EEVblog Electronics Community Forum

Electronics => Beginners => Topic started by: BinaryBits on October 09, 2021, 11:29:58 am

Title: PWM Fan frequency and motor speed controllers
Post by: BinaryBits on October 09, 2021, 11:29:58 am
So I have ordered a high flow Delta PWM fan which I want to try and build a fume extractor with. I have been looking at these cheap PWM speed controllers in order to regulate the speed of the fan, but most of them says that their frequency is 10 KHz, while the optimal frequency for the fan is I think around 25KHZ.

Would those 10KHz speed controllers work in order to adjust the fan speed between a 1-100% duty cycle, even though the PWM frequencies doesn't match exactly? And would there be any downsides to this? I think I read somewhere that the fan could make more noise because 10KHz is not outside what the human ear can hear?
I have also seen couple of speed controllers that says 100Khz, would that work, and maybe be better than the 10KHz ones?


Thanks in advance.
Title: Re: PWM Fan frequency different from speed controller frequency
Post by: SuzyC on October 09, 2021, 04:29:07 pm
You can control the speed of your fan without even using the PWM feedback output of a fan.

The easiest way to control speed is to control the voltage applied to the fan motor.

You must mind that there is a minimum voltage that the fan will start up and run at and you would have to limit the voltage controlling the fan to something above this voltage to start the span spinning. However, once the fan is started, you can usually decrease the fan motor voltage below the starting voltage for even slower fan speeds. Many CPU chassis-type cooling fans will start to operate at around 6-V and most can withstand voltages up to 14.6 volts without a problem.

You can can control fan speed if you generate a PWM output from a 555 timer or (best) from a MCU and any multi-KHz PWM frequency driving a fan motor will not be audible.

Then, if the PWM output is fed to a transistor or N-Channel MOSFET as a low-side driver(the fan is in series with the negative fan motor terminal and ground), then the PWM will directly control the fan motor voltage=speed, just by varying the PWM.  You may have to use a small capacitor across the motor to filter out spikes cause by the rapidly switching PWM signal to the motor.

Title: Re: PWM Fan frequency different from speed controller frequency
Post by: BinaryBits on October 09, 2021, 08:39:13 pm
Thanks for your answer. Yes I have read about the 555 timer, it seems one should be able to build a PWM controller with very few components, I think it was a 555, a diode, a resistor, a potentiometer, and a few capacitors?
Is the 555 suitable also for 12V fans, or would a 556 be better?

So whats the downside of controlling the fan with voltage, instead of PWM, more than that the fan cant be run with as low RPM?
Title: Re: PWM Fan frequency different from speed controller frequency
Post by: Benta on October 09, 2021, 08:58:09 pm
How about if you show us what fan we're talking about? Even a "røvsvensker" should be able to do that :)

Title: Re: PWM Fan frequency different from speed controller frequency
Post by: charlie.n255 on October 09, 2021, 10:04:46 pm
If you switch a mosfet with the output of the 555 it should be fine to drive the fan.  :-+
Title: Re: PWM Fan frequency different from speed controller frequency
Post by: Benta on October 09, 2021, 10:15:24 pm
If you switch a mosfet with the output of the 555 it should be fine to drive the fan.  :-+

As we still don't know if it's a 2-, 3- or 4-wire DC fan (or some completely different), talking about 555 and MOSFETs makes no sense.

 :palm:
Title: Re: PWM Fan frequency different from speed controller frequency
Post by: charlie.n255 on October 09, 2021, 11:01:02 pm
Sorry, I am a beginner also.
Title: Re: PWM Fan frequency different from speed controller frequency
Post by: Benta on October 09, 2021, 11:21:39 pm
Sorry, I am a beginner also.
Welcome.

We all were at some point. Don't be shy, listen, read and feel the athmosphere first. Then I'm sure you'll be a good addition/contributor here. :)

Title: Re: PWM Fan frequency different from speed controller frequency
Post by: eblc1388 on October 10, 2021, 08:58:22 am
Would those 10KHz speed controllers work in order to adjust the fan speed between a 1-100% duty cycle, even though the PWM frequencies doesn't match exactly?

