• Welcome to Overclockers Forums! Join us to reply in threads, receive reduced ads, and to customize your site experience!

Making high power fan controllers - guide

Overclockers is supported by our readers. When you click a link to make a purchase, we may earn a commission. Learn More.
Another update:
- I added some info about 2-transistor circuit, for better understanding and tweaking possibilities
- new section - eye candies. Now controller can have not only performance, but also look. There are now instruction on illumination, and indicating fan voltage in both digital and analog ways. I think that for example 2-5.25" bay, 4-rail kickstart controller with analogue volt-o-meters on each line with white LED lighting would definietly look cool. Unfortunatly, things that make it look good, cost usually more than parts for voltage regulation.

//edit//
Not to give update to update, I can say that there are some changes to dual transistor controller. All credits go to Shroomer, who has made that controller and found that load on transistor Q1 is close to none. I checked simulation again, and I found that measurements were done incorrectly, showing too high load. A result - you can use low-power transistor for Q1, and no heatsink with it. Cost and complication are decreased. Q2 remains high-power with heatsink.
 
Last edited:
Update:
Have you ever used Speedfan to control fan speeds? Did you like it?
But there was a problem - fan had to be low-power to connect it to mobo. Times changed - now you can use any fan you want with latest circuit including Tornados and Deltas - check it out.

A small warning - circuit hasn't been tested (mostly because of lack of time and my mobility problems). If you test it - please send me PM with report.
 
Got a tornado? Wanna run 5v, or 12v? Got an old **JUNK** psu laying around with a 115/220 switch on the back?

Take out the 115/220 switch, an old molex adaptor (like the ones that come with a DFI infinity-- female->Sata) with a female, and wire the 12v and 5v into opposite ends of teh switch. ground in the middle ofcourse, and then wire the tornado to the top sets. And there ya go...

A simple 12v/5v switch for a tornado :D

**disclaimer** make sure you know what your doing if you open a PSU. Im not responsible for anything you do because your dumb.

Incase you got lost reading that, I made a picture. works wonderfully :D

125vswitch5wz.jpg
 
Last edited:
It shouldn't hurt PSU - I used to do such things with some fans.
However I cannot recommend this method in any way. In my opinion it lacks flexibility too much. You can easily make 0/5/7/12v "controler" using 2 switches, but basically you get 3 useful sets, no RPM readout at 7v and possibility to turn fan off accidently. Complication of transistor controller and switches-based one is similiar enough, to recommend that first.
 
Ven0m said:
It shouldn't hurt PSU - I used to do such things with some fans.
However I cannot recommend this method in any way. In my opinion it lacks flexibility too much. Complication of transistor controller and switches-based one is similiar enough, to recommend that first.

I run either 5v (24/7 use) or the occasional 12v for full speed benching. I dont really wanyt/need/have a use for a 7v, so this works very well :)

Only problem is getting the solders on.. otherwise, its working well right now. :cool:

(was looking for a FREE/easy mod, for fan controller. :D )
 
nice guide ... good to see i am not the only one tackling this sort of thing. my goal is to controll 9 fans in unison. i decided to go with PWM and FETs (and a cap for stabilization). the primary reason for PWM is that i can controll it digitally. there are no pots or the like, it is totally digital.

just my $0.02.
 
Please post either datails of your tested circuit here or post a link to your own post. I know that I'm definately interrested in PWM. Hopefully Ven0m will have no objections.
 
Hehe, sure :D
PWM without stabilization is just something that completely doesn't work for me - that adds more noise than fan at full speed, at least with fans I own. It works for some people, so it can't be that bad. Basically I've seen no reason to post PWM circuits. Perhaps there should be some update.

Also - you can see "Speedfan controller" - it's just a circuit that uses "fan out" characteristics, so it's PWM and digitally controlled.

There could be also something with USB control, but I don't think I have enough skills and knowledge to make it. If someone could help - that would be great.

femtoghoti - if you don't mind, I'll add your scheme to guide.
 
my design was a 'proof of concept' design that was a couple months ago, so i will not have exact values or part numbers. the goal was to use PWM to controll a FET so i can vary the average voltage to a DC fan. the fan was a 92mm Panaflow.

in the end what worked was hooking the drain up to the PWM, emmitter to +12v and collector -> cap -> fan -> ground. it turned out i needed the cap to clean up the voltage because the Panaflow didn't like not having a 'clean' source.

the tricky thing with PWM as i found out is that not all fans "like" it. so i think it isn't as straight forward as straight DC, but simpler then rocket science. i guess it depends on where you want to take the guide.

i will generate a scheme this weekend.
 
Oh thanks. I know general PWM stuff, I even have some schemes and understand them. The most interesting thing was how do you control the fan - to be clear, your method to generate pulses.
The most common, and one I used for my schemes was 555 timer with some pot to adjust high/low ratios. I'm even more interested with your scheme, as you mentione no pot to use.
 
part of my design involves having it operate independantly of my CPU. as a result i will be programing a FPGA. this FPGA will then set my duty cycle to my FET's. this will afford me to be able to controll many fan independantly or together.

in the bigger picture there is going to be a PIC as the brain of all this that will be controlling the show.
 
I have no slightest idea what FPGA and PIC are, but if you described whole thing, it would be great.
 
dippy_skoodlez said:
PIC- programmable interrupt controller?
That was my first guess, but then I thought that "it can't be" - that seems to be too "high-tech" for a controller. Or maybe just electronics I know is from stone age?
 
sorry for being obscure .. here they are

PIC - i believe it does stand for Programmable Interupt Controller, but only because they started out as controllers for disk drives. long story short it is a microcontroller. here is where you can find the data sheet for the one i am using. http://ww1.microchip.com/downloads/en/DeviceDoc/39026c.pdf

FPGA - Field Programmable Gate Array. this is a programmable microchip. basically you lay out a scheme with all your gates (and, or, xor, ect) and flip-flops and the "program" the chip. they can do all kinds of fun stuff.

the FPGA affords me additional I/O so i can use smaller FET's to controll multiple fans. this also will lead to the possibility of operating independantly, should i decide. each fan will have its own channel and own FET, and i can do with it as i see fit :) the FET's can handle more power, but i am trading that for flexibility.

controlling the fans is just a piece of my current project ... it is the rest of the project that the PIC handy for. as mentioned this could be done with 555 or a FPGA.
 
Last edited:
femtoghoti said:
sorry for being obscure .. here they are

PIC - i believe it does stand for Programmable Interupt Controller, but only because they started out as controllers for disk drives. long story short it is a microcontroller. here is where you can find the data sheet for the one i am using. http://ww1.microchip.com/downloads/en/DeviceDoc/39026c.pdf

FPGA - Field Programmable Gate Array. this is a programmable microchip. basically you lay out a scheme with all your gates (and, or, xor, ect) and flip-flops and the "program" the chip. they can do all kinds of fun stuff.

the FPGA affords me additional I/O so i can use smaller FET's to controll multiple fans. this also will lead to the possibility of operating independantly, should i decide. each fan will have its own channel and own FET, and i can do with it as i see fit :) the FET's can handle more power, but i am trading that for flexibility.

controlling the fans is just a piece of my current project ... it is the rest of the project that the PIC handy for. as mentioned this could be done with 555 or a FPGA.

Interesting - How are you getting the duty cycle out of the FPGA? Can we get some circuit diagrams and pseudocode out of you ;)

I was planning a parallel-port driven set of registers feeding a poor-man's-DAC for controlling 7 fans or so, but university work and the lack of soldering irons has rather halted progress on that :-/
 
Back