Resistor for Fan control?

Got hold of some 10w resistors, thinking of using them to lower the voltage to fans, but I'm not sure about the values yet.
To cover the most likely area I bought 15,120 and 1K Ohm. Think the 120 is the most suitable. Anyone have some ideas? What values to get..

---
Added a pic of the res. alonge a ordinary very familiar thing to show scale.

The problem with fixed value resistors is as you suspect, they are but one value. How well a fixed value resistor serves to regulate a fans speed depends upon how much current that fan draws, the rpm it spins at without any voltage dropping resistor and how fast you want it to spin. Since most manufacturers do not list the RPM for any voltage other than full voltage, it's almost impossible to guess what value you need. The more current the fan draws, the less resistance you need to impact its speed. A resistor that is adequate for a fan that draws 100ma will be too high when used with a fan that draws 750ma and vice versa. That is why high wattage variable resistors (rheostats) are so popular. You can adjust the resistance to fit the fan and your speed preference. Now, if you are deploying several of the same fans, you can use a rheostat to determine the amount of resistance needed to hit the speed you like and then substitute fixed resistors in place of it. Generally, a 35 or 50 ohm rheostat will serve for a wide range of fans.

For a while last year, I had a line on 35 ohm 12W rheostats and bought all I could get for members from a local electronics surplus store for about 1/5 their normal going price. I bet I shipped out over a hundred of them.

Most larger (>50k people) metropolitan areas have at least one or more electronics surplus stores in them. Check the Yellow PAges or ask the local Ham radio operator, they always know where those kind of stores are. Perhaps the Electronics Shop teacher at school. Otherwise, there are plenty on the Web.

Hoot

Before I got a high wattage Rheostat. I was using high wattage resistors connected to a switch with a 9 Watt Tornado fan. The switch I used was a single pole double throw, with center off. I could change the fan from full 12 volts, to around 9 volts and 7 volts. Slow , Medium or High.

I don't know the resistance values you need for the fan you have, but I used 2 - 10 Ohm resistors. When the switch is in the bottom postion you get the two resistors in parallel or 5 Ohms. So the Values would be 0 Ohms, 10 Ohms and 5 Ohms with the switch. I was getting around 12 Volts, 7 volts and 9 volts with these resistances.

Here's a picture sorry I did it very quick so its messy.


Also here is a link for a very cheap high wattage Rheostat that works great with Tornado's. I also think it will work with other fans as well.

http://www.allelectronics.com/cgi-bin/category.cgi?category=470&item=RHE-15&type=store

Here is the same exact Rheostat at another web site, but they want $12 instead of $1.50.

http://www.frozencpu.com/cgi-bin/frozencpu/ele-50.html?id=huYdLMvX

Keep in mind if the voltage for the fan is set too low at system start the fan won't start. Tornado's need around 6 or 7 Volts at the very least to start.

I got the resistors from a electonics-repairshop.
I was actually goning to my a variable resistor, but they just had 1/2w rheostats.
----
My plan is to quiet down a old AT-psu-fan, guess the chances are best for success with the 15Ohm resistor...
----
To keep costs low I always try to buy local.
For major purchases I use "www.rselectronics.com", they have a shop in norway too. I have their whole catalogue on cdrom.
---
Thanx for the input guys. I can't start modding today... Merry Christmas!

An alternative is to use diodes. Cheap, simple to implement and use.

Each diode drops 0.7v when forward-biased, so put 4 or 5 in series to drop the voltage to the fan. The el-cheapo types (1N4001) is like 2p in the UK and can handle 1A.

Lets see if I got this correct... You put a series of diodes in serial with the +12v lead... Correct?

Yup - 12v - diode - diode (etc) -diode - fan - 0v

DaWiper said:

Lets see if I got this correct... You put a series of diodes in serial with the +12v lead... Correct?

Click to expand...

Yes he meant in series, which is a pretty good idea. You could also use diodes in place of the resistors in the circuit I posted. Having diodes in series and in parallel.

I get it. But why use a diodes in parallel?

Beats me. You will still get a 0.7v drop, unlike resistors where you will get 1/2 the voltage drop if you put 2 of the same in parallel (assuming that the fan current doesn't change, and it will, hence the really complicated mess of maths.)

I get the serial stuff, but why use one in parallel?

If you wire up 4 diodes in series to the center connection of the switch like in the picture I posted below. When the switch is in the center postion that will give you around 9.2 volts for the fan.
Now take another 3 diodes and wire them in series on the bottom connection. When the switch is in the bottom postion the 4 Diodes and the 3 Diodes are now in parallel which will give you around 7.1 Volts.
So the switch will be 12 Volts, 9.2 volts, and 7.1 volts.

The switch has to be a three position toggle switch. With middle connection being off, when toggle is in middle postion.

OOOooo! Diodes in parallel with each other... I thought that you meant diodes in parallel with the fan: Diode shorting + and -.... Ok I get you now!

I have used the diode approach when I didn't have the right value fixed resistor or the right wattage, which brings up a good point. Make sure the diodes or rectifiers if you choose, are rated for the right current. Small switching diodes, though inexpensive, may not be able to take the amount of current your fan draws. For example, a 1N914 is a common and inexpensive switching diode, but it can only tolerate 200ma continuous forward current. If you were using some of them to drop the voltage to a fan that takes 500ma (1/2 amp), the weakest in the series string would burn out. I typically try to use a rectifier (hefty diode) that is rated for twice the current that my fan needs. If your fan needs 500ma (1/2 amp), then use a 1000ma (1 amp) rectifier. You can get 1N4004, 4005, 4006or 4007 rectifiers for about $.20 each from just about any component store. They take 1 Amp continuous current. The difference between the 4,5,6 or 7 digit represents how high a forward voltage they tolerate and any of them will work just fine at 12V.

If you already have some diodes, but they are not rated for the current you need, connecting them in parallel will increase the current capability. Say you have some 100ma diodes and your fan needs 400ma. Just parallel 8 of them (remember twice the current) to get a capacity of 800ma. You should try to use all the same model diode, or even better the same lot, so that the current is fairly evenly divided among them. The voltage drop across mass produced diodes and rectifiers varies slightly from one to the next. If you parallel four of them and they are not fairly matched, the one with the lowest forward voltage drop will try to take the most current and possible burn out. That's why I try to avoid having to parallel them, favoring instead just getting those where a single one has enough current capacity to handle the fans needs. Less opportunity for failure.

Like resistors, diodes that are used to derive a voltage drop will get warm, even hot. that's where the dropped voltage goes. It is dissipated in the form of heat.

Hoot

If I find this mod defendable in terms of security I will probably end up with a fixed resistor at 10W or more to be sure that burnout doesn't happen.
The whole project is about making a p233 w/48MB ram into a gateway/firewall with a minimum of noise. I've already removed the cpu fan, leaving the cpu with a duron-class heatsink. It sure gets hot...
This machine will be placed in our livingroom almost always switched on...
Found a linux distro(dachstein) that runs straight off cd or floppy...

It's amazing how quiet a pc can be....