I did like this video and it gave me some ideas for my situation. Question for you though. I just built a pc with the Cooler Master H500m on the ASUS Crosshair VIII Hero (Wifi) mobo. The case has 2 200mm DC fans that turn on and off on silent mode like you were talking about. I was thinking of replacing them but now I'm wondering if there's a controller that can treat them like pwm and also make them visible in Fan Xpert 4. Any thoughts on that? Also, why does it seem like 200mm fans don't really come PWM except for Thermaltake but those have a proprietary connector that isn't just a regular PWM fan connector? Confusing stuff... Anyways, appreciate any help you could provide. Thanks!
Some server fans are 4 pin with some of the pins wired a little differently. I've completely forgotten the "standard" they use. I think it was just an annoying OEM rewiring. I was (about 10 or 15 years ago) able to re-pin the server fan to work with out issue on a standard pc motherboard. It was a nice fast dangerous delta fan from memory :D
@@tamaica7770 Stepping up and down voltages is rather easy to do. Stepping up and down current involves more work and other various reasons. More current = more heat for example.
skillet pan heat depends on P=IV. Lets say the fan can work at 6V, and mobo gives it current to operate. Increase current, increase the speed of fan. just not sure how they decided to choose V instead of I. maybe for safety?
4 года назад+176
Finally - someone explaining it in a way that can be understood
I love how thorough you are about this subject and that everything mentioned is relevant and on track. The only thing I don't like is there is no discussion about the 4 pin molex connector fan type.
Damm! This man is unbelievable! I just wish I had known this channel before. Of the dozens of computer channels I know, this one is by far the best. What a fantastic teaching method! he doesn't need edits, animations or a ton of fancy equipment to perfectly explain a subject. That's talent! Thank you!
Voltage control uses resistance to lower the tension, therefore as a result it consumes more power due to the loss of energy as a form of heat; this makes PWM a power saver. I really liked your channel right away after watching a couple of videos because obviously it is not made for showing off but for teaching purposes, “not too many infos + accuracy at its best = subscribe”. Thank you.
- 3 Pin regulates through the amount of "juice" it sends to the fan. The more power the faster the fan spins. - 4 Pin regulates through timings, for how long the fan is on on end. Pretty simple. Can sum it up in two paragraphs. Great video, well done.
Realy nice video. I've enjoyed it a lot. Few time ago i've to buy a new processor fan because mine have burned after 6 years of use. I almost bought the 3 pin fan but after some internet research i discovered that processors use pwm (4 pin) fans. Then i bought the right fan for my processor and i am still waiting for it to come. Thanks for nice content. Regards. Rodrigo.
Nice visual of "pulse width modulation".............. hope folks appreciate the lesson!....................... Nicely done!........................... AJ
Brilliant explanation, this is so useful in understanding how car radiator fans work and when your ac doesnt work with the fan going into full speed mode!
Video is fantastic, these are found on central heating appliances too, which now have modulating fans and pumps which work via a PWM signal from PCB. Normally heating engineers don't carry oscilloscopes, even cheap ones, so we are reduced to using Hz on a multi-meter.
THANK YOU! I bought 2 new fans for an older pc and I did not realize until today that I have 3 pin fans while the sockets on the motherboard are 4 pins. So I was sad that I must send back and buy new ones. You are a lifesaver and your explanations are amazing and easy to understand.
Another beauty of PWM control: I'm a model train enthusiast. My DC trains can run on straight DC or PWM, but using PWM allows the lighting to remain at full brightness, and the sound function at full volume, at any speed. The full voltage is always available for a fraction of the time and is used to charge capacitors to power the accessories. Wall warts / power adapters can be made without heavy, expensive copper windings if they use PWM.
Wow! This was a fantastic demonstration of how PC fans work. I've been struggling to understand the difference between 3 and 4 pin fans and now I have a clear understanding of the two.
I didn't know whether the fan was driven by the power of PWM signal or not, but now I see the maximum voltage of the PWM line only goes up to about half a volt, so my question is answered. This video is exactly what I was looking for. The power if provided by the 12-volt line and the PWM acts as a sort of relay that switches the 12 volts on and off. Thanks for the video.
Duty cycle is still correct. even voltage controlled fans are controlled over a PWM dutycycle on the 12V rail. The fact that a motor is an inductor flattens the 12v-0v difference and averages it out to a voltage between the two values. with an oscilloscope you'd be able to see, that the line in fact has 12V peak and just switches on and of rapidly.
There is probably a capacitor also on the power rail which will average out the voltage so you will probably not see the full 12-to-zero volt swing on the line using a scope. But I agree that the voltage is most probably varied using PWM even for a 3-pin fan.
A few years ago I designed an AtTiny85 based DIY fan controller, and checking the tach in DC mode was a royal pain, because in the off/low part of the duty cycle, there was no power for the tach signal, so I had to run a burst of 100% DC (pulse stretching) every time I wanted to check the time between two tach pulses, which led to audible noise, especially at low RPMs. With a capacitor between the 12V and GND maybe the voltage would be flat enough to get a readable tach signal even in the off part. OTOH, the capacitor would cause the circuit to draw more current in the on part as the capacitor was refilled, and for the same reason it would also make the fan run faster at lower duty cycles because the average current would be higher, making speed control more difficult and unpredictable at low duty cycles. Unless I somehow found a way to feed the fan a PWM signal not between 0V and 12V, but between 5V and 12V, so that the tach (hopefully) always stayed readable with 5V logic. But then again, that would prevent the fan from spinning down below around 40%. (Hey, did I just stumble upon the reason for that particular limit in the BIOS?) Nah. My conclusion remains, 3 pin fans suck for speed control.
I just got back into building PCs again since 2002. This was a great video for clearing some confusion I have with my case fans in my new PC. Thank you.
