Delta PSU fan replacement woes

grimwood · 26th October 2004, 4:34 PM

Hi All,

I have a Delta server PSU, forgot what model number, but it is SSI and has two 350W hotswap units in it. It has two 80mm fans in it and they are loud, at 38dBA. They also run at low rpm while the machine is off. They are 3-pin Delta AFB0812VH fans, rated at 7-13.8V. I want to replace the fans with my quiet 3-pin Papst fans.

First thing I did was simply replace the Deltas with the Papsts. The Papsts didn't fire up either when the machine was off or powered up. Put the Deltas back in and they worked.

Although they have 3 pin connectors, the Papsts have only two wires. I figured it had something to do with the PSU wanting rpm monitoring, so soldered on an extra wire to the fans to make them 3-wire and tried again, still nothing.

Just in case I botched the soldering or the Papst's can't do rpm monitoring, I grabbed some 60mm Globalwin fans from I think Fop-32s, and tried those. Nothing either. Tried the Deltas again and they worked fine.

Anyone know what's going on? Does it sound plausible that the PSU wants the fan to have a certain resistance or rpm rating? The Delta fans themselves look like they don't have anything special about them, but maybe they do. Any suggestions? (I don't have easy access to a multimeter).

Thanks,
Daniel.

aokman · 27th October 2004, 10:54 AM

Have you checked the fan output to with a multimeter to see what voltage the PSU is supplying, perhaps its putting out less than what the fan requires to start

You sure all the wires are around the right way?

Its also a good idea to make sure your replacing the Delta's with fans of an equal CFM so the PSU doesn't overheat.

Goth · 27th October 2004, 2:21 PM

Non standard pinout?

larrymoencurly · 27th October 2004, 5:03 PM

What are the pinouts of the fans, including the colors of the wires? Black usually means ground, red goes the the supply voltage (+12V here, even though yellow is normally reserved for +12V, and red for +5V), and white for the RPM signal. Have you tried running the Pabst fans by connecting them directly to +12V (yellow and black), +5V (red and black), or +7V (yellow to yellow, black fan wire to red PSU wire) to see how low the voltage can be for them?

Goth · 27th October 2004, 5:18 PM

tacho monitoring wires are often yellow as well as white

grimwood · 27th October 2004, 5:29 PM

Thanks for the replies.

It looks like I should get a multimeter on the weekend - an el cheapo should be good enough for this.

I definitely got the wires the right way around. I only added the rpm-monitoring wire, and left the other two as well. The globalwins that I tried came with 3 wires, I didn't touch them. It's possible the deltas are wired differently, I'll plug them into the mainboard and see if they work there (I don't think it can get fried through the rpm monitoring as I've heard that's at 12V too). Would be wierd if that's the problem.

The Papst fans don't go as low in voltage by about half a volt (8V vs 7.something). I did crank the machine up so that the fans should be going 12V (when the deltas go full speed), then plugged in a papst and tried to nudge it along by hand and it didn't help. I don't know the difference between starting and operating voltage requirements for these things, but it didn't make a difference. I have a Zalman fanmate in my other machine, I could try putting one of those on a delta and see if it takes much effort to stop one of those working.

As for cooling, thanks for the tip but I think it should be fine. The deltas are rated at about 38CFM (from memory) and the papsts are about 27CFM. I don't have many hard drives like the server that the PSU came out of, and the case has better airflow.

Thanks,
Daniel.

Goth · 27th October 2004, 6:11 PM

How did you add the RPM sense wire? Directly onto the controller IC or Hall Effect sensor?

larrymoencurly · 28th October 2004, 11:07 AM

Is it possible to locate the thermistor that controls the fans and put a capacitor across it to boost the start-up voltage? I found that this allowed some of my fans to start reliably at 1V lower than normal, even when they wouldn't normally start with a push.

grimwood · 28th October 2004, 11:55 AM

Goth was right, it's a non-standard pinout. I took the Delta out and plugged it into the mainboard and it didn't start, it even gave that lovely odour of when you plug things in the wrong way

. I then pulled the pins out of the Papst socket and touched the two main ones to the PSU slots in various ways until I worked out the order. (hopefully that has the rpm sensor in the right place, I could use the Delta in the mainboard to work out the order properly). I did this with the system powered up. Once I worked out the order, I put them back in the socket and plugged it in, and it works nicely (dunno about rpm sensing at this stage). Turned it off and the Papst spins while idle, so I'm guessing it's around 8-9V.

BTW, the wires on the Delta were red, black and blue. I don't remember which was which, will compare the other Papst with it tonight. For the RPM sensing wire, there was a spot on the pcb waiting for it. I haven't tested to see if it actually works, maybe I'll do that before I reorder the pins on the other Papst.