You may have some misunderstanding between fan's PWM frequency and speed control signal PWM frequency.

The fan's PWM frequency is fixed to the fan by design and user can not change. Its duty cycle will control the fan's speed.

However, the control signal PWM frequency, which user send to fan to control the fan speed is *NOT* used directly for hardware switching. The duty cycle information of this PWM control signal is extracted and used instead. So the control PWM frequency can be 10KHz, 20KHz or 30KHz(within fan's acceptable range) and the fan speed will be the same.

Title: Re: PWM Fan frequency different from speed controller frequency
Post by: BinaryBits on October 10, 2021, 12:07:45 pm
Yes I'm defenitely confused about this PWM thing. So thanks.
Okay so approximately what would be an acceptable range for a fan that is 25KHz?

The fan will be a Delta AFC1212DE, its a 4-pin PWM fan. I have been searching for a data sheet for that exact fan but could not find it. I found data sheets for other Delta fans and they had a lot of detailed and useful information.
The AFC1212 is made in different configurations it seems, but this will be the DE version.
There is another fan with the same model number that is 1.6Amps, but mine is a AFC1212DE that was used in Dell computers and is 3.0Amps, its also branded with "Dell P/N:Y210M-A00"

So by the way, these cheap PWM controllers that are made for motors and such, has only power + and minus, and then + and minus for the motor/fan. So it seems none of them are using the PWM lead that is available on PWM fans.
So is it possible to control the fan with PWM through only the fans + and minus leads, or how would the fan be connected?
Title: Re: PWM Fan frequency different from speed controller frequency
Post by: Ian.M on October 10, 2021, 12:50:16 pm
Odds are its compliant with “4-Wire Pulse Width Modulation (PWM) Controlled Fans”, Intel Corporation September 2005, revision 1.3, as that's what all standard PC motherboards use.
https://www.glkinst.com/cables/cable_pics/4_Wire_PWM_Spec.pdf (https://www.glkinst.com/cables/cable_pics/4_Wire_PWM_Spec.pdf)

For a simple way to control such a fan see my reply #13 in https://www.eevblog.com/forum/beginners/trying-to-build-a-pwm-fan-controller/ (https://www.eevblog.com/forum/beginners/trying-to-build-a-pwm-fan-controller/)  The 555 needs to be a CMOS one as the 'classic' bipolar NE555 doesn't have a rail to rail output.  You can substitute any small signal Schottky diodes for the ones shown. 
Title: Re: PWM Fan frequency different from speed controller frequency
Post by: oPossum on October 10, 2021, 12:54:06 pm
30 Hz to 300 kHz, 0 to 100% PW.

spec sheet attached
Title: Re: PWM Fan frequency different from speed controller frequency
Post by: Siwastaja on October 10, 2021, 01:08:07 pm
This signal is just communicating to the fan the speed you want. With higher frequency, you can communicate the new speed setting more frequently. But because the fan can't change its physical speed quickly anyway, any frequency is fine as long as it's within the limits of the protocol. According to the previous poster above, those limits are wide and why wouldn't they. Unless the fan is defective by design, the end result should be the same: fan adjusting its RPM using whatever internal means that do not matter, to follow the speed communicated using the duty cycle (high-to-low ratio) of the control signal. The frequency of that control signal is irrelevant because there is just one parameter to be communicated and that's done using the duty cycle of the signal.


Now, controlling the speed of a classic simple 2- or 3-wire fan by PWMing one of its power lead is completely different; it's not just communication, it's actual power switching. Those fans do not have internal speed control, but they are designed in such way that they are happy if you just cut their power frequently, allowing PWM control of either positive or negative wire.
Title: Re: PWM Fan frequency different from speed controller frequency
Post by: BinaryBits on October 14, 2021, 06:59:37 am
Thanks, but the data sheet provided in an earlier post is not the correct one, that is the "D" model, mine is the "DE" and is a much higher Amp fan (3A), I have only found the datasheet for the 9 cm version of my fan (AFC0912DE).