Well done on the growth of your channel and thanks for a brilliant explanation, I just got really nice be quiet case with 3x3pin fans and 2x 4 pin fans, I was a tad disappointed but eh still very cool, now I know that I can use them together reasonably, maybe not as 'cool' as 5x4pins but I think I can live with it for now, AUTO sounds really useful! ;D
My question is, with a PWM fan, is the actual motor coil attached to the fans internal controller still two wires? If so couldn't they just keep the 3 wire method and change the motherboards control to do the PWM directly? I'm guessing that the fan motor itself is more complicated, possibly with multiple windings or something similar and uses a dedicated motor driver to provide more granular control?
Main issue is that if you drive the fan by its 12v wire, your fan controller has to switch the load of the fan. But with a dedicated PWM wire, the fan controller just sends a digital signal, and the switching is done by the fan itself - so a big fan that pulls a lot of power won't overload the controller. Additionally, yes a brushless motor usually has four coils in it that need to be sequenced, so there will be a driver for that in the fan, as well as the speed sensor. These two chips may behave unpredictably if their 12v power supply is getting switched on and off at 23KHz. There were some very early PWM fan controllers, mostly custom jobs, that were designed to drive the 12v pin, and while I never used one they were known for causing an electronic humming noise from the fan.
You explained it really well. Now I am in a problem with colour codes of wire. My old 3 wire fan had brown, red, and black colour wires, and my new 3 wire fan of same model and same company(ebmpapst) has red, white and blue colour, so how to connect this new fan? Can you help me get the solution
16:30 Note: Apparently not all motherboards support 3-pin fan control on 4-pin headers; users should check the manual first. If not supported, 3-pin fans will run at full speed.
He is right all about what was said about how frequency works. We don't usually say times per second though. But more of how many cycles per second that the pulses are being made. "Cycles per second" or "pulses per second" sounds more straightforward instead of "times per second." Other than that, this is still one of the best educational video for people who would like to learn a little more about their computer fans and learn the differences between DC and PWM.
Hi, I have purchased a fan 12 volt and 2.8 Amps it draws about 33 watt full speed but it is too much air and consumption. If connected with the motherboard the fan spins and then turn off every 3 seconds. Maybe the motherboard connector release 12 volt but not the amount of amps required? So if I lower the voltage with a potentiometer, should I be able to lower the fans speed? I have a module that gives all the energy I want so 12 volt 2.8 amps I would like to try with this module to a smoother speed and lower consumption. Any tips to how do that?? thanks
Some washing machines run with 3 phase AC induction motor with its own motor control board (MCU as they call it). It converts 120VAC to 3 phase AC. The main board on the washing machine seem to basically supply the PWM. There is plug in from the main board to the motor board with 3 wires. The 3 pins are ground and 5VDC plus one more which I assume is the PWM. I supplied the Motor control with 120VAC and 5 volts to the PWM to see what happens. Nothing happened. So I am assuming the third pin ( pwm) is what sends the signal to the motor control to get going. If I have this right, is there a simple pwm supply one can get to do the job since this has its own inverter, I am just wondering.
Great video mate. I was wondering why my 3-pin fan that's plugged to a 4-pin connector is not controllable, checked the bios and turns out its still in PWM mode on that particular fan header. Switched to DC and it finally works. Cheers!
One little bit to add is that you can use a 4 pin fan on a 3 pin motherboard connector. The 4 pin fan just stays at 100% speed by default but by controlling the voltage it will slow down.
TLDR warning. :) Your oscilloscope may not be the issue... Setting the fan to a low value, it STILL needs enough power to run. The computer, at a lower duty-cycle, has to stay within an "operable cycle level". Your desires, by settings, are just a suggestion, ball-park, of what is a safe lower end, and a maximum higher-end. Some fans have an internal controller to assist with the desired signal too. So, even though the BIOS says go to 1%, the fans controller may have its own duty-cycle for the actual fans 1% desire. (So it doesn't stall. They monitor the hall-sensor to get the RPM and can sense when the fan has stalled, or is about to stall. Saving the fan from damage, due to USER or BIOS stupidity or failures. Like when someone tells a fan to be 1% duty, or the signal is lost from a PWM control-pin. That is usually a $10 fan vs a $20 fan. Relative to a $2-$5 Voltage-controlled fan.) Though, most normal fans can be controlled with a PWM controller. They just don't have the (on-board control wire). So you would just not be able to use BIOS to control the fans speed. Which is why they sell separate fan-speed controllers. Those can take the voltage-level, and turn the output into a PWM-12v signal to feed the fan power. At the end of the day, the reasoning of using PWM over Voltage-control, is for efficiency, control and longevity. Feeding a 12v pulse, for half the cycle, which is equal to a 6v load, will consume half the power/watts. It gives the motor the voltage it needs to "spin", where a solid 6v might not give the motor enough power to spin. Reducing the voltage, in essence, the motor is struggling to function, as it is TRYING to get to the top RPM, but it doesn't have enough power to. While a PWM pulse, the fan has adequate power to spin at the full RPM, but the fan loses power for a specified time-period, using momentum to sustain a slower speed-average. With the aid of capacitors and diodes to sustain the average power, in the power-outages. Resulting in a drop of AMPS, at a sustained voltage. As opposed to a drop of VOLTS, which results in raised AMPS being drawn, with a voltage-controlled setup. (Thus, lower wattage consumed.)
Excellent video! a question: Why does a new 4 pin fan spin then turn off after apprx 3 seconds on a 9v battery? The old 3 pin fan does the same.... why? I knew nothing about fans when ordering the replacement, so it is a 4 pin replacement for a Toshiba satellite C655. The old fan (3 pin) ran for approx 3 seconds and stopped (thought the problem was the fan) BOTH fans do not run continuously on a 9v battery! HOWEVER if I pulse the power to the red wire (+) maybe about 5 to 10Hz by hand... the fans run! Why?
The Toshiba will boot, the fan spins for 3 seconds and stops. (keyboard open so I see that it has stopped) The laptop runs normally for about 3 to 5 minutes, then slows due to overheating. I changed the connector on the 4 pin fan to a 3 pin (red black and yellow.. easy to change the pins in the plastic connector) Both fans stop after 3 seconds!
