PWM VS Variable Voltage | For controling DC Fan Speed

[xReM1x]

Hi,

I was wondering, is there any advantages to PWM a fan to control it speed instead of changing the input voltage?
I know that 1 advantage is that you dont waste alot of power with PWM, But for 12V 160mA fan, its not a problem.

My fan is arctic cooling F8 PWM REV2. it is not used inside a pc, its for a project.

So what are the advantages and disadvantages of PWM a fan and what are the advantages and disadvantages of changing directly the input voltage?

 

[Ryan_78]

Pwm is when it auto changes.

 

[xReM1x]

Pwm is when it auto changes.

What?

 

[TJ Hooker]

I don't think there's much difference. As you said, the biggest benefit of PWM is power efficiency. The only advantage of voltage control I can think of is that it may be easier to implement.

 

[xReM1x]

I don't think there's much difference. As you said, the biggest benefit of PWM is power efficiency. The only advantage of voltage control I can think of is that it may be easier to implement.

I've heard that the fan can be damaged if the input voltage changes.
Is it true?

 

[Paperdoc]

Compared to older Voltage Control Mode fans, the newer PWM fans have two small advantages. One is that they can be run to lower minimum speeds, and the second is that they can start up more easily and hence don't require a full power start-up signal. Both of these are minor unless you normally run the fan very slowly.

As I understand it, one part of the "backwards compatibility" for 4-pin (PWM) fans is that if they are fed from an older-style 3-pin fan port (Voltage Control Mode) they still WILL work under the mobo's speed control system. Apparently this is because, give NO PWM signal on the 4th pin of the port (which does not exist on a 3-pin port!) the fan will simply operate from the DC voltage supplied on Pin #2, whatever that is, and its speed will be determined by that, just like a 3-pin fan. This means you do NOT get the advantages of low-speed reliability from the PWM feature that is not operating, but that's not a big deal.

If you are not using the fan inside a PC and powering it from a mobo 4-pin port operating in true PWM Mode, it would be quite difficult for you to create the "normal" signals from a mobo to operate the fan in PWM Mode. You would be much better to arrange a fan power supply with Pin #1 as Ground, and Pin #2 as a DC voltage which varies. You do not need to connect anything to Pins 3 or 4. The Voltage to Pin #2 should max out at 12 VDC, and normally it should not go lower than about 5 VDC, because at lower voltages it may stall. You should also be aware that starting a 12 VDC fan with a low voltage (lower than 6 VDC or so) may not work. A standard 3-pin fan port on a mobo always STARTS up at full 12 VDC for a couple of seconds to ensure the motor starts turning, then reduces voltage to whatever is needed for the fan to do its job.

There are lots of PWM speed controllers for DC motors on the market. Do NOT attempt to use one of those with a computer PWM (4-pin) fan. The computer fan is built differently from a standard DC motor, and several makers of computer PWM fans warn you not to power them from common PWM speed controllers. Doing so may not achieve what you expect, and very likely will damage some of the internal electronic components of a computer PWM fan.

As a side piece of info in case it affects your project, the automatic fan control function in a PC is NOT a speed control, even though that's the term everyone uses. It is a TEMPERATURE control system. For the case fans, for example, there is a temperature sensor built into the mobo by its maker, and the BIOS has a target for that measured temperature. The control system compares the actual measured temperature to the target and manipulates the voltage supplied to the fan to whatever it takes to get that temperature to the target. Of course this really does change the fan speed so that it changes the air flow rate generated by the fan. The point is, the control system does not aim to control fan speed; it aims to control a measured temperature, and fan speed and air flow (via fan supply voltage or PWM signal) is just the variable the control loop can manipulate to achieve that aim.

Regarding speed measurement, the fan generates inside itself a train of pulses for this. The signal is 2 pulses per fan revolution, and I believe the pulses are 5VDC magnitude. It is output from the fan on its connector Pin #3. The mobo has a counting circuit for that so it can display to speed for you. Many mobos also monitor this pulse train and, if it shows NO pulses, that is interpreted as fan failure to generate an alarm message. But interestingly, that measured speed is NOT used for "speed control" because, as I said, the control system does not care about speed - it is focused on temperature.

 

[TJ Hooker]

I've heard that the fan can be damaged if the input voltage changes.
Is it true?

I can't really think of a reason why varying voltage would damage the fan. I suppose large, rapid voltage swings causing the fan speed to drastically change could maybe cause some mechanical stress on the fan. Doesn't seem like something that's likely to be an issue though.

@Paperdoc I think you're splitting hairs. The mobo is controlling the speed of the fan... it's speed control. Just because the tachometer output isn't fed back into the control loop doesn't mean it's not speed control. The input to the control system is temperature, and the output is fan speed.