This fan is from a Dell computer so the wiring is not a standard fan wiring, so I'm not sure which is the PWM wire and which is the Tach wire, the datasheet for similar Delta fans suggests that the Yellow wire is PWM, while online most people says that the blue Wire is PWM, even on Dell branded Delta fans, and blue is usually the PWM wire on fans.

So my question is, would there be any harm done if I happened to connect the Tach/RPM wire in place of the PWM wire?

About the speed controller connection, maybe I misunderstood the earlier answer, but mine is a 4-pin fan with a PWM Control wire, but the standard PWM motor speed controllers doesn't have a dedicated PWM input, they have only 4 inputs, 2 for power, and then 2 inputs for the motor (+ and -), so could the fan PWM Wire be connected to the speed controllers motor inputs in place of the negative or positive wire? Or how should the fan be wired to the speed controller, or are these not the correct type of PWM speed controller for a PWM fan?
Title: Re: PWM Fan frequency different from speed controller frequency
Post by: Ian.M on October 14, 2021, 07:45:27 am
According to the Intel 4 wire fan spec, both the tach output signal and PWM speed control input signal are supposed to be open drain/collector signals.  i.e. the driving side actively pulls the signal low, but when it goes high, it is pulled up by a passive resistor.   You can wire any number of open drain/collector outputs in parallel without damage, provided the pullup current is less than the current sinking capability of the weakest output and the pullup voltage is similarly below the max. voltage limit of all the outputs.

My circuit (linked above. N.B. use a CMOS '555') has a true open drain output (pulled up to +5V by the R3:R4 potential divider) so is safe if accidentally connected to the tach wire as I just explained.
(https://www.eevblog.com/forum/beginners/trying-to-build-a-pwm-fan-controller/?action=dlattach;attach=587633;image)

However many DIY Intel fan control circuits *DONT* follow the specs and actively drive the PWM pin high.  Those circuits can damage your fan if you mix up the signal wires.

It is *ESSENTIAL* that you properly identify the fan's +12V power and ground wires.  If you get those mixed up there is a high probability you'll damage the fan, probably instantly killing it.  If you've still got the Dell motherboard in question, try checking continuity back to the main power connector to positively identify the power and ground pins, or if you can still power the board up even momentarily, take a DMM and see which pin gets 12V.
Title: Re: PWM Fan frequency different from speed controller frequency
Post by: eblc1388 on October 14, 2021, 08:49:42 am
About the speed controller connection, maybe I misunderstood the earlier answer, but mine is a 4-pin fan with a PWM Control wire, but the standard PWM motor speed controllers doesn't have a dedicated PWM input, they have only 4 inputs, 2 for power, and then 2 inputs for the motor (+ and -), so could the fan PWM Wire be connected to the speed controllers motor inputs in place of the negative or positive wire? Or how should the fan be wired to the speed controller, or are these not the correct type of PWM speed controller for a PWM fan?

What's in your mind a standand PWM controller for fan is an immediate control device installed between +12V and old two wire fan, to achieve some form of speed control similar to a modern 4-wire PWM fan. Very likely the +12V or the 0V is being switched On/Off to emulate some form of PWM current control to varies the fan speed. That device is intended for two wire fans. It may or may not work with a 4-wire fan.

You are correct that those are not the correct type of speed controller for your existing fan. A proper PWM fan controller for a 4-wire PWM fan would generate the correct PWM control signal, but with a steady +12V supply for the control of the fan, and comes with a 4-pin socket to mate with your 4-wire fan, as shown below. Hope this clear up your misunderstanding.
 
Title: Re: PWM Fan frequency different from speed controller frequency
Post by: BinaryBits on October 17, 2021, 08:00:23 am
Thanks for the answers, yes that cleared it up a bit. I have identified the GND and positive lead, I have already tested the fan with DC voltage, but at its rated 12V it obviously was running at full speed, I could get it to run slower when I lowered the voltage, tested down to 8V.
I have not a Dell motherboard, I bought this fan solely to build a 3D printed fume extractor.

