Fan voltage control via. potentiometer...

8

8Complex

Limp Gawd
Joined
Dec 15, 2003
Messages
272
Hey guys, I've read through the FAQ and came up with no conclusive answers. I've google searched around, however I don't think I was asking the right questions. I'm not too informed in way of electronics, so forgive me if my terminology is off. Also, forgive me if this is the wrong forum, but all the wiring and fan information was in the FAQ in here, not in the OC&Cooling forum -- mods, please move if you feel necessary.

I've got a fan I'm putting in my case as a blow-hole unit, and I want to be able to throttle it back (and hopefully that'll dim the LED's as well). Most everything I've read online about making controllers seems to use a far more complex circuit then I would think necessary.

Through my dealings with as few electronic projects as I've been involved with before, I know that potentiometers have the ability to cut and average the voltage depending on how you have it hooked up. Granted, I've only used it in very rough automotive applications (don't ask), but I think I've got the basic mechanical concept for them down.

I was thinking of two possibilities for wiring, though I don't know the imedence value I'd need for either.

(no connect)
   |
   |
POT------- Output to + fan
   |
   |
+12v

----OR----

+5v
   |
   |
POT------- Output to + fan
   |
   |
+12v


The way I figure, the first one will cut the voltage depending on how much resistance the pot will allow. I imagine a certain impedence (very low) will just cut the voltage enough to bring the fan output voltage to a decent level.

I believe the second method of hooking it up will allow me to adjust the voltage from +12v down to +5 volt -- though I'm not sure that the fan will run fine all the way down to +5v as I've been reading fans like a certain voltage range, usually fairly close to +12v.


Does anyone have any good suggestions, or ideas on what I've proposed above that'll help me out?


My main requirements on this be that the parts be available at Radio Shack, not mail order, and that it's kept simple as I don't have a whole ton of room in my case to do anything fancy like large circuit.
 
this is also the in FAQ, under rheo bus
 
Ahhh... what's a BayBus? Heh, and I thought that FAQ's were for people that didn't already know the terminology. ;)


According to this link from the FAQ, I need a rheostat, not a potentiometer. Though I'm still unclear why people would jump hoops to make a circuit using that LM317 chip and all when they could just use a rheostat... is there an advantage I'm missing?

It looks like Radio Shack has a 25ohm, 3watt rheostat that I can use that will more then likely be perfect. I don't want to be able to shut the fan down, just lower it's speed and maybe turn down the lights a bit.

Thanks all, hopefully soon I'll be posting pics of project half-assed. :)
 
pots/rheos generate more heat and are *much*less effecient than a voltage regulator. The advantage of a voltage regulator is that you don't have to buy an expensive, high power rated rheostat for the adjustment. 25ohm will be perfect, if you do decide to stick with that route, although some fans, especially tornados and larger 120mm fans draw ~6w, so a 3w probably won't be sufficient, depending on the fan.
 
Hmmmm... they do create more heat. Strange, I've even seen people glue a small heatsink on their LM317's when making those circuits, too. *shrug*

Not a big deal with the fan I've got... I've just an Antec 80mm LED fan (.14A draw)... I think I did the math (Watts = Volts x Amps) and it came out to 1.68w. Hopefully staying 55-60% of it's limit will help in heat issues as well.
 
Well, I got that rheostat in and it cuts RPM down by about 20%. I think I'm on the hunt for another higher impedence unit, though, as it's still pretty loud. I have to look into noise insulation methods as this blowhole is DAMN loud right on the sheet metal without a dampener.

If anyone has an idea where to find a standard sized rheostat (1/4" adjuster, maybe about 1" diameter body) that is 75ohm and 3-5watts, please let me know. I can't find a single place that sells ones that is around that size. :-/
 
None of those seemed quite right. I actually went through all the links for electronics components in the FAQ and came up empty-handed.

Oddly enough, I found the exact one I was looking for in a 2001 Radio Shack commercial customer catalog here at work, but when I called them on it, they said that they don't carry it and more and don't have anything like it any more. Doh! Fortunately I found out the company they were using was called 'Alpha' and a few things came up when googling for 'Alpha potentiometer'... I'll have to see if any of those will work ok.


Oddly enough, I seem to have issues starting my machine with the fan installed now. It reboots fairly randomly during load and doesn't stop until I turn it off, disconnect the fan, and restart it. I think in the end I will have to end up going with an actual fan controller. Kinda sucks since that means I'll have to put yet another fan in the case, and I'll have to probably disassemble everything to get to the spot in the chassis where I'll need to install it. ARGH!
 
... if mouser doesn't have it, it probably doesn't exist. I'm sure they have something.

Also, digikey is a good electronics provider
 
You could use the poteniometer to drive an operational amplifier which would power your fans, it would be relativly simple. Pick up a LM714 single op amp (or a triple op amp if you need more power). Connect 12 volts to Vcc, and the ground to ground. Connect the output from the potentiometer to the non-inverting terminal (+) and then connect the interting terminal (-) and the output together. A second wire attached to the output will now give you up to 4 watts of power per op amp at the voltage selected by the poteniometer, and a very low source internal resistance. This would be a $2 way to implement a fan control system, you could even have a triple op amp with three potentiometers, allowing you to control your fans seperatly. HTH

edit: i just looked it up for you and found that you can get a LM675 for less than $2 and it can supply 4 amps if you want only one poteniometer/amplifier combo. This would allow up to 48 watts in devices to be power cheaply by the potentiometer. This amplifier also can work in ambient temps up to 70 celcius.
 
