
'The right to privacy matters'
I'm confused.  Now they are giving us two temp readings.  In the past, for the S7 and others have we just been seeing the pcb temp (labeled: Temp).
Then why now they showing us the chip temp.  Why is that important now?  Poor design?
My pcb temps are running 56, 56, 59 at 625M.  On any other Antminer I would say there is plenty of room for some over-clocking.  Or in this case to move to the default 650M. (EDIT)
Chip temps are 87, 86, 91.  Are these temps higher than an S7 with the above pcb temps?  In other words is there more of variance on this miner than the older ones.  Where we can burn out a chip before the pcb gets to 80C.  That's still the safety shutdown temp. I believe.
Inquiring minds want to know.

PCB temps are useless...just a cheap way to guesstimate what the die temps would be. My guess would be that with more real-estate afforded by 14/16nm it became worth it it put temps sensors on the dies, and the fact that chips at these node levels are more sensitive to high temps. Anything over 90c continuous on die temps will shorten the life of the chip. I would be concerned if they are running over 80c.

if 80c on the chips is bad then the s-9 should be clocked at 575 vs 650 as to go to 79c will not be easy in rooms without ac.

i went to 625 but may indeed go to 600.

mean while i fixed the mad whistle


https://bitcointalksearch.org/topic/m.15223731
https://bitcointalksearch.org/topic/m.15223315
legendary
Activity: 2174
Merit: 1401
I'm confused.  Now they are giving us two temp readings.  In the past, for the S7 and others have we just been seeing the pcb temp (labeled: Temp).
Then why now they showing us the chip temp.  Why is that important now?  Poor design?
My pcb temps are running 56, 56, 59 at 625M.  On any other Antminer I would say there is plenty of room for some over-clocking.  Or in this case to move to the default 650M. (EDIT)
Chip temps are 87, 86, 91.  Are these temps higher than an S7 with the above pcb temps?  In other words is there more of variance on this miner than the older ones.  Where we can burn out a chip before the pcb gets to 80C.  That's still the safety shutdown temp. I believe.
Inquiring minds want to know.

PCB temps are useless...just a cheap way to guesstimate what the die temps would be. My guess would be that with more real-estate afforded by 14/16nm it became worth it it put temps sensors on the dies, and the fact that chips at these node levels are more sensitive to high temps. Anything over 90c continuous on die temps will shorten the life of the chip. I would be concerned if they are running over 80c.
legendary
Activity: 4256
Merit: 8551
'The right to privacy matters'
Yes  this is for 2 miners at freq 625 fans on 70%.


almost at 28th  with 625 freq




3 blocks so far.  A long way to go for roi

sr. member
Activity: 464
Merit: 301
Phil do you have overnight numbers for the miner ?
legendary
Activity: 4256
Merit: 8551
'The right to privacy matters'
I'm confused.  Now they are giving us two temp readings.  In the past, for the S7 and others have we just been seeing the pcb temp (labeled: Temp).
Then why now they showing us the chip temp.  Why is that important now?  Poor design?
My pcb temps are running 56, 56, 59 at 625M.  On any other Antminer I would say there is plenty of room for some over-clocking.  Or in this case to move to the default 650M. (EDIT)
Chip temps are 87, 86, 91.  Are these temps higher than an S7 with the above pcb temps?  In other words is there more of variance on this miner than the older ones.  Where we can burn out a chip before the pcb gets to 80C.  That's still the safety shutdown temp. I believe.


Still confused.

when the pcb hit 80c it should shut down


There is zero info on chip and what is bad.

kano.is is  26.31th to 27.13 th  at this moment.

thesee two machine at 7th more the the network was when I started in summer of 2012 !

  network was about 20th
legendary
Activity: 1456
Merit: 1000
Damn that's sweet stuff.  Always fun reading up on the solar array you have going Smiley

I would agree  very cool to see 2 S9's connected the the array.   I like the loft how does it do with dust out of curiosity?
legendary
Activity: 1593
Merit: 1004
I'm confused.  Now they are giving us two temp readings.  In the past, for the S7 and others have we just been seeing the pcb temp (labeled: Temp).
Then why now they showing us the chip temp.  Why is that important now?  Poor design?
My pcb temps are running 56, 56, 59 at 625M.  On any other Antminer I would say there is plenty of room for some over-clocking.  Or in this case to move to the default 650M. (EDIT)
Chip temps are 87, 86, 91.  Are these temps higher than an S7 with the above pcb temps?  In other words is there more of variance on this miner than the older ones.  Where we can burn out a chip before the pcb gets to 80C.  That's still the safety shutdown temp. I believe.