Im just not sure about the PWM lead and the Tacho lead, since Dell does not follow standard fan wiring, and internet gives conflicting answers, thats why I asked about the risk if getting those two wires mixed up? I read that the PWM control signal is 5V I think?

I have seen that fan controller shown in the previous post, on Amazon, but I didn't see a potentiometer on it, to adjust the fan speed, but maybe thats done from the button?

I went ahead and bought a cheap PWM Speed controller to try it out, its called "Motor Governor ZK-MG" and supposedly has an adjustable frequency between 1Khz and 99Khz, and adjustable duty cycle between 0-100%. Bought it just to see if it works?

There is this writeup over at the overclockers website that shows how to build a dedicated PWM fan controller with very few components, its for a 5V PWM fan, but it says that the build for 12V fans is almost the same except that one should instead use a 556 and an extra resistor.
If somebody knows if that schematic on that website is correct and should work fine with a 12V, 3Amp fan I would be glad to get a confirmation that it looks okay?
My concern is that my fan can draw up to 3Amps.
Its this website:

https://www.overclockers.com/pwm-fan-controller/ (https://www.overclockers.com/pwm-fan-controller/)
Title: Re: PWM Fan frequency different from speed controller frequency
Post by: eblc1388 on October 17, 2021, 06:09:40 pm
Im just not sure about the PWM lead and the Tacho lead, since Dell does not follow standard fan wiring, and internet gives conflicting answers, thats why I asked about the risk if getting those two wires mixed up? I read that the PWM control signal is 5V I think?

After you have identified the +12V and 0V lead, and with the fan running, you can connect each of the remaining leads to 0V without risk of damaging the fan. The lead which short to 0V would reduce the fan's speed will be the PWM control input. Shorting the tach lead to 0V will not hurt anything.

I went ahead and bought a cheap PWM Speed controller to try it out, its called "Motor Governor ZK-MG" and supposedly has an adjustable frequency between 1Khz and 99Khz, and adjustable duty cycle between 0-100%. Bought it just to see if it works?

I think you have most likely bought a real PWM controller for 2-wire fan. The picture of ZK-MG shows also 150W and 5A, which is a dead give away that it can control 150W with 3A current, for a 2-wire fan. Hopefully you can still use its output as a control signal for you newly bought PWM fan. You may be able to do that by adding an extra PNP or NPN transistor to its output.

There is this writeup over at the overclockers website that shows how to build a dedicated PWM fan controller with very few components, its for a 5V PWM fan, but it says that the build for 12V fans is almost the same except that one should instead use a 556 and an extra resistor.

As I have mentioned in previous post, PWM fan needs a PWM control signal. It can be 5V or 3.3V. With this signal, the size, voltage and power of the fan is irrelevant. With one signal you can control hundred of fans. You should use the circuit provided by Ian.M in the above reply#14 with a CMOS 555 instead.
Title: Re: PWM Fan frequency different from speed controller frequency
Post by: perieanuo on October 18, 2021, 07:36:25 am

My circuit (linked above. N.B. use a CMOS '555') has a true open drain output (pulled up to +5V by the R3:R4 potential divider) so is safe if accidentally connected to the tach wire as I just explained.

hi
 according to this https://www.delta-fan.com/Download/Spec/AFC0912DE-AF00.pdf (https://www.delta-fan.com/Download/Spec/AFC0912DE-AF00.pdf) you need to drive the pwm fan input with some controller like Ian.m suggested:
- control signal frequency 30-300KHz (20KHz preferred)
- controlling the duty factor you control the speed from 0 to 100%