Amazing, the most comprehensive video I have seen on this topic. But what happens if I use a simple pwm fan speed controller in a mix of 3 and 4 pin fans?
Please answer me! When you set the mode to pwm and duty to 50% , your multimeter had to measure the average value of PWM signal. Right??!! How did it measure the max value ?? Can anybody answer me ??
An excellent video, I just want to add a note, not all motherboard manufacturers are so restrictive like ASUS about the DC fan control, some allow to set the voltage down until the fans stop, even in DC mode. I have four 200mm Cooler Master fans, and the difference between 50% and 60% is VERY audible, in an ASUS motherboard the sound is really annoying at 60% but in my new motherboard I can set the voltage I want in the BIOS, if I remember correctly, I have the fans at 6 or 7 volts at the beginning of “the curve” and the sound difference is really big. It’s a shame that ASUS don’t give the user full control of the fans… That was why I change brand. Still, I really think that this was the best video I’ve watched on the subject, congratulations to the guy with the cool beard.
İ dont know English much but to my understand pwm fans are working to CPU temparature exactly 4-pins fans are working auto (İf there is wrong thing please comment)Thanks
Absolutely awesome video....I have been working and building computers for 35 years I know about 3 pin and 4 pin PWM but I never knew exactly how it was doing it...never to old to learn. I do have a question though. Will the 3 pin last longer that a 4 pin fan because of the pulsing on and off? You know how computers are sending voltage on and off is the sure fire way to kill a electronic equipment.
hello my helios 300 one fan was stuck and another was working fine, I cleaned cooling system and repast thermal paste, now both fans sre spinning slow but not on system control and not showing fan details on software
I have a question? I have been using a 12 VDC fan and a 12 VDC really both connected to turn on 110 V compressor. This is used for a fridge. Well now this company which makes refrigerators I won’t say its name has installed a four cable fan and now it says it’s a 13 V fan. So I’m having issues because the fan stays on so the compressor stays on. So I’m wondering if there’s anything else I can do
im not sure you are right about dc control not being pwm...? did you try it with the scope? how would the motherboard generate variable voltage the circuits to pwm are so much simpler?
It's PWM in a different place. Yes, the voltage is stepped down via PWM on the motherboard, but the fan isn't being switched on/off by that voltage. A PWM fan at 50% has 100% torque due to being supplied with 12v, but a 3pin fan at "50%" is getting a constant voltage of 6v, massively reducing its torque. It's debatable as to how much this affects overall performance, but the important thing is that they're getting power in two very different ways.
Interesting, I always assumed that the motherboard dealt with the PWM directly on the power pins. In fact, I wonder why they don't just do that instead of sending the PWM signal on a separate pin.
Great explain. One video per year watched entirely. I'm serious. I'm going to change my case and wanted to know about the exact difference between those two and I got it. That was really helpful. I watched this for future to what should I choose when I would like to buy two fans for maybe new case. Maybe it will not be necessary. I don't also comment too much videos and give thumbs up but there I should for giving 1:1 all knowledge about fans. Keep it going!
My scope is terrible and I wouldn't trust its voltage measurements, but in any case, the PWM signal is just a signal, it doesn't do any work per se. So it can be very low, somewhere between 0.5 and 1v, and switch a mosfet inside the fan on and off.
What is the best way for me to run 8 pwm SickleFlow blue led 120mm 12VDC 4-pin fans, to my 4-pin headers on my mobo Asus Sabertooth 990fx R2.0 without the fans running full throttle like it has been while using a hub like connection.
Which one of the mods ( dv or pwm ) can let you have your system auto adjust the speeds dependant on temperaturs and other variables that the system monitors and does on it's own ?
Auto-control can be done with voltage or PWM, just depends on what fan controller you're using. Most modern mobos will allow voltage control down to low levels. The main difference is purely what range a fan has. 4 pin fans tend to have a much wider speed range.
9:35 I believe you can. With the rotary knob in the Hz position, press the Hz/Duty soft button right under the screen, and the meter will probably show you the duty cycle in percents.
What about the fans that use a molex connector? are they at least controllable like the 3pin or it just gives the maximum electricity that the fan can handle
Molex just runs the fan at full speed. It is possible to re-pin the molex connector for different set voltages, you can reverse the pin order to get 5v, or swap 2 and 4 to get 7v across the fan. We used to do this all the time in the early 00s before fan controllers were a thing - but these days I would advise against it, as the mods were pretty jank and will melt something if you get it wrong.
Would it be a good rule of thumb that on modern motherboards you should only combine 4-pin pwm fans on the same header and 3-pin fans should have their own headers?
Correct yes, if you're using fan splitters, each splitter should have the same type of fan on it. Mixing won't cause any damage, but you won't have proper control over the fans. Either the PWM fan would be DC controlled (resulting in lower performance) or the 3pin fan would run at full speed all the time, as it can't receive the PWM signal.
@@Adamant_IT Yeah. I ran out of headers and tried to use a splitter on a pwm cpu and a 3-pin top case fan. My 3-pin fan wasn't even spinning. So I paired two pwm cpu fans together on their own header and left the 3-pin on it's own header and everything works fine. I think pwm fans are better for combining onto a single header -- I have 4 fans on the case door and they are working fine combined on the high-voltage fan header.
4-pin fans default to full speed if they don't receive a PWM signal (ala, if you disconnect the 4th pin). But if you're doing that, you might as well use a 3-pin fan.
I have a gigabyte aorus Rtx 4090 extreme watherforce. It has a aio cooler with 3 fans. Unfortunately I don’t have the original fans, and I don’t know the original fan amperage. My question is, do you think it damage my card if the amperage is a bit lower or higher by 0.5 A ? Is there possibility for me to measure the ampere on the graphic card ? It has one 4pin and two 3pin socket.
