Fan speed controller

Control: management of fans, temp/rpm monitoring via soft/hardware

Moderators: NeilBlanchard, Ralf Hutter, sthayashi, Lawrence Lee

Post Reply
NotSilentYet
Posts: 19
Joined: Sat Jan 11, 2003 9:27 am
Location: Lund, Sweden

Fan speed controller

Post by NotSilentYet » Sat Jan 11, 2003 11:52 am

Hi,

I am new to these forums so its possible this has been posted before. I just wanted to share the way that I control the speed of my fans with the simple circuit posted below. I use this instead of a rheostat because i found them too expensive and I had all the components for this circuit lying around. Instead of a 2n6292 you could use basically any npn power transistor such as the mje340 and tip31.

Image

The optional resistor sets the lowest fan voltage. With a 10 k resistor this equals 4.7 V with my 92 mm papst fan. If you leave the resistor out it will be adjustable to 0 V. The downside to this circuit is that the lowest voltage is load dependant. So if you attach a larger fan the lowest voltage will go down.

Henrik

powergyoza
Patron of SPCR
Posts: 543
Joined: Mon Oct 21, 2002 1:01 am
Location: Vancouver, BC, Canada
Contact:

Post by powergyoza » Sat Jan 11, 2003 1:12 pm

Wow.

Can R2 be substituted with a thermister? Or better yet, can a 10k thermister be placed inline with R2 to make a user-adjustable thermal control circuit?

clamrade
Posts: 22
Joined: Mon Jan 13, 2003 8:34 am
Location: MN

resistor?

Post by clamrade » Mon Jan 13, 2003 9:49 am

Would a simple potentiometer do the same thing?

NotSilentYet
Posts: 19
Joined: Sat Jan 11, 2003 9:27 am
Location: Lund, Sweden

Post by NotSilentYet » Mon Jan 13, 2003 3:12 pm

Yes it would. However it must be rated at a higher wattage than a normal potentiometer. These are often referred to as rheostats and tend to be rather expensive. That's why I use the above circuit.

clamrade
Posts: 22
Joined: Mon Jan 13, 2003 8:34 am
Location: MN

not sure yet

Post by clamrade » Mon Jan 13, 2003 8:03 pm

I see where you are coming from. But an L1A panaflow draws only 68mA. Thus, one would need at most a cheap 1/4 watt potentiometer.

Am I miscalculating? I only ask because I was going to do this, and I would like a sanity check. Thanks.

TerryW
Posts: 85
Joined: Sun Aug 11, 2002 3:26 pm
Location: Coquitlam, B.C. Canada

Post by TerryW » Mon Jan 13, 2003 9:58 pm

Hi all.

I feel compelled to reply to these messages. My electronics is a little rusty, so please bear with me. :wink:

To NotSilentYet: Your circuit is an emitter-follower design. The base-emitter junction voltage of your transistor is around 0.6VDC at saturation. With your load (fan) in the emitter lead, the combined current through the tansistor could cause the base-emitter voltage to increase past 0.6VDC and damage the transistor. Hence no fan operation.

A slight circuit redesign should be to relocate the load (fan) to the collector lead of the transistor, a series base resistor to limit the base-emitter junction current (voltage) and tie the emitter to gnd.

To Clamrade: A rheostat is better suited for direct control of a fan, and a 1/4-watt potentiometer will burn up if the fan draws 68mA. Simple ohms law tells us to calculate wattage, we need to multiply the voltage by the current. I.E. 12V X 0.068A (68mA) = 0.816-watts. I would recomend using at least a 2-watt-rated potentiometer or rheostat.

Please don't get me wrong. If you are a techy & handy with a soldering iron, build up a transistorized fan controller. It may give you the results you require and provide you with a circuit that you built yourself.

If you don't have the electronics wherewithall, maybe use a rheostat.

Good luck :!:

TerryW 8)

Marvin
Posts: 64
Joined: Mon Nov 18, 2002 11:54 pm
Location: Tampere, Finland

Post by Marvin » Tue Jan 14, 2003 2:11 am

I made this myself. It was easy, components easy to find and rather cheap. And this one doesn't change its setting according to load, all it needs is that the PNP can handle enough current.

I myself added this to the circuit, and now it looks pretty nice in front of my 'puter, even if I say in myself :D, not just two knobs that handle the fans.

clamrade
Posts: 22
Joined: Mon Jan 13, 2003 8:34 am
Location: MN

not quite

Post by clamrade » Tue Jan 14, 2003 6:27 am

TerryW,

Not quite right. The fan draws 68mA at 12V. The pot dissipates far less. The pot at 10ohm, would drop at most 50mW and above 100ohm, would drop at most 200mW. I'll let you know if it doesn't work out.