Originally posted by CubedPC
You could use the poteniometer to drive an operational amplifier which would power your fans, it would be relativly simple. Pick up a LM714 single op amp (or a triple op amp if you need more power). Connect 12 volts to Vcc, and the ground to ground. Connect the output from the potentiometer to the non-inverting terminal (+) and then connect the interting terminal (-) and the output together. A second wire attached to the output will now give you up to 4 watts of power per op amp at the voltage selected by the poteniometer, and a very low source internal resistance. This would be a $2 way to implement a fan control system, you could even have a triple op amp with three potentiometers, allowing you to control your fans seperatly. HTH

edit: i just looked it up for you and found that you can get a LM675 for less than $2 and it can supply 4 amps if you want only one poteniometer/amplifier combo. This would allow up to 48 watts in devices to be power cheaply by the potentiometer. This amplifier also can work in ambient temps up to 70 celcius.
Click to expand...
Well, looking up an LM675 I found a datasheet here. I followed along with your hookup instructions and that is pretty easy (assuming -VEE is ground).

Using my 25ohm, 3 watt rheostat/potentiometer (which I'm gathering is actually one in the same by way of terms at this point), what type of range would I be able to get with this chip? I notice something on the datasheet stating that it's supply range is 16-60V, which I'm assuming will need a rather decent sized resistor to linearly cut the voltage to 3.2-12V (divide by 5 linearly).All that from a simple 12V input?

Also, would the input of the rheostat still need the +12V, or would it be different?

Sorry, I'm somewhat a newb when it comes to electronics, but I somehow fumble my way through. :) On a bright note, now that I've added some rubber washers to the installation of my fan, it's MUCH quieter. I still need to find a way to cut the resonance of the panel it is attached to, but I'm assuming a piece of wood and some superglue can take care of that one (both stiffen and weigh the panel down a bit).
 
The chip does have a 16-60 volt input range, which means that the Vcc+ has to be between 8 and 30 volts and Vcc- has to be between -8 and -30 volts. However, since you are going to have a DC input signal, you can have Vcc- be ground (0 volts) and it will work exactly the same, and be fine.
The point of the setup I described is that it does not require a large poteniometer. Since you already have one though, it would proably be easiest to just use it to power your fans directly without the operational aplifier.
The point of the op amp is that it not produce as much heat as a potentiometer does and you can use a small radio shack potentiometer that is about the size of a pencil eraser (although you can use yours). The potentiometer will set the voltage that the amp will output, but will have a negligable current through it. The op-amp will supply all of the current, and will always keep the output voltage exactly the same as the output voltage from the poteniometer (that is why the output is looped back to one of the input pins, it forces all inputs and outputs to be equal voltage)
You should set your potentiometer up going from 12 volts to 5 volts (most fans won't start below about 6 volts). The op amp will have a very similar output voltage (will be slightly lower). This large range is the big advantage of the amp, because it can supply 3 amps at 6 volts, which would require a 18 watt potentiometer to do, and it does it with only a few watts of power dissapated.
Hope that answered your questions and didn't make you more confused.
-Steve
 
Heh, it was a little more confusing, but I think I sort've get it. :)

So the potentiometer is just inputting the standard signal and without a draw, it can be adjusted from pin1 to pin3 (output as 2). Now if I put a +12 on one side and a +5 on the other, i can adjust along that range linearly. The opamp will output that same voltage, but it will do it without the worries of amperage (up to4 or 8 amps, if I remember from what you said).

So to be clear on the inputs... I can set +Vcc to +12 and -Vcc to ground and still be able to output up to +12v from the opamp? Or would it have the ~1.5v drop that most of the LM317 circuits have, maxxing at 10.5v+?

Oddly enough, you've reminded me that I can get these chips for free as company samples from certain places (it's good to work in engineering at times ;) )... so this mod would end up costing me nothing above what I've spent. :)
 
I haven't sat down and thought about all the pluses and minuses, of the different approaches, but here are a few.

The number one reason I like the voltage regulators: Short-circuit protection. They are tough critters to kill. There's thermal protection too. The main down side is the voltage drop, you won't get 12V out with 12V in. Typically 10.5 or so, maybe less.

The opamp is an option I hadn't considered, I dunno wha happens if you short the output. Probably let the smoke out.

Even cheaper if you don't mind giving up the short-circuit protection: just use a transistor as an emitter-follower.
http://www.cpemma.co.uk/ef.html
Several commercial fanbuses use this. People burn out the transistors regularly, so I know that sensitivity to shorting the output is a fairly realistic concern.
 
I just took a look at some of the graphs, and it looks like the internal output resistance is right around 4 ohms. Since an average fan draws about .25 amps at 12 volts, the average fan has about a 48 ohm resistance, the output voltage would only be aproximatly 12-.9*X where X is the number of fans being powered. This means that this circut deisign could only be used with about one fan at most :(
This circut would work if the desired voltage was less than the max voltage from the formula above, but doesn't look like as good of idea anymore. Since you have a potentiometer, you could just use that alone to power your fans, that will allow 12 volts for sure (just might not be able to reduce the speed as much).
 
Well, for right now I'm fairly happy with the speeds I'm able to adjust to/at... I can prop it up to 12+ for about 5500rpm, and then back it down (not sure on voltage) to about 4400rpm. It cuts the noise significantly.

The only real trick now will be removing the resonance buzzing/humming from the panel that it is attached to. I've already got rubber washers on either side of the panel spacing it out, but I think it's just the right amount of vibration to transmit it down the line. I'm going to try insulating the connections of the panel a little more, and if that doesn't work I'll try glueing or JB Welding some trusses onto the backside of the panel for some rigidity.
 
You must log in or register to reply here.
Back
Top