Thank you! I've been hunting long and hard for PWM fans what will accept 0 duty cycle = 0 rpm and finding none of them wanting to claim less than 200rpm. That's fine, but I want them to STOP completely. Buying fans that clearly supported 0rpm or 0db mode seemed like it could cost me two hundred quid/dollars. Understanding that the DC fans, given the motherboard allows it (Crosshair Hero), WILL allow me to STOP "secondary" fans. Fans that provide filtered intake air for exhaust radiators, for example. I don't mind the main radiator fans being PWM doing 200rpm and not stopping, they are pretty silent. However for these only on under high loads fans.... DC fans will work I don't see many use cases where you would want to run intake support fans at anything less than 50% anyway. As I have a collection of DC fans and DC fans are cheaper that could save me a lot of money, especially as I'm 90% sure the CH H8 supports DC fan "Fan OFF" mode and 0% voltage.
Oddly the Alphacool rad fans are DC. Their label says 0-2500RPM. It will be interesting to see what their minimum stable speed is. I also have a box of fractal design DC fans I can test.
Also, note... The CPU fan headers are different. CPU FAN, OPT, PUMP and AIO_PUMP are all guarded to not go to 0 (PWM or DC). The CH H8 limits these to 40% minimum. This is where good quality PWM fans are king. They will adapt to the 40% minimum and be silent and slow at that, higher than normal limit on PWM. Chassis fans and aux fans will accept 0% and "Fan Off" states... at least on the CH.
Can someone explain please if there is a difference in power consumption? I would imagine both types consume roughly 50% of energy if operation speed is at 50% ?
Voltage control generally uses resistors to control voltage. The extra power is just dissipated through the resistors when the fan spins slower so overall, no energy is really saved. Whereas in PWM, the 12v is turned on and off internally in the fan, which does save some energy
Excellent explanation, thank you, I just had one question... Does this mean 3-pin fans use less energy? Or is the consumption similar because a PWM fan is "off" for a certain part of the time rather than just permanently drawing less power? Thank you
I guess that's why my BIOS starts the PC with 2 of my fans at 100% because those fans are only 3 pins so voltage. The other frontal and back fans are 4 pins and are pretty silent so PWM. How do I set the voltage for my 2 up fans from AsRock B450m Bios so they start silent? Thanks.
I don't know ASRock boards very well, but at a quick look, I think that your motherboard doesn't allow separate control of non-PWM fans. A cheap solution without replacing the fans or getting a dedicated fan controller would be some inline resistors like these: www.ebay.co.uk/itm/133216619324
How about 4 pin PWM fans with an breakout for a second fan, I note the extension has 3 pins, is it fair to assume the speed sense is off the primary fan and both fans will be PWM controlled to the same signal/speed?
I have bought some 4-pin PWM fans but I can't get them to turn off when not needed (at 0% duty cycle the still spin at 600 rpm). The only way is to put them in DC mode and then select 0%. Will it cause any problem in the long run to use DC mode on PWM fans (maybe on the PWM controller)?
Won't damage them, no. The fan controller must not have a true 0% on it, as 0% by definition must switch off. It's not uncommon to see fan controllers not go below 20%, so maybe your one is like that, but just doesn't report it properly.
I would have wished to see schematics, like, is there a mosfet in the PWM fan ? What are the pins ? How can I hack fans for other projects ? Excellent explanation non the less.
If you liked this video, I demonstrate this in action on LEDs in this video here: ruclips.net/video/sPkRS2sLXh8/видео.html
I did like this video and it gave me some ideas for my situation. Question for you though. I just built a pc with the Cooler Master H500m on the ASUS Crosshair VIII Hero (Wifi) mobo. The case has 2 200mm DC fans that turn on and off on silent mode like you were talking about. I was thinking of replacing them but now I'm wondering if there's a controller that can treat them like pwm and also make them visible in Fan Xpert 4. Any thoughts on that? Also, why does it seem like 200mm fans don't really come PWM except for Thermaltake but those have a proprietary connector that isn't just a regular PWM fan connector? Confusing stuff... Anyways, appreciate any help you could provide. Thanks!
Some server fans are 4 pin with some of the pins wired a little differently. I've completely forgotten the "standard" they use. I think it was just an annoying OEM rewiring. I was (about 10 or 15 years ago) able to re-pin the server fan to work with out issue on a standard pc motherboard. It was a nice fast dangerous delta fan from memory :D
Nice explanation. So why they don’t control the DC fan via current?
@@tamaica7770 Stepping up and down voltages is rather easy to do. Stepping up and down current involves more work and other various reasons. More current = more heat for example.
skillet pan heat depends on P=IV. Lets say the fan can work at 6V, and mobo gives it current to operate. Increase current, increase the speed of fan. just not sure how they decided to choose V instead of I. maybe for safety?
Finally - someone explaining it in a way that can be understood
I love how thorough you are about this subject and that everything mentioned is relevant and on track. The only thing I don't like is there is no discussion about the 4 pin molex connector fan type.
A year on from making this video and still the best explanation on RUclips I've seen 👏🏻
Damm! This man is unbelievable! I just wish I had known this channel before. Of the dozens of computer channels I know, this one is by far the best. What a fantastic teaching method! he doesn't need edits, animations or a ton of fancy equipment to perfectly explain a subject. That's talent! Thank you!
agree
"My fan doesn't work."
"Have you tried turning it on and off again 25 000 times a second?"
😂
-"off and on". What you wrote is both an incorrect reference -*-AND-*- doesn't make sense.-
I'm a dope.
Voltage control uses resistance to lower the tension, therefore as a result it consumes more power due to the loss of energy as a form of heat; this makes PWM a power saver. I really liked your channel right away after watching a couple of videos because obviously it is not made for showing off but for teaching purposes, “not too many infos + accuracy at its best = subscribe”. Thank you.
- 3 Pin regulates through the amount of "juice" it sends to the fan. The more power the faster the fan spins.
- 4 Pin regulates through timings, for how long the fan is on on end.
Pretty simple. Can sum it up in two paragraphs.
Great video, well done.
