Pulse Width Modulation EXPLAINED

▶ Our studio gear on Amazon:

Subscribe to the channel!

MUSIC:
'Astronaut' by JWeo
Provided by Silk Music

[MrStereo]

Knew the PWM thingy already, but luv getting re-schooled/refreshed on these topics (cause Iam old :p ) as SS makes these vids so interesting vs online reads or paperback.Thaaaanks Greg

[whocares_yes]

I knew about PWM and how it worked, but dammit I learned something new. You got me again, Greg. Nice vid, as always.

[TrevorSuarez]

FANTASTIC explanation, Greg. Your use of a guitar to explain the amplitude and attenuation was clever and incredibly helpful. One of my favorite videos of yours so far. It felt very much like an "Engineering Explained" but for computers rather than for cars/mechanical-engineering.

[Exponaut_R-]

I had a vague idea of what PWM fans were but this video gave me a better understanding.

[semuhphor]

Thanks
I was just talking about PWM with a colleague. Great explanation!!

[OldBuford]

Very nice explanation, I never really understood how the fans could be getting constant 12v yet be able to run at 25% speed but now that you explained it, it makes perfect sense, thanks!

[Atilolzz]

Can't wait to watch, glad you keep making such nice content !

[deepanchatterjee]

The higher the ON TIME of a signal, the higher the UP TIME of the fan will be. The fan basically stops when the N TIME[ data bit = 1, on time] ends and tries to stop due to next bit being 0 i.e. OFF TIME but, the motion keeps it going and as the PWM waveform switches states in nanosecond scale, the state change of fans[ off-on-off-on-...] is NOT visible to us human eyes. GG, Electronics student here, the first tech channel I've seen who at least knows what he is dealing with and let's the community know it. A big thank you Greg sir, good wishes from India. :)
*Science Studio, hence justified! <3

[FKoto]

I'm using a retro Cooler Master ATC 200 case so the I definitely appreciate the PWM technology. Since my case only allows 80mm fans, letting them sync to my bios' fan curve to modulate their rpms makes it much nicer when I'm just casually using my desktop.

[Drew-ixih]

nice man, good informative video. Love to see this kind of content, I use and suggest your videos to people getting into PC building and videos like this make explaining things much easier.

[TheSickness]

In addition, in the LED age and lighted fans, running lighted fans in varying dc mode, lights will change brightness accordingly. In pwm 12v are always there and lights will stay bright

[TheIan]

So just one week after presenting my PWM circuit as my final project and getting a bad score, you upload this video...

[edgar]

Nice guitar dude. You think you would do more videos on instruments, amps, or pedals. I would love to see a Science Studio take on these

[BIGBASSSAMA_]

never stop making informative vids like these thanks man

[EastAngliaUK]

Surprised a video about fans is interesting.

[Psyopcyclops]

Perfect explanation, thanks! Personally I decided to go for the Silent Wings 3 DC fans though, as I found them for half the price of the PWM. Would’ve preferred the PWM, but I can’t go past a deal lol.

[gazlink]

You didn't explain WHY PWM control is better than DC control. DC fans can be adjusted almost as well as PWM fans, the low speed operation is just a minor bonus that PWM control gets, but it's absolutely not why they are used, and better.
The reason PWM is better than DC control for fans is all in the power supply.
Providing variable power to a fan by varing it's DC voltage requires another power supply, similar to a Vcore VRM.. a hell of alot simpler, it's typically a simple buck converter, but this is inherently less efficient than simply giving the fan 12V and the fan doing the work of regulating it's power draw (which is either on or off so no wasted power).
This is why your PWM motherboard headers can easily take 2A current, if not more (the simply connect straight through to the 12V of the main PSU), but most DC controlled headers can only provide up to 1A safely. Above that and the buck converter built into the board will overheat, especially in the 50-80% region.
This is the same principle as any mosfet consuming power - it consumes most power when switching between on and off, either fully on or fully off it doesn't consume much power at all, (and why your CPU draws almost no power when not switching) - but when switching has it's highest value of current*impedance, meaning the I*I*R power losses are highest. The mosfets in the buck converter will be in this state most when providing between 50-80% of their input voltage of 12V. Some well designed fan power supplies will sink the most power at full load, and be more efficient at lower voltages.
Either way, with PWM all of this is negated, other than the much simpler and more efficient mosfet in the fan hub itself with far lower switching frequency (it only needs to switch typically once per 1/4 fan revolution if at all, so about 100 Hz) which is much less than a typical buck converter mosfet switching frequency in the DC-DC converter, which switch in the 300 KHz range. This is the only way to do DC to DC for power electronics, it's still a switched mode power supply and so ~85% efficient, but chopping it to AC, and back to DC is the only way. Using a linear step-down like a variable resistor or those resistors you can put on fan cables would be more like 50% efficient, so it has to contain switching mosfets, capacitors and inductors - again not needed for PWM, other than a much simpler more power efficient design that switches only when the fan needs to - for powering the coil for part of it's cycle.
Sharing a PWM signal to control 10 PWM fans is easy, anyone with a soldering iron can make the connections, assuming powerful fans or water pumps etc.. - 5A draw at 12V, no problem, and nothing else needed other than the PWM signal from the CPU fan header, and 60W on the 12V rail of your main PSU (and routing 5A through the board or wires, however they are connected, which is no issue).
Controlling 10 DC fans from a DC signal that can only provide 1A max (so can only be used as a signal not power supply), or using a PWM signal to modulate a buck converter to make 5-12V DC up to 60W - the only way to control those fans - means an actual 70W power supply that costs as much as 70W DC-DC converters cost, and dissipating 10W of heat from the extra power supply, and routing 6A to it from your PSU).

This is why PWM control is better than DC control, and why if you deck out your system with fans, only a few can be DC controlled if your motherboard does DC control at all.

[EvelynRoods]

Beautiful guitar. I don’t have any Les Paul shape guitars, only superstrat style (Ibanez RG320 from 2001, and a squier strat that I’ve modified).

[Nostrum]

Nice video. Some fans are already so silent you won't hear them even in DC mode much, especially with low TDP CPUs. Still don't see the killer advantage in PWM apart from extremely low RPMs (versus not running at all).

[TheRguru]

Excellent video. I literally just learned about PWM at university, but this still taught me a thing or two.