SuperSilencer back from the brink, but new fan is slow??
Moderators: NeilBlanchard, Ralf Hutter, sthayashi, Lawrence Lee, Devonavar
SuperSilencer back from the brink, but new fan is slow??
And the ongoing saga of my supersilencer modification continues... If you recall the first episode, entitled "I killed my Supersilencer..." I'd learned that touching the heatsink to ground while the PSU was in operation was a not a particularly bright idea... After soldering a new fuse to the board, I resuscitated the PSU.
I've now proceeded to finish the original modification by replacing the stock SuperRed fan (34CFM) with a Panaflo M1BX (rated 32CFM) and attached the motherboard monitoring wire so I can see what's going on... As you might remember from the first episode, I attached the M1BX fan and found that it was not starting. I've now come to the conclusion that this fan can not and will not start without a little push -- everytime I turn it on. So the interim solution? Never turn it off. Well, it's silent enough that I can live with that for the time being.
Problem is, as I'd suspected, the M1BX fan is running way low. Abit fan monitor's (Abit uGuru EQ & BIOS) indicate that the fan sits at 1000rpm and isn't budging (that is, once I kick start it with the flick of a car key). I'm running two instaces of Prime95 on my P4 3.0C hyperthreading CPU, and have it maxed out at 100%, CPu temp at 65C, Systemp at 42C, PWM temp at 71C and the PSU fan still sits at around 1080rpm. When I touch the PSU housing, it feels very hot, as if it's not being adequately cooled.
So to sum up the fan situation, here's what my choices appear to be so far:
Stock SuperRed:
Starts reliably at 1500rpm in my machine, and goes upward of 3000rpm at full load (annoying whine starts around 1800rpm -- the reason I wanted to mod it in the first place).
Panaflo M1BX:
Doesn't start without a kickstart (ever!), and runs at 1000rpm at 0% and about 1080rpm at full load. Silent as the night sky, but PSU seems to be getting very hot under load.
I guess this Panaflo is supposed to have similar CFM to the stock one, but at what voltage???? I can accept the idea that this particular fan might be a "reluctant starter", but once it's going, shouldn't it at least ramp up rpms a little bit? I mean the PSU must be pushing close to 12V at it when the rest of the system is at full load for half an hour right?
Not sure what to consider now. Trying another M1BX might resolve the no starting issue, but what about the the fact that it's just not ramping? Is this inherent to all M1BX fans? I'd rather not have to buy a wholoe bunch of $15-$20 fans to troubleshoot this issue if possible... Should I perhaps consider a Papst NGML, an L1A (both lower CFM), or another M1BX, or give up and put the stock SuperRed back in?
Gordo
I've now proceeded to finish the original modification by replacing the stock SuperRed fan (34CFM) with a Panaflo M1BX (rated 32CFM) and attached the motherboard monitoring wire so I can see what's going on... As you might remember from the first episode, I attached the M1BX fan and found that it was not starting. I've now come to the conclusion that this fan can not and will not start without a little push -- everytime I turn it on. So the interim solution? Never turn it off. Well, it's silent enough that I can live with that for the time being.
Problem is, as I'd suspected, the M1BX fan is running way low. Abit fan monitor's (Abit uGuru EQ & BIOS) indicate that the fan sits at 1000rpm and isn't budging (that is, once I kick start it with the flick of a car key). I'm running two instaces of Prime95 on my P4 3.0C hyperthreading CPU, and have it maxed out at 100%, CPu temp at 65C, Systemp at 42C, PWM temp at 71C and the PSU fan still sits at around 1080rpm. When I touch the PSU housing, it feels very hot, as if it's not being adequately cooled.
So to sum up the fan situation, here's what my choices appear to be so far:
Stock SuperRed:
Starts reliably at 1500rpm in my machine, and goes upward of 3000rpm at full load (annoying whine starts around 1800rpm -- the reason I wanted to mod it in the first place).
Panaflo M1BX:
Doesn't start without a kickstart (ever!), and runs at 1000rpm at 0% and about 1080rpm at full load. Silent as the night sky, but PSU seems to be getting very hot under load.
I guess this Panaflo is supposed to have similar CFM to the stock one, but at what voltage???? I can accept the idea that this particular fan might be a "reluctant starter", but once it's going, shouldn't it at least ramp up rpms a little bit? I mean the PSU must be pushing close to 12V at it when the rest of the system is at full load for half an hour right?
Not sure what to consider now. Trying another M1BX might resolve the no starting issue, but what about the the fact that it's just not ramping? Is this inherent to all M1BX fans? I'd rather not have to buy a wholoe bunch of $15-$20 fans to troubleshoot this issue if possible... Should I perhaps consider a Papst NGML, an L1A (both lower CFM), or another M1BX, or give up and put the stock SuperRed back in?
Gordo
There's a remote possibility of the fan being connected incorrectly. I once put a Papst 4412F/2GL in my 300W Tornado but plugged it in backwards, so it was basically being powered from the RPM sensing line. The fan would never start on its own, and once manually started it always spun slowly (as in I could see the hub turning unblurred) and the PSU got pretty hot.