There are few people who can articulate tech (and engineering) stuff like this so well and more importantly present it as well...good work!
Definitely this is the best PWM vs Dc explanation on RUclips. Please keep up your great job!
Realy nice video. I've enjoyed it a lot. Few time ago i've to buy a new processor fan because mine have burned after 6 years of use. I almost bought the 3 pin fan but after some internet research i discovered that processors use pwm (4 pin) fans. Then i bought the right fan for my processor and i am still waiting for it to come. Thanks for nice content. Regards. Rodrigo.
Nice visual of "pulse width modulation".............. hope folks appreciate the lesson!....................... Nicely done!........................... AJ
I wasn gonna skip the video when i saw the length but hearing you explaining something is so interesting i watched the whole thing.
Brilliant explanation, this is so useful in understanding how car radiator fans work and when your ac doesnt work with the fan going into full speed mode!
Your explanation and demonstration of voltage controlled fans versus PWM control was brilliant. Thank you.
Understand clearly now why my TUF Gaming X570 MB has the "fan tuning" function that calibrates fan speeds. Appreciate the content.
As others have stated, this is an excellent and underrated video.
It answered all my questions and more. Well done.
Video is fantastic, these are found on central heating appliances too, which now have modulating fans and pumps which work via a PWM signal from PCB. Normally heating engineers don't carry oscilloscopes, even cheap ones, so we are reduced to using Hz on a multi-meter.
THANK YOU! I bought 2 new fans for an older pc and I did not realize until today that I have 3 pin fans while the sockets on the motherboard are 4 pins. So I was sad that I must send back and buy new ones. You are a lifesaver and your explanations are amazing and easy to understand.
Interesting and very useful, thank you
Another beauty of PWM control: I'm a model train enthusiast. My DC trains can run on straight DC or PWM, but using PWM allows the lighting to remain at full brightness, and the sound function at full volume, at any speed. The full voltage is always available for a fraction of the time and is used to charge capacitors to power the accessories. Wall warts / power adapters can be made without heavy, expensive copper windings if they use PWM.
Wow! This was a fantastic demonstration of how PC fans work. I've been struggling to understand the difference between 3 and 4 pin fans and now I have a clear understanding of the two.
I didn't know whether the fan was driven by the power of PWM signal or not, but now I see the maximum voltage of the PWM line only goes up to about half a volt, so my question is answered. This video is exactly what I was looking for. The power if provided by the 12-volt line and the PWM acts as a sort of relay that switches the 12 volts on and off. Thanks for the video.
Underrated video... This deserves more views.
been looking everywhere for that information and there it is explained clearly and succinctly. thank you for the great video sir!
Duty cycle is still correct. even voltage controlled fans are controlled over a PWM dutycycle on the 12V rail. The fact that a motor is an inductor flattens the 12v-0v difference and averages it out to a voltage between the two values. with an oscilloscope you'd be able to see, that the line in fact has 12V peak and just switches on and of rapidly.
There is probably a capacitor also on the power rail which will average out the voltage so you will probably not see the full 12-to-zero volt swing on the line using a scope. But I agree that the voltage is most probably varied using PWM even for a 3-pin fan.
A few years ago I designed an AtTiny85 based DIY fan controller, and checking the tach in DC mode was a royal pain, because in the off/low part of the duty cycle, there was no power for the tach signal, so I had to run a burst of 100% DC (pulse stretching) every time I wanted to check the time between two tach pulses, which led to audible noise, especially at low RPMs. With a capacitor between the 12V and GND maybe the voltage would be flat enough to get a readable tach signal even in the off part. OTOH, the capacitor would cause the circuit to draw more current in the on part as the capacitor was refilled, and for the same reason it would also make the fan run faster at lower duty cycles because the average current would be higher, making speed control more difficult and unpredictable at low duty cycles. Unless I somehow found a way to feed the fan a PWM signal not between 0V and 12V, but between 5V and 12V, so that the tach (hopefully) always stayed readable with 5V logic. But then again, that would prevent the fan from spinning down below around 40%. (Hey, did I just stumble upon the reason for that particular limit in the BIOS?) Nah. My conclusion remains, 3 pin fans suck for speed control.
This is invaluable! You need a Patreon, I'd love to support your channel. Thanks!
Thanks for making this video. I learned a lot about these fans and their respective differences. Keep up the great informative work!
I just got back into building PCs again since 2002. This was a great video for clearing some confusion I have with my case fans in my new PC. Thank you.
Well done on the growth of your channel and thanks for a brilliant explanation, I just got really nice be quiet case with 3x3pin fans and 2x 4 pin fans, I was a tad disappointed but eh still very cool, now I know that I can use them together reasonably, maybe not as 'cool' as 5x4pins but I think I can live with it for now, AUTO sounds really useful! ;D
My question is, with a PWM fan, is the actual motor coil attached to the fans internal controller still two wires?
If so couldn't they just keep the 3 wire method and change the motherboards control to do the PWM directly?
I'm guessing that the fan motor itself is more complicated, possibly with multiple windings or something similar and uses a dedicated motor driver to provide more granular control?
Main issue is that if you drive the fan by its 12v wire, your fan controller has to switch the load of the fan. But with a dedicated PWM wire, the fan controller just sends a digital signal, and the switching is done by the fan itself - so a big fan that pulls a lot of power won't overload the controller.
Additionally, yes a brushless motor usually has four coils in it that need to be sequenced, so there will be a driver for that in the fan, as well as the speed sensor. These two chips may behave unpredictably if their 12v power supply is getting switched on and off at 23KHz.
There were some very early PWM fan controllers, mostly custom jobs, that were designed to drive the 12v pin, and while I never used one they were known for causing an electronic humming noise from the fan.
Man, this video explained much more than I anticipated. Great job!
The Asus motherboard does allow under 60% voltage but only if you run the automatic fan configuration first. It tests the RPM range of the fans.
I didn't understand the differences until now :) Thanks Graham
Thank you for the best explanation of how DC and PWM fans work. Thank you for your hard work.
