I Replaced my Nexus 92mm with a JMC 92mm PWM Fan, so far I am very happy with it, at 980rpm it's quieter than the nexus @ 12v and I get the same temp.
In the bios I can set a desired speed of 10%, but while the fan is still spinning the rpm is not reported anymore.
Is it safe to use that setting? Could the fan just stop and not start again?
PWM question
Moderators: NeilBlanchard, Ralf Hutter, sthayashi, Lawrence Lee
I think that PWM is safer than dropping the voltages, but still poses a risk of stopping your fan altogether. I don't know how long it takes for a fan to degrade to the point where it won't start though. I wouldn't worry about the rpm not being reported. I've noticed that a lot of boards don't recognize below 600-700 RPMs. I'd keep an eye on the fan for a while if you're going to keep it that low.
-
ToasterIQ2000
- Posts: 36
- Joined: Fri Feb 25, 2005 10:36 pm
The circuitry for reporting a fan's RPM (either in the fan itself or on the fan header side) may require more voltage than the fan motor.
I have LCD modules from Crystalfontz that can control fan speed basaed on auxillary temperature sensors. If I set the minumum fan speed to 10% and leave the RPM reporting turned on, then I can hear the fan spin up twice a second: the fan controller needs to send a higher voltage pulse to the fan in order to read the RPM sensor. At twice a second it is annoying. If I turn the RPM reporting off, the fan spins at a stable 10% speed without pulsing upward. If I set the minimum fan speed to 20% I can leave the RPM sensing turned on without a regular variation in fan noise. (This is with a 5 or 6 year old Pabst 80 mm -- I'd have to pull my radiator to check the model number.)
A question comes to mind: does setting the fan speed in BIOS to 10% mean 1.2 volts ( either by PWM or not ) or does the BIOS consider the fan to be stopped at anything less than say 5 or 7 volts, making 10% fan speed 5.5 volts perhaps? My experience is that most 12 volt fans can be driven at 7-12 volts without sounding odd, but many fans do not like to run below 5 or 7 volts and sound labored or clicky or lumbering or grindy at lower voltages. And that sort of noisiness suggests to me that a fan will be dammaged in the long run.
You might try setting it at 20% or 30% or even 50% and see if there is some minimum threshold for getting RPM reports from your particular fan & fan controller & BIOS program combination.
For myself and PCs I build as a hobby -- PCs that I won't get sued over if they burn: if the fan sounds smooth but isn't giving stable RPM reports I do not worry.
I have LCD modules from Crystalfontz that can control fan speed basaed on auxillary temperature sensors. If I set the minumum fan speed to 10% and leave the RPM reporting turned on, then I can hear the fan spin up twice a second: the fan controller needs to send a higher voltage pulse to the fan in order to read the RPM sensor. At twice a second it is annoying. If I turn the RPM reporting off, the fan spins at a stable 10% speed without pulsing upward. If I set the minimum fan speed to 20% I can leave the RPM sensing turned on without a regular variation in fan noise. (This is with a 5 or 6 year old Pabst 80 mm -- I'd have to pull my radiator to check the model number.)
A question comes to mind: does setting the fan speed in BIOS to 10% mean 1.2 volts ( either by PWM or not ) or does the BIOS consider the fan to be stopped at anything less than say 5 or 7 volts, making 10% fan speed 5.5 volts perhaps? My experience is that most 12 volt fans can be driven at 7-12 volts without sounding odd, but many fans do not like to run below 5 or 7 volts and sound labored or clicky or lumbering or grindy at lower voltages. And that sort of noisiness suggests to me that a fan will be dammaged in the long run.
You might try setting it at 20% or 30% or even 50% and see if there is some minimum threshold for getting RPM reports from your particular fan & fan controller & BIOS program combination.
For myself and PCs I build as a hobby -- PCs that I won't get sued over if they burn: if the fan sounds smooth but isn't giving stable RPM reports I do not worry.
No. This thread is about 4 wire PWM fans, not PWM control of the fan supply voltage. "10%" setting means the PWM control signal duty cycle, the fan will rotate either at it's minimum specified RPM or at 10% of its maximum speed, whichever is larger. The fan voltage remains at 12V.ToasterIQ2000 wrote:A question comes to mind: does setting the fan speed in BIOS to 10% mean 1.2 volts ( either by PWM or not ) or does the BIOS consider the fan to be stopped at anything less than say 5 or 7 volts, making 10% fan speed 5.5 volts perhaps?
-
ToasterIQ2000
- Posts: 36
- Joined: Fri Feb 25, 2005 10:36 pm
I'd not heard of 4-pin PWM fans before. Thankyou for the link.
I have a Asrock 939Dual SATA2, Rev. 1.05. It has a three pin CPU_FAN1 connector (although on one manual page it is schematically 4-pin.)
Reasearching this 4-Pin PWM system I found http://www.analog.com/library/analogdia ... speed.html most usefull.
Figure 6 "Pulse stretching to gather tach information" and accompanying text on that page describes what I was thinking of regarding noise and tach reporting well.
By my reading of Figure 2 "3- and 4-wire fans" there: the move from a 3 to 4 wire design might not alter the coincidence of low power to the fan coils with a loss of tach output signals from the fan. Would a hall sensor for generating a fan tach signal respond differently to 10% duty cycle at 20 hz versus 10% duty cycle at 20 khz?
I have a Asrock 939Dual SATA2, Rev. 1.05. It has a three pin CPU_FAN1 connector (although on one manual page it is schematically 4-pin.)
Reasearching this 4-Pin PWM system I found http://www.analog.com/library/analogdia ... speed.html most usefull.
Figure 6 "Pulse stretching to gather tach information" and accompanying text on that page describes what I was thinking of regarding noise and tach reporting well.
By my reading of Figure 2 "3- and 4-wire fans" there: the move from a 3 to 4 wire design might not alter the coincidence of low power to the fan coils with a loss of tach output signals from the fan. Would a hall sensor for generating a fan tach signal respond differently to 10% duty cycle at 20 hz versus 10% duty cycle at 20 khz?
Not sure about the bios but I have used speedfan with a speed setting of 0% controlling the stock PWM CPU fan from Intel ... and at this speed setting.. speed fan reports around 900 rpmIn the bios I can set a desired speed of 10%, but while the fan is still spinning the rpm is not reported anymore.
Is it safe to use that setting? Could the fan just stop and not start again?
At 100% speed which i presume is full speed of fan.. it reports around 2000 rpm
I am not sure if this is the answer you are looking for but i do wish there were more PWM fans out there with such ease of control from software.
regards
Runn3r