Bitcoin Forum
December 13, 2024, 01:45:10 AM *
News: Latest Bitcoin Core release: 28.0 [Torrent]
 
   Home   Help Search Login Register More  
Pages: [1]
  Print  
Author Topic: S17 APW9+ FAN problem  (Read 175 times)
Gabrics (OP)
Full Member
***
Offline Offline

Activity: 193
Merit: 124

Just digging around


View Profile WWW
August 07, 2020, 06:36:40 PM
Last edit: August 08, 2020, 03:20:13 AM by frodocooper
 #1

Hi Guys,

We have a few Bitmain S17 units and the 15K RPM FANs in the PSU started to die (~ 1y old). No problem (we thought). We ordered a bunch of 40x40x28mm 12V ~same FANs and installed them (originals and the replacement are both 12V are 2 wire no PWM type).

And here comes the surprise: Original FAN works from a given FAN header (inside the PSU). Replacement FAN doesn't work from FAN header (inside the PSU) (several PSUs and FANs tested).

But the replacement FANs do work from: 12V test PSU; from the 12V coming from the APW9 PCI-E connector.

Voltage measured on the FAN connector ~12V. Voltage measured on wire of the new FAN, ~12V. So the correct (looks correct) 12V is reaching the new FAN, but the FAN just won't work at all.

Now I am stunned. How can a simple two wire FAN not working if correct voltage applied (and working from the same PSU's PCI-E 12V)?

Any idea would be welcomed as these are relatively hard to come by and not that cheap either.

Original FAN: Nidec UltraFlo  W40S12BMD5-01Z90.

Replacement FAN: Sunon PDM1204PQBX-A Double Ball Fan 12v 8w Fan 40x40x28mm.

Thanks.

PS: Please also advise if you know any online store selling two wire 40x40x28 15k RPM fan's and not shipping from China.
HagssFIN
Legendary
*
Offline Offline

Activity: 2506
Merit: 1714


Electrical engineer. Mining since 2014.


View Profile WWW
August 07, 2020, 07:21:33 PM
 #2

Did you measure the voltage in the fan connector in the psu? What was the result?

Maybe the problem is not in the fan, but in the psu circuit.

euhctcyn
Newbie
*
Offline Offline

Activity: 28
Merit: 14


View Profile
August 07, 2020, 08:29:23 PM
Last edit: August 08, 2020, 03:21:32 AM by frodocooper
 #3

The one that you bought should work, is pretty equivalent in terms of CFM/performance/specs:

https://imgur.com/a/TpJJVgG – First graphic is the Sunon specs, second one the original Nidec.

PS: Please also advise if you know any online store selling two wire 40x40x28 15k RPM fan's and not shipping from China.

Why 15k RPM? Afaik the 17 series fans work @ 7k RPM at max.
Gabrics (OP)
Full Member
***
Offline Offline

Activity: 193
Merit: 124

Just digging around


View Profile WWW
August 08, 2020, 02:24:20 PM
Last edit: August 09, 2020, 12:13:07 AM by frodocooper
Merited by frodocooper (3), HagssFIN (2), Artemis3 (1)
 #4

Original FAN works at 15K: http://www.equipment-coolingfans.com/sale-12327231-12v-15000rpm-utra-flow-axial-fan-plastic-frame-dc-nidec-type-w40s12bmd5-01z90.html

Voltage on the FAN pins is a good 12V. 12.23V to be precise.

Yepp, it should work. And it doesn't Sad. We a bought like 50 and none of them work tested in 6 PSUs. I am thinking that they (Bitmain) may pulse the voltage on the FAN pins and that somehow doesn't work with the Sunon replacement FAN (we don't have an oscilloscope at the shop).



We found the problem finally. The FAN pins (two pin one way header!) 12V PINs inside the APW9+ are reversed on the PCB. The original FAN works with the reverses polarity, but the new fan only works if the polarity is correct. Weird.
euhctcyn
Newbie
*
Offline Offline

Activity: 28
Merit: 14


View Profile
August 09, 2020, 10:20:52 AM
 #5

Good to know, thanks for sharing!
Pages: [1]
  Print  
 
Jump to:  

Powered by MySQL Powered by PHP Powered by SMF 1.1.19 | SMF © 2006-2009, Simple Machines Valid XHTML 1.0! Valid CSS!