Thanks for the help

.
Daniel.

aokman · 28th October 2004, 12:35 PM

BLUE = Sense Wire
BLACK = GND
RED = +

That what it should have been

grimwood · 29th October 2004, 11:58 AM

Thanks for that aokman, that's what I suspected.

For the record, the connector to the mboard (from the direction I looked at it) has the order sensor, +, ground. The connector to the PSU has the order +, ground, sensor.

I can't tell if the rpm monitoring on the papsts actually works because I can't seem to get any sensor programs to understand the STL2 mainboard. I might try plugging them into my other machine, but for now it's working and I can't be bothered.

So quiet now

.

Thanks,
Daniel.

Lord OverKill · 1st November 2004, 1:23 AM

From my expirence with Delta fans they use like 1.35A or something, more than a normal fan, check the Papst fans on a normal fan header, i would think they work, could be to much current... but i think in this case they would overspin not cease to function... dont know.

I'd think they have enough juice, the third wire (monitor) would possibly make a difference in that the PSU could have a built in switch to not boot unless the fans are functioning.

stroggos2k · 1st November 2004, 10:39 AM

Quote:

Originally Posted by Lord OverKill

From my expirence with Delta fans they use like 1.35A or something, more than a normal fan, check the Papst fans on a normal fan header, i would think they work, could be to much current... but i think in this case they would overspin not cease to function... dont know.

I'd think they have enough juice, the third wire (monitor) would possibly make a difference in that the PSU could have a built in switch to not boot unless the fans are functioning.

I think you'll find that you cant force amps down a fan, it'll only draw what it needs.

aokman · 1st November 2004, 10:49 AM

What is it with the Delta name and people assuming it draws plenty of current

they do make quiet low speed fans aswell you know

grimwood · 1st November 2004, 4:20 PM

Quote:

Originally Posted by Lord OverKill

From my expirence with Delta fans they use like 1.35A or something, more than a normal fan, check the Papst fans on a normal fan header, i would think they work, could be to much current... but i think in this case they would overspin not cease to function... dont know.

I'd think they have enough juice, the third wire (monitor) would possibly make a difference in that the PSU could have a built in switch to not boot unless the fans are functioning.

The Deltas in question use 0.26A according to the delta website. The PSU being very clever is what I first feared, but as the rest of this thread points out, it was just a different pinout on the header.

Regards,
Daniel.

26th October 2004, 4:34 PM	#1
grimwood Member Join Date: Feb 2002 Location: Mt Richon, Perth Posts: 1,546	Delta PSU fan replacement woes Hi All, I have a Delta server PSU, forgot what model number, but it is SSI and has two 350W hotswap units in it. It has two 80mm fans in it and they are loud, at 38dBA. They also run at low rpm while the machine is off. They are 3-pin Delta AFB0812VH fans, rated at 7-13.8V. I want to replace the fans with my quiet 3-pin Papst fans. First thing I did was simply replace the Deltas with the Papsts. The Papsts didn't fire up either when the machine was off or powered up. Put the Deltas back in and they worked. Although they have 3 pin connectors, the Papsts have only two wires. I figured it had something to do with the PSU wanting rpm monitoring, so soldered on an extra wire to the fans to make them 3-wire and tried again, still nothing. Just in case I botched the soldering or the Papst's can't do rpm monitoring, I grabbed some 60mm Globalwin fans from I think Fop-32s, and tried those. Nothing either. Tried the Deltas again and they worked fine. Anyone know what's going on? Does it sound plausible that the PSU wants the fan to have a certain resistance or rpm rating? The Delta fans themselves look like they don't have anything special about them, but maybe they do. Any suggestions? (I don't have easy access to a multimeter). Thanks, Daniel.

27th October 2004, 10:54 AM	#2
aokman Member Join Date: Jul 2001 Location: Melbourne Posts: 8,053	Have you checked the fan output to with a multimeter to see what voltage the PSU is supplying, perhaps its putting out less than what the fan requires to start You sure all the wires are around the right way? Its also a good idea to make sure your replacing the Delta's with fans of an equal CFM so the PSU doesn't overheat. __________________ EpicPhotos.com.au - Image Is Everything Next Expedition - Storm Chasing with ETT from June 3rd, 2013 Last edited by aokman; 27th October 2004 at 10:58 AM.

28th October 2004, 12:35 PM	#10
aokman Member Join Date: Jul 2001 Location: Melbourne Posts: 8,053	BLUE = Sense Wire BLACK = GND RED = + That what it should have been __________________ EpicPhotos.com.au - Image Is Everything Next Expedition - Storm Chasing with ETT from June 3rd, 2013

1st November 2004, 1:23 AM	#12
Lord OverKill Member Join Date: Dec 2003 Location: Bendigo, VIC, Australia Posts: 244	From my expirence with Delta fans they use like 1.35A or something, more than a normal fan, check the Papst fans on a normal fan header, i would think they work, could be to much current... but i think in this case they would overspin not cease to function... dont know. I'd think they have enough juice, the third wire (monitor) would possibly make a difference in that the PSU could have a built in switch to not boot unless the fans are functioning. __________________ lowkey: give me the learnin' stick.