My favorite way is cheap arduino or anything alike, you can implement that duty pwm easy-peasy with 2 euros already manufactured driver.
and when you change your mind, you reassign the arduino to another project https://create.arduino.cc/projecthub/tylerpeppy/25-khz-4-pin-pwm-fan-control-with-arduino-uno-3005a1 (https://create.arduino.cc/projecthub/tylerpeppy/25-khz-4-pin-pwm-fan-control-with-arduino-uno-3005a1)
here https://fdossena.com/?p=ArduinoFanControl/i.md (https://fdossena.com/?p=ArduinoFanControl/i.md) the gut controlled 3 fans to stabilise case temp with one arduino
cheapest arduino will work, buy whatever you'll find useful for future projects
Title: Re: PWM Fan frequency and motor speed controllers
Post by: BinaryBits on October 20, 2021, 04:27:56 pm
Thanks for the info, but the datasheet for the Delta AFC0912DE fan says "Caution: The lead wire of RD signal can not touch the lead wire of positive or negative"
After reading a bit of Delta documents it seems to mean "Rotational Detector, is that something different from the Tacho wire??

I checked the voltage on the two wires and the blue showed a steady 3.62V, and the yellow showed close to 0V, sometimes jumping around a bit? I suspect that the blue wire is PWM?

I tried the speed controller that I bought, with the black and red wire (DC control) at 25Khz, and it adjusts the speed of the fan, but not smoothly/linearly, firstly It can only go down to about 40% Duty cycle before the fan stalls/makes strange noises, and secondly after around 75% it feels as if the speed is not increasing anymore. So its not the ultimate way of controlling the fan.

That writeup I mentioned is also a dedicated PWM fan controller, the reason I liked that one was because it was very clear and helpful for a beginner like me, with a parts list and everything, and seemed extremely simple. I would gladly try to build the provided circuit by Ian.M, but unfortunately that circuit diagram alone is not enough for me to be able to build that particular PWM controller. Thats why I asked about the one I pointed too.

About the Arduino, I actually tried that earlier successfully with another PWM fan, which is only a 0.25A fan (but which draws only 0.09A at 12V), but I was afraid that the Arduino couldn't handle the 3A of the Delta fan, or maybe the Arduino part of the setup is not affected by the amp draw? Also I would need to be able to adjust the duty cycle externally through a potentiometer and not in software.
Title: Re: PWM Fan frequency and motor speed controllers
Post by: perieanuo on October 20, 2021, 04:39:25 pm
read this https://aerocooler.com/delta-afc1212de-dell-p-n-y4574-120x38mm-extreme-hi-speed-fan-dell-5pin/
Quote
It comes with 4 wire lead (white=TAC sensor, red=power, black=gnd, blue=PWM control) to 5 pin Dell connector
so blue is pwm control. here you apply the pwm. have you buyed yet something to drive the pwm, or do you already got some arduino/stm board?
if you do, get to work, write the controller, put the wires and it will do
Title: Re: PWM Fan frequency and motor speed controllers
Post by: perieanuo on October 20, 2021, 04:56:31 pm
Thanks for the info, but the datasheet for the Delta AFC0912DE fan says "Caution: The lead wire of RD signal can not touch the lead wire of positive or negative"
After reading a bit of Delta documents it seems to mean "Rotational Detector, is that something different from the Tacho wire??

I checked the voltage on the two wires and the blue showed a steady 3.62V, and the yellow showed close to 0V, sometimes jumping around a bit? I suspect that the blue wire is PWM?

I tried the speed controller that I bought, with the black and red wire (DC control) at 25Khz, and it adjusts the speed of the fan, but not smoothly/linearly, firstly It can only go down to about 40% Duty cycle before the fan stalls/makes strange noises, and secondly after around 75% it feels as if the speed is not increasing anymore. So its not the ultimate way of controlling the fan.

That writeup I mentioned is also a dedicated PWM fan controller, the reason I liked that one was because it was very clear and helpful for a beginner like me, with a parts list and everything, and seemed extremely simple. I would gladly try to build the provided circuit by Ian.M, but unfortunately that circuit diagram alone is not enough for me to be able to build that particular PWM controller. Thats why I asked about the one I pointed too.