Have you tested the fan on its own, i.e. outside the PSU running off something like a Fanmate?
Have you tested the fan on its own, i.e. outside the PSU running off something like a Fanmate?
Hmmm, though it's possible I wired it wrong, I don't see how the fan monitoring wire would be incorrectly connected if it's plugged into the MB and providing an rpm indication. I'll try taking it apart though and connecting the fan to a fan mate. If that worked, I suppose that would rule out the fan as the cause of the problem.nap_loaf wrote:There's a remote possibility of the fan being connected incorrectly. I once put a Papst 4412F/2GL in my 300W Tornado but plugged it in backwards, so it was basically being powered from the RPM sensing line. The fan would never start on its own, and once manually started it always spun slowly (as in I could see the hub turning unblurred) and the PSU got pretty hot.
Have you tested the fan on its own, i.e. outside the PSU running off something like a Fanmate?
Systematic approach:
- Test the fan on its own and make sure it starts at low voltages
- Test the PSU with the stock fan again, to make sure it's truly fixed (in case something in addition to the fuse died)
If both are fine, it's definitely a wiring issue. Assuming the Tornado and Silencer use the same circuit board, the fan should be plugged in so the smooth side of the fan plug faces the back (exhaust side) of the PSU.
Actually I should check something: does your PSU fan header have 3 pins or 2? If it's only 2 (like some OEM Tornados have) then a lot of what I've said doesn't apply to you.
- Test the fan on its own and make sure it starts at low voltages
- Test the PSU with the stock fan again, to make sure it's truly fixed (in case something in addition to the fuse died)
If both are fine, it's definitely a wiring issue. Assuming the Tornado and Silencer use the same circuit board, the fan should be plugged in so the smooth side of the fan plug faces the back (exhaust side) of the PSU.
Actually I should check something: does your PSU fan header have 3 pins or 2? If it's only 2 (like some OEM Tornados have) then a lot of what I've said doesn't apply to you.
-
- SPCR Reviewer
- Posts: 8636
- Joined: Sat Nov 23, 2002 6:33 am
- Location: Sunny SoCal
The 2 SS400's that I've checked both put out about 4.8 to 4.9V to the fan header at their slowest setting. Assuming that yours is the same, that should be plenty of voltage to run you fan and it should be turning somewhere in the neighborhood of 1200-1300 at that voltage. You might want to check your voltage at the fan header to see what it's putting out. Maybe it's somewhat less than 5V.
Well, I never got around to figuring out what was up with the PSU fan signal, but I checked the M1BX fan itself by plugging it into a FanMate. It works beautifully, starting up every time at the lowest setting and pushing out decent airflow at a full 12V. So clearly the problem is the PSU or some strange combination of incorrect wiring and the M1BX's reluctance to start at less than 5V. Who knows, I may have fried the PSU fan controller circuitry during one of the many blunders I made during this mod...
So in the end, I decided to route the M1BX through the FanMate and out to the motherboard permanently. I've set it slightly higher than the lowest FanMate setting, so it's not quite silent, but definitely not audible @ about 6 ft from the box. That's definitely better than the yo-yo like whining the old fan put out.
To test the setup, I ran the system at full throttle for 30 minutes by maxing out my cable connection, writing to both drives, running two instances of Prime95 and allowing a Flight Simulator Cessna to fly straight and level, (taxing both CPU and video). Ok, certainly not a scientific torture test, but it's likely more combined load than I'll ever put the system through. After 30 minutes, the drives were each at 33C (elastic suspended with a 7V fan blowing on them) the CPU temp sat at 60C, the System hit 33C and PWM at 65C. The PSU was noticeably warm, but not noticeably moreso than when the SuperRed was in it at full throttle.
If I really need to push the system, I've placed the FanMate in an accessible location so I can just reach into the 5-1/4" drive bay and turn it up. I don't know that this is the overall safest situation for the PSU, but it's basically all I can think of doing right now (short of spending more cash) to satisfy my requirement for a quiet PSU.
Gordo
So in the end, I decided to route the M1BX through the FanMate and out to the motherboard permanently. I've set it slightly higher than the lowest FanMate setting, so it's not quite silent, but definitely not audible @ about 6 ft from the box. That's definitely better than the yo-yo like whining the old fan put out.
To test the setup, I ran the system at full throttle for 30 minutes by maxing out my cable connection, writing to both drives, running two instances of Prime95 and allowing a Flight Simulator Cessna to fly straight and level, (taxing both CPU and video). Ok, certainly not a scientific torture test, but it's likely more combined load than I'll ever put the system through. After 30 minutes, the drives were each at 33C (elastic suspended with a 7V fan blowing on them) the CPU temp sat at 60C, the System hit 33C and PWM at 65C. The PSU was noticeably warm, but not noticeably moreso than when the SuperRed was in it at full throttle.
If I really need to push the system, I've placed the FanMate in an accessible location so I can just reach into the 5-1/4" drive bay and turn it up. I don't know that this is the overall safest situation for the PSU, but it's basically all I can think of doing right now (short of spending more cash) to satisfy my requirement for a quiet PSU.
Gordo