1st November 2004, 10:49 AM	#14
aokman Member Join Date: Jul 2001 Location: Melbourne Posts: 8,053	What is it with the Delta name and people assuming it draws plenty of current they do make quiet low speed fans aswell you know __________________ EpicPhotos.com.au - Image Is Everything Next Expedition - Storm Chasing with ETT from June 3rd, 2013

27th October 2004, 2:21 PM	#3
Goth Grumpy Member Join Date: Jun 2002 Location: /dev/null Posts: 9,303	Non standard pinout?

27th October 2004, 5:03 PM	#4
larrymoencurly Member Join Date: May 2002 Posts: 216	What are the pinouts of the fans, including the colors of the wires? Black usually means ground, red goes the the supply voltage (+12V here, even though yellow is normally reserved for +12V, and red for +5V), and white for the RPM signal. Have you tried running the Pabst fans by connecting them directly to +12V (yellow and black), +5V (red and black), or +7V (yellow to yellow, black fan wire to red PSU wire) to see how low the voltage can be for them?

27th October 2004, 5:18 PM	#5
Goth Grumpy Member Join Date: Jun 2002 Location: /dev/null Posts: 9,303	tacho monitoring wires are often yellow as well as white

27th October 2004, 5:29 PM	#6
grimwood Member Join Date: Feb 2002 Location: Mt Richon, Perth Posts: 1,546	Thanks for the replies. It looks like I should get a multimeter on the weekend - an el cheapo should be good enough for this. I definitely got the wires the right way around. I only added the rpm-monitoring wire, and left the other two as well. The globalwins that I tried came with 3 wires, I didn't touch them. It's possible the deltas are wired differently, I'll plug them into the mainboard and see if they work there (I don't think it can get fried through the rpm monitoring as I've heard that's at 12V too). Would be wierd if that's the problem. The Papst fans don't go as low in voltage by about half a volt (8V vs 7.something). I did crank the machine up so that the fans should be going 12V (when the deltas go full speed), then plugged in a papst and tried to nudge it along by hand and it didn't help. I don't know the difference between starting and operating voltage requirements for these things, but it didn't make a difference. I have a Zalman fanmate in my other machine, I could try putting one of those on a delta and see if it takes much effort to stop one of those working. As for cooling, thanks for the tip but I think it should be fine. The deltas are rated at about 38CFM (from memory) and the papsts are about 27CFM. I don't have many hard drives like the server that the PSU came out of, and the case has better airflow. Thanks, Daniel.

27th October 2004, 6:11 PM	#7
Goth Grumpy Member Join Date: Jun 2002 Location: /dev/null Posts: 9,303	How did you add the RPM sense wire? Directly onto the controller IC or Hall Effect sensor?

28th October 2004, 11:07 AM	#8
larrymoencurly Member Join Date: May 2002 Posts: 216	Is it possible to locate the thermistor that controls the fans and put a capacitor across it to boost the start-up voltage? I found that this allowed some of my fans to start reliably at 1V lower than normal, even when they wouldn't normally start with a push.

28th October 2004, 11:55 AM	#9
grimwood Member Join Date: Feb 2002 Location: Mt Richon, Perth Posts: 1,546	Goth was right, it's a non-standard pinout. I took the Delta out and plugged it into the mainboard and it didn't start, it even gave that lovely odour of when you plug things in the wrong way . I then pulled the pins out of the Papst socket and touched the two main ones to the PSU slots in various ways until I worked out the order. (hopefully that has the rpm sensor in the right place, I could use the Delta in the mainboard to work out the order properly). I did this with the system powered up. Once I worked out the order, I put them back in the socket and plugged it in, and it works nicely (dunno about rpm sensing at this stage). Turned it off and the Papst spins while idle, so I'm guessing it's around 8-9V. BTW, the wires on the Delta were red, black and blue. I don't remember which was which, will compare the other Papst with it tonight. For the RPM sensing wire, there was a spot on the pcb waiting for it. I haven't tested to see if it actually works, maybe I'll do that before I reorder the pins on the other Papst. Thanks for the help . Daniel.

29th October 2004, 11:58 AM	#11
grimwood Member Join Date: Feb 2002 Location: Mt Richon, Perth Posts: 1,546	Thanks for that aokman, that's what I suspected. For the record, the connector to the mboard (from the direction I looked at it) has the order sensor, +, ground. The connector to the PSU has the order +, ground, sensor. I can't tell if the rpm monitoring on the papsts actually works because I can't seem to get any sensor programs to understand the STL2 mainboard. I might try plugging them into my other machine, but for now it's working and I can't be bothered. So quiet now . Thanks, Daniel.