Superior explanation AND illustration of the differences between PWM(4 pin) and 3 pin fan operation and logic.
You explained it really well. Now I am in a problem with colour codes of wire. My old 3 wire fan had brown, red, and black colour wires, and my new 3 wire fan of same model and same company(ebmpapst) has red, white and blue colour, so how to connect this new fan?
Can you help me get the solution
Thank you for this video. It was well worth watching and easily understood. Terry from Australia. 11/05/2024
16:30 Note: Apparently not all motherboards support 3-pin fan control on 4-pin headers; users should check the manual first. If not supported, 3-pin fans will run at full speed.
some people are very good at explaining tech related topics, but they should call you "Master", best explanation ever, thanks alot.
He is right all about what was said about how frequency works. We don't usually say times per second though. But more of how many cycles per second that the pulses are being made. "Cycles per second" or "pulses per second" sounds more straightforward instead of "times per second." Other than that, this is still one of the best educational video for people who would like to learn a little more about their computer fans and learn the differences between DC and PWM.
Really informative video, I'm curious what probes you use. Mine are really wide and hard to get into a small area like that
Hi, I have purchased a fan 12 volt and 2.8 Amps it draws about 33 watt full speed but it is too much air and consumption. If connected with the motherboard the fan spins and then turn off every 3 seconds. Maybe the motherboard connector release 12 volt but not the amount of amps required? So if I lower the voltage with a potentiometer, should I be able to lower the fans speed? I have a module that gives all the energy I want so 12 volt 2.8 amps I would like to try with this module to a smoother speed and lower consumption. Any tips to how do that?? thanks
Can I connect two wires + and - 12V DC to this 4pin PWM FAN???? PLZ TELL ME!!! THANKS U👍
Hard disk Drive ,cpu , Motherboard, Case fans
Which of these components does connect directly to the power supply? Choose two
Very good video, ty! I had a question. Can u also connect a 4 pin fan into a 3 pin connector (using the same logic 3 into 4)? thx a lot!
Yes you can, however you'll only be able to do voltage control on it, because PWM pin won't be hooked up.
I come from a time where we had 2 pin and 3 pin fans xD
Some washing machines run with 3 phase AC induction motor with its own motor control board (MCU as they call it). It converts 120VAC to 3 phase AC. The main board on the washing machine seem to basically supply the PWM. There is plug in from the main board to the motor board with 3 wires. The 3 pins are ground and 5VDC plus one more which I assume is the PWM. I supplied the Motor control with 120VAC and 5 volts to the PWM to see what happens. Nothing happened.
So I am assuming the third pin ( pwm) is what sends the signal to the motor control to get going.
If I have this right, is there a simple pwm supply one can get to do the job since this has its own inverter, I am just wondering.
Great video mate. I was wondering why my 3-pin fan that's plugged to a 4-pin connector is not controllable, checked the bios and turns out its still in PWM mode on that particular fan header. Switched to DC and it finally works. Cheers!
Absolutely perfectly explained!! If only all teachers in engineering colleges had this ability.
Can I plug a 3 pin fan into a 4 pin header with the tach pin missing? Where the layout is: Ground, 12V, blank, PWM.
Good explanation and examples of the differences. Thanks very much.
Can you use a PWM fan on a 3 pin header?
Yes, but it will function as a 3pin fan, so you'll lose the PWM benefits entirely.
Absolutely BRILLIANT explanation of this. Thank you!
One little bit to add is that you can use a 4 pin fan on a 3 pin motherboard connector. The 4 pin fan just stays at 100% speed by default but by controlling the voltage it will slow down.
How can I power a noctua 4-pin fan to the mains? I wired the +12v connector and the ground connector but the fan doesnt work
TLDR warning. :)
Your oscilloscope may not be the issue... Setting the fan to a low value, it STILL needs enough power to run. The computer, at a lower duty-cycle, has to stay within an "operable cycle level". Your desires, by settings, are just a suggestion, ball-park, of what is a safe lower end, and a maximum higher-end. Some fans have an internal controller to assist with the desired signal too. So, even though the BIOS says go to 1%, the fans controller may have its own duty-cycle for the actual fans 1% desire. (So it doesn't stall. They monitor the hall-sensor to get the RPM and can sense when the fan has stalled, or is about to stall. Saving the fan from damage, due to USER or BIOS stupidity or failures. Like when someone tells a fan to be 1% duty, or the signal is lost from a PWM control-pin. That is usually a $10 fan vs a $20 fan. Relative to a $2-$5 Voltage-controlled fan.)
Though, most normal fans can be controlled with a PWM controller. They just don't have the (on-board control wire). So you would just not be able to use BIOS to control the fans speed. Which is why they sell separate fan-speed controllers. Those can take the voltage-level, and turn the output into a PWM-12v signal to feed the fan power.
At the end of the day, the reasoning of using PWM over Voltage-control, is for efficiency, control and longevity. Feeding a 12v pulse, for half the cycle, which is equal to a 6v load, will consume half the power/watts. It gives the motor the voltage it needs to "spin", where a solid 6v might not give the motor enough power to spin. Reducing the voltage, in essence, the motor is struggling to function, as it is TRYING to get to the top RPM, but it doesn't have enough power to. While a PWM pulse, the fan has adequate power to spin at the full RPM, but the fan loses power for a specified time-period, using momentum to sustain a slower speed-average. With the aid of capacitors and diodes to sustain the average power, in the power-outages. Resulting in a drop of AMPS, at a sustained voltage. As opposed to a drop of VOLTS, which results in raised AMPS being drawn, with a voltage-controlled setup. (Thus, lower wattage consumed.)
Excellent video! a question: Why does a new 4 pin fan spin then turn off after apprx 3 seconds on a 9v battery? The old 3 pin fan does the same.... why? I knew nothing about fans when ordering the replacement, so it is a 4 pin replacement for a Toshiba satellite C655. The old fan (3 pin) ran for approx 3 seconds and stopped (thought the problem was the fan) BOTH fans do not run continuously on a 9v battery! HOWEVER if I pulse the power to the red wire (+) maybe about 5 to 10Hz by hand... the fans run! Why?
The Toshiba will boot, the fan spins for 3 seconds and stops. (keyboard open so I see that it has stopped) The laptop runs normally for about 3 to 5 minutes, then slows due to overheating. I changed the connector on the 4 pin fan to a 3 pin (red black and yellow.. easy to change the pins in the plastic connector) Both fans stop after 3 seconds!
Great work. Fantastic explanation. This is an absolute gem of a video.
This is the first video I have watched on this channel and I liked it. Therefore: Subscribed.
Hows that vc99 run? I was looking at multimeters and reviews on amazon...it didnt do so well...whats your experience and thoughts on it?
Do you know how can I find the 12v and GND in a 4 wire fan if all the wires are the same colour? Note: without frying the motor like I just did?
Amazing, the most comprehensive video I have seen on this topic. But what happens if I use a simple pwm fan speed controller in a mix of 3 and 4 pin fans?
Please answer me!
When you set the mode to pwm and duty to 50% , your multimeter had to measure the average value of PWM signal. Right??!! How did it measure the max value ?? Can anybody answer me ??
An excellent video, I just want to add a note, not all motherboard manufacturers are so restrictive like ASUS about the DC fan control, some allow to set the voltage down until the fans stop, even in DC mode. I have four 200mm Cooler Master fans, and the difference between 50% and 60% is VERY audible, in an ASUS motherboard the sound is really annoying at 60% but in my new motherboard I can set the voltage I want in the BIOS, if I remember correctly, I have the fans at 6 or 7 volts at the beginning of “the curve” and the sound difference is really big.
It’s a shame that ASUS don’t give the user full control of the fans… That was why I change brand.
Still, I really think that this was the best video I’ve watched on the subject, congratulations to the guy with the cool beard.
You my friend are a life saver!!
Great respect!
İ dont know English much but to my understand pwm fans are working to CPU temparature exactly 4-pins fans are working auto (İf there is wrong thing please comment)Thanks
Absolutely awesome video....I have been working and building computers for 35 years I know about 3 pin and 4 pin PWM but I never knew exactly how it was doing it...never to old to learn. I do have a question though. Will the 3 pin last longer that a 4 pin fan because of the pulsing on and off? You know how computers are sending voltage on and off is the sure fire way to kill a electronic equipment.
hello my helios 300 one fan was stuck and another was working fine, I cleaned cooling system and repast thermal paste, now both fans sre spinning slow but not on system control and not showing fan details on software
And what if you check gnd and+12v of 3-pin fan with oscilloscope while changing that "duty cycle". )
I have a question? I have been using a 12 VDC fan and a 12 VDC really both connected to turn on 110 V compressor. This is used for a fridge. Well now this company which makes refrigerators I won’t say its name has installed a four cable fan and now it says it’s a 13 V fan. So I’m having issues because the fan stays on so the compressor stays on. So I’m wondering if there’s anything else I can do
im not sure you are right about dc control not being pwm...? did you try it with the scope? how would the motherboard generate variable voltage the circuits to pwm are so much simpler?
It's PWM in a different place. Yes, the voltage is stepped down via PWM on the motherboard, but the fan isn't being switched on/off by that voltage.
A PWM fan at 50% has 100% torque due to being supplied with 12v, but a 3pin fan at "50%" is getting a constant voltage of 6v, massively reducing its torque. It's debatable as to how much this affects overall performance, but the important thing is that they're getting power in two very different ways.
I have a pwm fan from fractal that doesn't rpm lower that 580 but when i put it to DC it can low as 300 rpm.
What are your thoughts?
Interesting, I always assumed that the motherboard dealt with the PWM directly on the power pins. In fact, I wonder why they don't just do that instead of sending the PWM signal on a separate pin.
Great explanation. It really has shown how much more control de PWM fans have, thanks!
Great explain. One video per year watched entirely. I'm serious. I'm going to change my case and wanted to know about the exact difference between those two and I got it. That was really helpful. I watched this for future to what should I choose when I would like to buy two fans for maybe new case. Maybe it will not be necessary.
I don't also comment too much videos and give thumbs up but there I should for giving 1:1 all knowledge about fans.
Keep it going!
Great content. Subscribed and thanks for all the hard work.😀
I'm confused about voltage needed on singal cable. Does your osciloscope shows 0-0.5V range? Does this need to be scaled to 12V?
My scope is terrible and I wouldn't trust its voltage measurements, but in any case, the PWM signal is just a signal, it doesn't do any work per se. So it can be very low, somewhere between 0.5 and 1v, and switch a mosfet inside the fan on and off.
Sir: Thank you very much for such a clear concise explanation. I was able to share this video and the concept was understood perfectly. 👍
What is the best way for me to run 8 pwm SickleFlow blue led 120mm 12VDC 4-pin fans, to my 4-pin headers on my mobo Asus Sabertooth 990fx R2.0 without the fans running full throttle like it has been while using a hub like connection.
Which one of the mods ( dv or pwm ) can let you have your system auto adjust the speeds dependant on temperaturs and other variables that the system monitors and does on it's own ?
Auto-control can be done with voltage or PWM, just depends on what fan controller you're using.
Most modern mobos will allow voltage control down to low levels. The main difference is purely what range a fan has. 4 pin fans tend to have a much wider speed range.
9:35 I believe you can. With the rotary knob in the Hz position, press the Hz/Duty soft button right under the screen, and the meter will probably show you the duty cycle in percents.
Hi! Do 3pin pumps work the same? So do you need to change the setting in the BIOS?
Yup, same applies to water pumps.
What about the fans that use a molex connector? are they at least controllable like the 3pin or it just gives the maximum electricity that the fan can handle
Molex just runs the fan at full speed. It is possible to re-pin the molex connector for different set voltages, you can reverse the pin order to get 5v, or swap 2 and 4 to get 7v across the fan. We used to do this all the time in the early 00s before fan controllers were a thing - but these days I would advise against it, as the mods were pretty jank and will melt something if you get it wrong.
