Thermistor substituting for thermal fan control circuit

[370forlife]

http://www.virtual-hideout.net/guide...an/index.shtml

I want to make that little circuit for my ATNG AP-500X that has no temp control on the fan. I've got everything but I am having a hard time finding a 470 ohm thermistor.

The closest I can find that I have laying around (I want to just use parts I have as much as possible) is a 50 ohm thermistor out of a sunpro.

I assume I will have to play with the value of the other resistor. For a 470ohm thermistor, I will also need a 2.2K resistor. On the data sheet for the LM317T it shows a 220 ohm resistor in place of the thermistor and a 5K resistor in place of the 2.2K. So I was thinking when using a 50ohm thermistor I may want to use something like a 10K resistor. Am I on the right track, or completely off?

[PlainBill]

Re: Thermistor substituting for thermal fan control circuit

http://www.virtual-hideout.net/guide...an/index.shtml

I want to make that little circuit for my ATNG AP-500X that has no temp control on the fan. I've got everything but I am having a hard time finding a 470 ohm thermistor.

The closest I can find that I have laying around (I want to just use parts I have as much as possible) is a 50 ohm thermistor out of a sunpro.

I assume I will have to play with the value of the other resistor. For a 470ohm thermistor, I will also need a 2.2K resistor. On the data sheet for the LM317T it shows a 220 ohm resistor in place of the thermistor and a 5K resistor in place of the 2.2K. So I was thinking when using a 50ohm thermistor I may want to use something like a 10K resistor. Am I on the right track, or completely off?

Completely off. In the original design, the 2.2K resistor and the thermistor form a voltage divider. You want the ratio of resistances to remain the same.

So if you drop the resistance of the thermistor from 470 ohm to 50 ohm, you must drop the value of the series resistor by the same fraction. Algebraically, 470/2200 = 50/ X . Or for those who skipped algebra and took band, you will need a 220 ohm resistor.

NOW you may have a problem. This little gadget is powered from 12 volts. The full 12 volts is dropped across the series circuit of the resistor and thermistor. In the original design, you have 470 + 2200 ohms; call it 2700 ohms. E=I X R or I = E/R, so the original circuit draws 12/2700 or 4.4 ma. There is roughly 2 volts dropped across the thermistor at room temperature. doing the math, that gives a power dissipation of 9 mW. However, your redesign increases the current by a factor of 10. So the thermistor will be dissipating nearly .1 watt. What's more, the 220 ohm resistor will be dissipating about 1/2 watt. Not something noticeable on your electric bill, but enough to increase the temperature of the thermistor AND require a 1 watt or larger 220 ohm resistor.

The net effect will probably be that once the temperature in the case starts to increase, the fan will ramp up to full speed and stay there, even when the case cools down.

By the way, Digikey has 470 ohm NTC thermistors in stock for $.21. You can even do this as a surface mount project.

PlainBill

[370forlife]

Re: Thermistor substituting for thermal fan control circuit

Problem with the surface mount is I want to be able to put the thermistor on two extension wires and be able to put it somewhere where it gets hottest.

It would be do-able, but a little tricky. Though, I don't really feel like paying $7 shipping on a $0.21 part.

The other thermistor I have is a 5K, the other is a 100K.

5K would need about a 22K resistor.

[PlainBill]

Re: Thermistor substituting for thermal fan control circuit

Problem with the surface mount is I want to be able to put the thermistor on two extension wires and be able to put it somewhere where it gets hottest.

It would be do-able, but a little tricky. Though, I don't really feel like paying $7 shipping on a $0.21 part.

The other thermistor I have is a 5K, the other is a 100K.

5K would need about a 22K resistor.

Actually, shipping would be about $2.00. Digikey will ship by First Class Mail, but they usually use a box.

HOWEVER, check the specs on the LM317T. You aren't looking for a precision voltage source here. The 5K should work, just bypass the input and output with caps. A .01 - .1 uF cap at the voltage regulator between Vout and Adj pins would be a good idea to filter out any noise picked up by the thermistor.

PlainBill

[c_hegge]

Re: Thermistor substituting for thermal fan control circuit

https://www.badcaps.net/forum/showth...rature+control

I quite like these as they allow you to set the target temperature.

[Scenic]

Re: Thermistor substituting for thermal fan control circuit

Zalman uses a similar circuit in the FanMate II ..
but using a 7805 v-reg and a potentiometer..

[370forlife]

Re: Thermistor substituting for thermal fan control circuit

https://www.badcaps.net/forum/showth...rature+control

I quite like these as they allow you to set the target temperature.

I was considering that one and I had all the parts at hand (except the 10K thermistor,) but I don't want to be fiddling with it trying to set the temp while a psu is running with the cover off.

HOWEVER, check the specs on the LM317T. You aren't looking for a precision voltage source here. The 5K should work, just bypass the input and output with caps. A .01 - .1 uF cap at the voltage regulator between Vout and Adj pins would be a good idea to filter out any noise picked up by the thermistor. PlainBill

Thanks for the help! The caps shouldn't be a problem, though I may have to pick up a 22K at ratshack. I do need a breadboard for this too, so might as well.

[PlainBill]

Re: Thermistor substituting for thermal fan control circuit

Zalman uses a similar circuit in the FanMate II ..
but using a 7805 v-reg and a potentiometer..

First mistake. It's NOT a 5 volt regulator. The LM317T is an adjustable output voltage regulator. The output voltage will be 1.25 volts above the adjust input.

You do realize that effectively they are the same circuit, don't you? In each case the semiconductor is being used as a voltage follower. The power FET will achieve a voltage closer to 12 volts, but lacks the thermal protection and current limiting of the voltage regulator.

PlainBill