NotSilentYet
Posts: 19
Joined: Sat Jan 11, 2003 9:27 am
Location: Lund, Sweden

Post by NotSilentYet » Tue Jan 14, 2003 11:01 am

Clamrade: You are correct. Driving a small fan like the L1A will probably not be a problem with a 1/4 W pot. I'm using it to control slightly larger fans.

TerryW: I can't see that happening. The fan voltage will always follow the base voltage minus the Vbe drop unless the base current becomes excessive. The transistor I'm using has a current gain of at least 30 at 2A. So I think everything should be safe up to about a 2 amp load. At 2 amps the base current will be up to about 66 mA which everything in the circuit should handle.

Marvin: The circuit you posted uses a Sziklai pair instead of a single transistor. This is a good thing if you want to drive larger loads and as you say the minimum voltage doesn't change depending on the load. I just felt it added unnecessary complexity in my case.

macday
Posts: 2
Joined: Sun Aug 11, 2002 3:26 pm
Location: Florida, USA

Post by macday » Tue Jan 14, 2003 7:40 pm

I have been using potentiometers (or rheostats) for controlling fans in my case for nearly two years. The biggest challenge I encountered was finding a source for these potentiometers (pots) that would handle at least 5 watts without breaking the bank.

One of the best sources I have found is Digi-key at www.digi-key.com (1-800-344-4539). They have 5W potentiometers available in a variety of resistance loads. I have been using the 500 ohm pots (Digi-Key part # CT2156-ND) which sell for $3.61 each. These pots are perfect for attaching to a 5 1/4 face blank or to the back of the computer. You can fit up to 6 of these on a face blank (if you have that many fans in your box) and wire them to your fans.

The major advantage to a pot over a transistorized fan control is it is immediately adjustable to support load, temperature changes, and different fans. Using the transistorized approach I found if I changed the unit from a 60mm or a 80mm fan to a 92mm my fan speed (noise) changed. With the ability to adjust the voltage on the fly I don't have these issues.

The 5W unit will easily handle the loads of most of the fans (or in some cases multiple fans) we are looking to quiet down. However, if you feel you need more power handling, or want to use it on one of the 7000 RPM 92MM Tornado fans then I would recommend the D25K500-ND or a D50K500-ND. These are 25w and 50w potentiometers with a 500 Ohm resistance that sell for $7.60 and $10.87 respectively.

Good Luck

Jim :lol:

Marvin
Posts: 64
Joined: Mon Nov 18, 2002 11:54 pm
Location: Tampere, Finland

Post by Marvin » Tue Jan 14, 2003 11:09 pm

NSY:

True, it is complex, having total amount of four components :lol: .

Still I would like the idea that someday I might change the fans attached to it, at the same time changing the load. That one extra transistor can be the cheapest one you can find, because currents it has to handle are so low that I think any will do.

I don't remember anymore, whitch NPN I used, but the total cost for my circuit was ~2EUR (2$), of whitch the NPN costed 0.3EUR. And I got most of the small components for free because I purchased some other stuff at the same time.

NotSilentYet
Posts: 19
Joined: Sat Jan 11, 2003 9:27 am
Location: Lund, Sweden

Post by NotSilentYet » Thu Jan 16, 2003 10:41 am

Marvin wrote:NSY:

True, it is complex, having total amount of four components :lol: .
Hehe. Although I said unnecessary complexity I didn't say it was complex. But I get your point. The reason is also that I was able to do the PCB extremely small by not adding the extra transistor. Money is not an object, the small transistor costs 0.03 EUR because I usually buy 30 or so at a time and the power transistor I got for free from onsemi. It might be good to know that most manufacturers have some sort of free sample program where you can order what you want and get it delivered by UPS for free. Examples of manufacturers that give away free samples are:

http://www.ti.com
http://www.maxim-ic.com
http://www.onsemi.com
http://www.analog.com

There is nothing like free electronic components :)

Marvin
Posts: 64
Joined: Mon Nov 18, 2002 11:54 pm
Location: Tampere, Finland

Post by Marvin » Fri Jan 17, 2003 3:11 am

NSY:

I'll have to check on those free components next time I'm building some project.

I do agree to your point, the extra component indeed increases the PCB size. In my case it was a matter that did not concern me because my PCB was going to big anyway, so one extra component didnt hurt that much.

But if one has only three components, one more truly makes a difference.

clamrade
Posts: 22
Joined: Mon Jan 13, 2003 8:34 am
Location: MN

zero to R

Post by clamrade » Wed Jan 22, 2003 12:00 pm

Most pots (if not all) are adjustable between zero ohms and the stated maximum resistance. So, a 500 ohm pot goes from zero to 500.

The minimum voltage on the fan would depend on the maximum resistance of the pot in series.

If you are actually building the above circuit, it more-or-less works the same. If you use a series resistor, then you set a minimum voltage (like the original poster stated).

Post Reply