Best DIY Fan controllers - PWM? Rheostat?

C

Citizen _insane

Gawd
Joined
May 2, 2006
Messages
565
I'm thinking of building a fan controller for my Rocketfish and was wondering what would be the best. I've got a decent amount of circuit building experience, but I don't know a whole lot of theory.

I've seen schematics for PWM controllers like here at Bit-Tech, but I don't know how well they work. Looks like it would be pretty cheap and also easy to fit 4 of them behind 1 5.25" bay cover. I would only be running 120mm fans (D12SL-12's) so I assume I'd have plenty of amperage overhead.

Lastly, does anyone know a good source for cheap knobs? They're super expensive almost eveywhere (which doesn't seem to make sense). I've seen em on digikey and mouser but most are $2-3 a piece or more if you go with aluminum.
 
I built a couple fan controllers several years ago.

I originally went with a PWM design, but I noticed that my Panaflo U1A fans made strange high pitched noises when running the PWM at excessively high frequencies. At lower frequencies, it was as though the fans were speeding up and slowing down. I think this had to do with the weight of the fan blade and inherent torquing when turning on and off.

I eventually went with a linear design using heatsinks. While it isn't as efficient in terms of power, it works for me. I think I used LM324 chips, or something close to that (I forget the model number). This eliminated the noise problem that I had with PWM chips.

Good luck!
 
Probably a LM317 design (variable voltage regulator). http://www.cpemma.co.uk/reg.html

I've done a PWM but had a "growling" coming from my fan that was more annoying than the full speed. I don't know if my frequency was too low or whether the frequency of the controller was producing an interference pattern since a brushless fan uses power in pulses rather than constantly.

For cheap knobs, I usually go with Mouser or Digi-key if I need consistency and good specs, and www.allelectronics.com if I'm feeling cheap. Solid aluminum are stupidly expensive though.
 
Does anyone know the name of the component that allows fans to attach to the motherboard? Those plastic L's with 3 pins? It'd be nice not to have to hack up all of my fans. I'd rather just order them from digikey w/ the rest of my components.

Oops, n/m I randomly ran into them on all electronics.

Fenris - the page you linked mentions the LM7805 and the LM713. Are they of the same family or do you recommend one over the other?

Wow, foot in my mouth again....the maximum amp draw on the fans are listed. Guess I know how large a HS to get now :)
 
I built a PWM fan controller a few years ago, and intentionally ran it at a frequency above 20kHz. (I used a simplified design that used a constant off time and varied the on time in order to adjust duty cycle) The advantage of PWM is that you can stick the electronics anywhere and not worry about heat buildup. Using an adjustable regulator like the LM317, you'll get a much simpler circuit, but you'll be limited in the number of fans you can run off it. In addition, you'll need to purchase heatsinks and also make sure that heat has somewhere to go.

Given the choice between the two methods, I'd choose the PWM circuit. It's a bit more on the design and soldering side, but you end up with a more compact circuit, lower power dissipation, and higher capacity.
 
Mohonri said:
I built a PWM fan controller a few years ago, and intentionally ran it at a frequency above 20kHz. (I used a simplified design that used a constant off time and varied the on time in order to adjust duty cycle) The advantage of PWM is that you can stick the electronics anywhere and not worry about heat buildup. Using an adjustable regulator like the LM317, you'll get a much simpler circuit, but you'll be limited in the number of fans you can run off it. In addition, you'll need to purchase heatsinks and also make sure that heat has somewhere to go.

Given the choice between the two methods, I'd choose the PWM circuit. It's a bit more on the design and soldering side, but you end up with a more compact circuit, lower power dissipation, and higher capacity.
Click to expand...

Do you recall what design you used or where you got it? I tried to do this since most humans cannot hear about 20kHz, but my fans flat out didn't turn. It may have been a limitation of my circuit design.
 
I'm having trouble finding any decent 1k ohm pots. Do they even exist in metal body form? I'd rather use the LM317 than the more expensive MIC29152BT (even if it does have higher performance). However, I'm not seeing any 1k ohm pots on Digi-Key.
 
Try searching Digikey for "Potentiometers", then choose "rotary-linear". Now filter by "In-stock", "bulk", "panel mount", "through-hole" and "1K". You should have 38 to choose from.
 
lol, I'm retarded. I filtered out the square ones b/c I thought they were trim-pots. Doh!

edit: none of those look any good. The only nice ones have shafts that are way too long (.75" wtf?). The 308NPC1K looks ok, but the shaft is on the long side and they're $6.76 a piece!!
 
I don't have a schematic handy, but here's how I did it:

Start with a 555 timer.
Connect the wiper of a 10kOhm pot to the discharge pin.
Connect a resistor (anything from 100 to 1000 ohms will do) between Vcc and one end of the pot.
Connect a diode between the other end of the pot and the Threshold pin, with the cathode towards the pot.
Connect a diode between the threshold and discharge pins, with the cathode toward the threshold pin.
Connect a 4.7nF capacitor between the threshold pin and ground.
Connect the output pin to the trigger pin
Pull the reset pin high with a 10kOhm resistor between Vcc and Reset.
Connect the output to the gate of an n-channel MOSFET that has a low drain-source on resistance, and tie the source of the MOSFET to ground.
Now hook up the fan with the red wire to +12V and the black wire to the drain on the MOSFET.

It's been a very long time since I've done it. Depending on what range you'd like, you can simplify the circuit, but this will give you something very close to 0-100% range.
 
I think they purposely make the shafts long so you can cut them to the desired length. It is easy enough to do with a dremel or hacksaw.
 
Yeah, I just added everything up on mouser and it's more expensive than buying one lol. Maybe I'll just wait until I buy a bulk order for building the motor controller for my CNC machine.
 
I just had an interesting thought--if you do anything with microcontrollers, you could build a PWM circuit very, very simply--take a cheap microcontroller with a few ADCs, attach the pots to the ADC pins, add a few lines of code to read the pins and output a PWM signal, and you're done!
 
Microcontrollers are fine, but it's like hitting a fly with a sledgehammer. A PWM is easy enough with a quad op-amp or 555 and a few passive components or a specific PWM controller chip.
 
Yeah, a microcontroller would be overkill, but think of how few parts you would have to use! Just the pots, matching capacitors to smooth the input, the uC, and the power transistors.
 
You must log in or register to reply here.
Back
Top