Adamant IT thank you
Would it be a good rule of thumb that on modern motherboards you should only combine 4-pin pwm fans on the same header and 3-pin fans should have their own headers?
Correct yes, if you're using fan splitters, each splitter should have the same type of fan on it. Mixing won't cause any damage, but you won't have proper control over the fans. Either the PWM fan would be DC controlled (resulting in lower performance) or the 3pin fan would run at full speed all the time, as it can't receive the PWM signal.
@@Adamant_IT Yeah. I ran out of headers and tried to use a splitter on a pwm cpu and a 3-pin top case fan. My 3-pin fan wasn't even spinning. So I paired two pwm cpu fans together on their own header and left the 3-pin on it's own header and everything works fine. I think pwm fans are better for combining onto a single header -- I have 4 fans on the case door and they are working fine combined on the high-voltage fan header.
So the only way to make the 4pin pwm fan work at full speed is through your computer. You can't cut the 4pin and make a 'do it your self project' ?
4-pin fans default to full speed if they don't receive a PWM signal (ala, if you disconnect the 4th pin). But if you're doing that, you might as well use a 3-pin fan.
Another great video, very clear to understand and a useful guide when buying PC fans.
Thank you for a very informative video with no unnecessary fluff. Good info!! I gave you a thumbs up & subscribed.
I have bought a 4 pin fan for a different application. Can I use a 0-5v on the PWM pin to control the fan speed?
No, it needs to be a a square-wave. If you can only do voltage control, then just do 0-12v on the power pins.
I have a gigabyte aorus Rtx 4090 extreme watherforce. It has a aio cooler with 3 fans. Unfortunately I don’t have the original fans, and I don’t know the original fan amperage. My question is, do you think it damage my card if the amperage is a bit lower or higher by 0.5 A ? Is there possibility for me to measure the ampere on the graphic card ? It has one 4pin and two 3pin socket.
Thank you! I've been hunting long and hard for PWM fans what will accept 0 duty cycle = 0 rpm and finding none of them wanting to claim less than 200rpm. That's fine, but I want them to STOP completely. Buying fans that clearly supported 0rpm or 0db mode seemed like it could cost me two hundred quid/dollars.
Understanding that the DC fans, given the motherboard allows it (Crosshair Hero), WILL allow me to STOP "secondary" fans. Fans that provide filtered intake air for exhaust radiators, for example. I don't mind the main radiator fans being PWM doing 200rpm and not stopping, they are pretty silent. However for these only on under high loads fans.... DC fans will work I don't see many use cases where you would want to run intake support fans at anything less than 50% anyway.
As I have a collection of DC fans and DC fans are cheaper that could save me a lot of money, especially as I'm 90% sure the CH H8 supports DC fan "Fan OFF" mode and 0% voltage.
Oddly the Alphacool rad fans are DC. Their label says 0-2500RPM. It will be interesting to see what their minimum stable speed is. I also have a box of fractal design DC fans I can test.
Also, note... The CPU fan headers are different. CPU FAN, OPT, PUMP and AIO_PUMP are all guarded to not go to 0 (PWM or DC). The CH H8 limits these to 40% minimum. This is where good quality PWM fans are king. They will adapt to the 40% minimum and be silent and slow at that, higher than normal limit on PWM. Chassis fans and aux fans will accept 0% and "Fan Off" states... at least on the CH.
Can someone explain please if there is a difference in power consumption? I would imagine both types consume roughly 50% of energy if operation speed is at 50% ?
Voltage control generally uses resistors to control voltage. The extra power is just dissipated through the resistors when the fan spins slower so overall, no energy is really saved. Whereas in PWM, the 12v is turned on and off internally in the fan, which does save some energy
Just the answers I was looking for. Thanks
I have a 4 wire fan with non-traditional wire colors, Blue, gray, black and brown. How can I determine which wire is which function?
www.ekwb.com/wp-content/uploads/2016/07/02-1.jpg
Excellent explanation, thank you, I just had one question... Does this mean 3-pin fans use less energy? Or is the consumption similar because a PWM fan is "off" for a certain part of the time rather than just permanently drawing less power? Thank you
Correct, the PWM fan is functionally "off" on the low cycles, so will be pretty similar to 3pin.
I guess that's why my BIOS starts the PC with 2 of my fans at 100% because those fans are only 3 pins so voltage.
The other frontal and back fans are 4 pins and are pretty silent so PWM.
How do I set the voltage for my 2 up fans from AsRock B450m Bios so they start silent?
Thanks.
I don't know ASRock boards very well, but at a quick look, I think that your motherboard doesn't allow separate control of non-PWM fans. A cheap solution without replacing the fans or getting a dedicated fan controller would be some inline resistors like these: www.ebay.co.uk/itm/133216619324
@@Adamant_IT Yes, I will replace them with some beautiful RGB ones, but until then, what software could I use in Windows ? Maybe I have luck.
How about 4 pin PWM fans with an breakout for a second fan, I note the extension has 3 pins, is it fair to assume the speed sense is off the primary fan and both fans will be PWM controlled to the same signal/speed?
How do you tell if your mobo can control 3 pin fans?
I have bought some 4-pin PWM fans but I can't get them to turn off when not needed (at 0% duty cycle the still spin at 600 rpm). The only way is to put them in DC mode and then select 0%. Will it cause any problem in the long run to use DC mode on PWM fans (maybe on the PWM controller)?
Won't damage them, no. The fan controller must not have a true 0% on it, as 0% by definition must switch off. It's not uncommon to see fan controllers not go below 20%, so maybe your one is like that, but just doesn't report it properly.
I would have wished to see schematics, like, is there a mosfet in the PWM fan ? What are the pins ? How can I hack fans for other projects ?
Excellent explanation non the less.
Thank you on behalf of all mankind. Perfect explanation.