Still confused.
legendary
Activity: 3892
Merit: 4331
I guess with these high temps we can assume that the fans are as loud or even louder than the S7s

loud as the s-7  and I have one unit that squeals like a stuck pig.

I know there is a fix for that if anyone had a squealing s-7 and wants to post the fix please do so.


my low tech solution to squealing s7 was to find out that fans screws to machine were over-tightened to the point of deforming fan's case.
i eased up one screw at the time starting from the closest to the point where I saw distortion. Half a turn-and it was OK-no squealing.
On one other machine this only helped a bit (partially), but after a day or two squealing went completely away anyway.
I guess that some part just got grinded or something.
legendary
Activity: 2294
Merit: 1182
Now the money is free, and so the people will be
Damn that's sweet stuff.  Always fun reading up on the solar array you have going Smiley
legendary
Activity: 4256
Merit: 8551
'The right to privacy matters'
I guess with these high temps we can assume that the fans are as loud or even louder than the S7s

loud as the s-7  and I have one unit that squeals like a stuck pig.

I know there is a fix for that if anyone had a squealing s-7 and wants to post the fix please do so.



My guess is we will mine at 600 to 625 with the s-9's until the fall  but no pressure at the moment to decide.

maybe simple box fans will cure all cooling as we only have 2 units with two on the way.



I had an S7 that the fan squealed.  The trick was to move a small piece of paper around the intake fan.  You keep moving it around until the you find the spot that the squeal stops and then you tape it there.  Another solution that worked for a bit was to loosen the screw a little bit.

Do you know what the ambient temps were when you got the screen grabs? 

I would think  it was 88f  because it is a loft.

I am not sure the safe chip 0c number ,

I know that

 spondoolies sp20 chips throttle at 125c

intel cpus' throttle at a 72c junction point could be wrong on this

amd gpu's throttle around 93c


but  the readings may not mean much.

the pcb number of 65c should be good
i do not like to see a chip number near 100c
i was near that at 97 c

but the fans are now pointing at the far wall not the close wall
the fans are set at 70% = 4300 rpm
freq dropped to 625 vs stock 650.

here is a screen shot of the two at kanos pool live time 10:10 pm

at freq 625 we are getting 26.82-27.33 th
two units = 4300 usd better safe then sorry Grin
legendary
Activity: 3822
Merit: 2703
Evil beware: We have waffles!
I'd go with the 1/4" thick stuff  which is what I used.
legendary
Activity: 4256
Merit: 8551
'The right to privacy matters'
I guess with these high temps we can assume that the fans are as loud or even louder than the S7s

loud as the s-7  and I have one unit that squeals like a stuck pig.

I know there is a fix for that if anyone had a squealing s-7 and wants to post the fix please do so.

re-read my post https://bitcointalksearch.org/topic/m.15212835 for the edit about that... The effectiveness varies with fan speed but helped immensely here. Should apply to s7's as well.

thank you  I will pick some up at lowes


http://www.lowes.com/pd/Frost-King-13-5-in-x-24-in-x-0-0312-in-Polyurethane-Foam-Air-Conditioner-Filter/3055595
legendary
Activity: 1096
Merit: 1021
I guess with these high temps we can assume that the fans are as loud or even louder than the S7s

loud as the s-7  and I have one unit that squeals like a stuck pig.

I know there is a fix for that if anyone had a squealing s-7 and wants to post the fix please do so.



My guess is we will mine at 600 to 625 with the s-9's until the fall  but no pressure at the moment to decide.

maybe simple box fans will cure all cooling as we only have 2 units with two on the way.



I had an S7 that the fan squealed.  The trick was to move a small piece of paper around the intake fan.  You keep moving it around until the you find the spot that the squeal stops and then you tape it there.  Another solution that worked for a bit was to loosen the screw a little bit.

Do you know what the ambient temps were when you got the screen grabs? 
legendary
Activity: 1593
Merit: 1004
I'm confused.  Now they are giving us two temp readings.  In the past, for the S7 and others have we just been seeing the pcb temp (labeled: Temp).
Then why now they showing us the chip temp.  Why is that important now?  Poor design?
My pcb temps are running 56, 56, 59 at 625M.  On any other Antminer I would say there is plenty of room for some over-clocking.  Or in this case to move to the default 650M. (EDIT)
Chip temps are 87, 86, 91.  Are these temps higher than an S7 with the above pcb temps?  In other words is there more of variance on this miner than the older ones.  Where we can burn out a chip before the pcb gets to 80C.  That's still the safety shutdown temp. I believe.
Inquiring minds want to know.
Pages:
Jump to: