Temperature control fan

[baxev2005]

Hello everybody,
I want to maintain a constant temperature of 60 deg C to a heat dissipating surface, varying the speed of a fan. The surface is inside a box and the fan will blow outside air to the surface. A typical circuit will compare the voltage of thermistor on the hot surface to a preset and power the fan. My question is the following:

Is a typical pc class fan be able to provide continuously variable speed from 0 to max or do I need a dc motor?

Of course any suggestion on the whole project is welcomed.

Thanks in advance

 

[maxx]

You want to increase the speed of fan in automatic mode or you want to keep the fan speed constant?

 

[colin55]

Do you want to turn the fan on and off?

 

[baxev2005]

I want to keep constant temperature at the surface by regulating airflow and the speed of the fan, no on/off, except of course during starting when temperature is below set point and the fan is off.

 

[colin55]

What is the temperature differential. In other words what is the allowable temp rise before cooling needs to the employed.

What is the cooling effect of the fan when operating at full speed? In other words how much will the fan over-cool the item and in how long does this take?

In most cases the fan will over-cool the item and switch off. This the most cost-effective way to run the fan.

 

[Mikebits]

Simplest way would be to run the fan at all times, the fans in my computer do, is this a cost issue? If so, why? If a device generates so much heat as to need a fan, I am sure it is wasting much more heat than a BLDC fan.

 

[baxev2005]

Collin55:
The objective is to keep the temperature as stable as possible by varying the fan speed only. Any temperature fluctuation will be because of the delay in changing fan speed and moving hot air outside the box. Again NO on/off but continuous fan speed adjustment by closed loop.

Mikebits:
I was wandering if I can adjust the usual pc fan speed by varying supply voltage (nominal fan voltage is 5 & 12 DC Volts), since they are not just DC motors but sort of BLDCs. If someone has already experimented with it please post the findings.

 

[maxx]

You can obtain almost an constant temperature using an on/off circuit .
You can increase or decrease speed of fan varying supply voltage .

 

[colin55]

What sort of DC fan are you using?

Fan motors used in computers are brushless and rely on a hall-effect switch to deliver pulses the transistors and coils. They do not have ANY start-up torque and will not start-up with a lower voltage.

 

[crutschow]

Since you are talking about a closed-loop control proportional control system, it is likely to be unstable (it will overshoot and undershoot the desired temperature without every stabilizing to a constant temperature) unless you design in some loop stability compensation circuits, such as PID or Fuzzy Logic. PID can be done with op amp analog circuits or a microprocessor. Fuzzy Logic is particularly easy to impliment with a microprocessor and is more intuitive (it's basically a series of If-Then statements, i.e. If "temperature error is large" Then "change fan speed by large amount").

The complexity of the required compensation depends upon how accurate and speedy you need the correction to be, and how fast the power dissipation may change on the heat dissipating surface.

 

[colin55]

You have to do things in a logical way.
I get hundreds of customers coming in with absolutley no data at their fingertips.
We have to know what type of fan, what dissipation it will produce at full RPM and what dissipation at a lower RPM before we can determine if the fan can be reduced in RPM.
It may only need 10% on to 90% off or 90% ON.
You cannot do anything until you know the thermal conditions.

 

[baxev2005]

colin55: I wanted just that to investigate the possibility to use a plain pc brushless fan and vary its speed, if possible, otherwise revert to the dc motor fan. Of course the amount of air moved by the fan is determined by the fan speed and size. But all these parameters are to be determined experimentally, as the project is not commercial but an amateur construction.

crutschow: I try to keep the circuit as simple as possible. A fluctuation of few degrees is acceptable and of course can be adjusted with the time constant of the feedback loop, as in the good old time of analog design. Nothing fancy.

 

[schilkek]

There are some nice thermal management chips that might get you closer to what you want to do. Here's one such beast that I've seen referenced a lot.

**broken link removed**

I haven't played with it, but I do remember that there were a number of people who solved "problems" in their MIC502 circuits by adding bypassing and filter capacitors. There is a "reset" phase which runs the fan full-on for 64 clocks prior to start of the PWM phase; if transients aren't handled well, the controller never gets out of this reset and the fan appears to run continuously.

Hope that helps!

 

[crutschow]

crutschow: I try to keep the circuit as simple as possible. A fluctuation of few degrees is acceptable and of course can be adjusted with the time constant of the feedback loop, as in the good old time of analog design. Nothing fancy.

OK. Then you just need to slow the loop down enough so that the fluctuations are within your requirements.

 

[ccurtis]

Since all you want to do is experiment to get the desired results, you may be interested in this link.

**broken link removed**

Edit: Opps, sorry. You want a continuously variable type.

 

[Cobalt60]

What sort of DC fan are you using?

Fan motors used in computers are brushless and rely on a hall-effect switch to deliver pulses the transistors and coils. They do not have ANY start-up torque and will not start-up with a lower voltage.

I have a 12V computer fan I use at work for when Im soldering, or hot. I power it with my bench supply. It starts and operates at a little over 4V, and I rarely use it over 8V.

-Chris P

 

[Andrew Leigh]

Hi,

I have used this fan circuit for some time. It extracts warm air from the surrounds of a gas fridge enclosure. As the ambient temperature increase so the fan speed increases.

A diode is used as the sensor to control the speed of the fan, in your application you can attach the diode to the surface and use the pot to adjust for the temperature.

In my case the unit is set to come on at ± 25deg C and the fan will reach maximum rpm at approx 38deg C.

**broken link removed**

See if this circuit could work for you.

Cheers
Andrew