 

[Paperdoc]

The warning about voltage supply to a 4-pin fan is based on two issues as I understand it. One is that the fan also contains a printed circuit board with components to do several jobs like simulating a commutator and brush system for this brushless motor, generating a speed pulse output, and modulating the 12VDC supplied on Pin #2 with the PWM signal supplied on Pin #4 before sending that to the motor windings. Those circuits require a stable 12VDC supply. Now, earlier 3-pin fans have many similar circuits, too, but they were designed to be used with a variable DC supply on Pin #2, so I have to assume that these components still can work with a supply of less than 12 VDC. How much less? I don't know for sure, but I'd guess they can work down to maybe 5 VDC, and a similar rule would apply to a newer 4-pin fan. A really low supply voltage probably would not damage the fan, but certainly would result is failure to perform as expected.

The other issue is HOW the "reduced voltage" is supplied. Simply providing a fixed DC voltage as a 3-pin fan port would do is one scenario that appears acceptable. BUT the fan manufacturers' warnings I have seen particularly caution you NOT to supply a "reduced voltage" which really is a DC supply already modulated by a PWM signal so that it is NOT a true DC supply. Its is, instead, a train of voltage pulses that switches rapidly from 0 to 12 VDC and back, often at a frequency of 10 to 30 kHz. Such a signal can be used for "normal" DC motors. It cannot be used with the special version motors used for computer "PWM Fans" because the voltage supply contains a lot of noise and potential voltage spikes that could damage the electronic components of the circuits included inside the motor case.

 

[xReM1x]

Compared to older Voltage Control Mode fans, the newer PWM fans have two small advantages. One is that they can be run to lower minimum speeds, and the second is that they can start up more easily and hence don't require a full power start-up signal. Both of these are minor unless you normally run the fan very slowly.

As I understand it, one part of the "backwards compatibility" for 4-pin (PWM) fans is that if they are fed from an older-style 3-pin fan port (Voltage Control Mode) they still WILL work under the mobo's speed control system. Apparently this is because, give NO PWM signal on the 4th pin of the port (which does not exist on a 3-pin port!) the fan will simply operate from the DC voltage supplied on Pin #2, whatever that is, and its speed will be determined by that, just like a 3-pin fan. This means you do NOT get the advantages of low-speed reliability from the PWM feature that is not operating, but that's not a big deal.

If you are not using the fan inside a PC and powering it from a mobo 4-pin port operating in true PWM Mode, it would be quite difficult for you to create the "normal" signals from a mobo to operate the fan in PWM Mode. You would be much better to arrange a fan power supply with Pin #1 as Ground, and Pin #2 as a DC voltage which varies. You do not need to connect anything to Pins 3 or 4. The Voltage to Pin #2 should max out at 12 VDC, and normally it should not go lower than about 5 VDC, because at lower voltages it may stall. You should also be aware that starting a 12 VDC fan with a low voltage (lower than 6 VDC or so) may not work. A standard 3-pin fan port on a mobo always STARTS up at full 12 VDC for a couple of seconds to ensure the motor starts turning, then reduces voltage to whatever is needed for the fan to do its job.

There are lots of PWM speed controllers for DC motors on the market. Do NOT attempt to use one of those with a computer PWM (4-pin) fan. The computer fan is built differently from a standard DC motor, and several makers of computer PWM fans warn you not to power them from common PWM speed controllers. Doing so may not achieve what you expect, and very likely will damage some of the internal electronic components of a computer PWM fan.

As a side piece of info in case it affects your project, the automatic fan control function in a PC is NOT a speed control, even though that's the term everyone uses. It is a TEMPERATURE control system. For the case fans, for example, there is a temperature sensor built into the mobo by its maker, and the BIOS has a target for that measured temperature. The control system compares the actual measured temperature to the target and manipulates the voltage supplied to the fan to whatever it takes to get that temperature to the target. Of course this really does change the fan speed so that it changes the air flow rate generated by the fan. The point is, the control system does not aim to control fan speed; it aims to control a measured temperature, and fan speed and air flow (via fan supply voltage or PWM signal) is just the variable the control loop can manipulate to achieve that aim.

Regarding speed measurement, the fan generates inside itself a train of pulses for this. The signal is 2 pulses per fan revolution, and I believe the pulses are 5VDC magnitude. It is output from the fan on its connector Pin #3. The mobo has a counting circuit for that so it can display to speed for you. Many mobos also monitor this pulse train and, if it shows NO pulses, that is interpreted as fan failure to generate an alarm message. But interestingly, that measured speed is NOT used for "speed control" because, as I said, the control system does not care about speed - it is focused on temperature.

Thanks, I got my answer - will the fan be damaged from varying the input voltage? no.

 

[latedev]

No because that is exactly what a PWM fan in effect does. Voltage on, Voltage off, which generates more Voltage spikes than a straight Voltage controlled fan does.

Low speeds are not difficult but do require more initial current to start the fan, better to start it at full tilt then reduce the speed. I cannot think of any computer that has pure Voltage control, as it is more expensive to produce, unless you are using a retro fitted device.