About the Arduino, I actually tried that earlier successfully with another PWM fan, which is only a 0.25A fan (but which draws only 0.09A at 12V), but I was afraid that the Arduino couldn't handle the 3A of the Delta fan, or maybe the Arduino part of the setup is not affected by the amp draw? Also I would need to be able to adjust the duty cycle externally through a potentiometer and not in software.
yellow is the output showing you some pulses mirroring motor rotation, if you put a pull-up resistor you'll see those pulses on scope (see page 7)

your speed controller adjust Vcc for your motor, it's not normal to control speed like this. NO NO NO. this motor is with FIXED voltage supply and his speed is imposed by driving PWM input (blue) like they say on page 8, not by changing Vcc on red pin.
leave the yellow alone for now (you won't monitor his speed, for now just set the speed you want with pwm control on blue wire), after succeding this control you can move to monitor if this speed is correctly done by the motor.

references are to the pdf to your actual motor, which are almost identical with the one i posted before.

maybe this will clear your ideas, exception is in your case your  yellow is the white wire from my example https://www.cgdirector.com/pwm-dc-system-fans/ (https://www.cgdirector.com/pwm-dc-system-fans/) or really basic and clear  https://www.ekwb.com/blog/what-is-pwm-and-how-does-it-work/ (https://www.ekwb.com/blog/what-is-pwm-and-how-does-it-work/)
Title: Re: PWM Fan frequency and motor speed controllers
Post by: BinaryBits on October 22, 2021, 10:04:20 am
Yeah, the speed controller that I have now uses the fan in DC mode, i.e, only using the Red and Black wires of the fan, but it still has a variable frequency setting, and adjustable duty cycle, so if the speed controller only changes the voltage then what is the variable frequency and Duty cycle good for? Then it would be more like just a simple adjustable buck converter?
I understand its not a real PWM fan controller, this one I have now is mainly for control of motors.

I have not a real PWM fan controller yet, I want to try and build one with a 555/556 timer, I just must order the parts first. I already have an Arduino and have successfully driven another lower Amp PWM fan through the Arduino. But I would rather build a dedicated PWM controller, it feels cleaner and smaller footprint than an Arduino?

Okay so I need to have a resistor with the Tacho wire to get a "normal" output? I tried probing the Tacho wire with my cheap oscilloscope without an resistor and the trace didn't make much sense to me.
What value resistor would I need, or how should that be calculated?
Title: Re: PWM Fan frequency and motor speed controllers
Post by: perieanuo on October 22, 2021, 02:46:59 pm
sorry, but you don't get the idea. your fan is designed to operate not from 0 to 12Vdc, but from 8.0 to 13.2 Vdc, and to be controlled in speed by pwm pin.
i can't validate something that is bad engineering just because you insist.
finally, choose to regulate him as you please, i'll stop here, you already have all the info on the table.
from my part, you can regulate him with energy from santa-klaus, i explained how it's done correctly and how everyone else does it.
Title: Re: PWM Fan frequency and motor speed controllers
Post by: BinaryBits on October 24, 2021, 12:57:43 pm
Maybe you didn't understood what I was writing, but nowhere have I insisted on running it in DC mode, thats just what I do right now because I don't have a real PWM fan controller, I have mentioned that I intend to build one though.

What I did though was questioning what the difference is between a motor speed controller and a buck converter?

And just like I have written in my previous comment I understand fully that a PWM fan is supposed to be run through the PWM control wire by adjusting the duty cycle, not by changing the voltage, but its hard to do that right now since I don't have the hardware to do that.

Thanks anyway.
Title: Re: PWM Fan frequency and motor speed controllers
Post by: eplpwr on October 24, 2021, 01:38:50 pm
The good thing with PWM control is that the fan itself takes care of the regulation. E.g. if you want to run at 25% speed, a DC control (switchmode or linear) would probably have to give the fan a short "boost" of a slighly higher voltage to get the fan spinning, then settle for the voltage that corresponds to 25% of max speed; with PWM control the fan itself takes care of this startup scenario.

AFAIK, 25 kHz PWM frequency was chosen as to be outside the audible band to make switching noise unaudible (to humans).

At 3 amps, there will be some power losses, quite high if linear regulating, less so for switchmode. This regulation is moved to the fan itself with PWM control - and the fan is a very good cooler.