Advanced DC fan controller [closed]

Question

I am wondering if I could get some advice for my final year project. Last year, for my 'mini project' I made a DC fan controller that activated the fan at a specified temperature, to cool it back down to the set temperature. The circuit is shown in the link below.

I have about eight months to make my new project. I was pondering on what I could add on to this to make it more advanced for final year standards; but I couldn't really think of anything too great. If you guys have any opinions or advice it would be greatly appreciated.

Answer 1

Several years ago I tried to make a fan controller that would allow two fans to match RPM's. The goal was to avoid having two fans at a slightly different speed, which would cause an audible effect called beating. This effect is commonly heard with dual-engine turboprop planes. When the fans/engines are close in RPM you hear a whah-whah-whah sound at the difference in RPS (revs-per-second). For example, if one fan is spinning at 1000 RPS while the other fan is at 1001 RPS then you will hear a whah-whah-whah sound at 1 Hz.

My circuit had 2 fans, both with tachometer outputs. One fan was controlled with something similar to what you have now. The second fan was slaved to the first. This was done by using a circuit very similar to a Phase-Locked-Loop (PLL). Here's a simplified diagram of a PLL:

Ref Clock is the tachometer signal from the first fan. The Voltage Controlled Oscillator (VCO) is actually the second fan, with the tach signal going to the divider. In your case, the divider can be bypassed (divide by 1). You'll have to do some research into Phase-Frequency-Detectors and Loop Filters, but you could probably use a simple XOR gate as the PFD.

The basic idea is this: The output of the loop filter IS the voltage used to drive the fan. Because the fan speeds up as the power voltage goes higher, it works as a voltage controlled oscillator. If you get the PFD and loop filter correct then the second fan will be the same speed as the first fan.

My circuit mostly worked, but I ran out of time to tweak it to work perfectly. The two fans were very close in speed, but not perfectly matched. The fans I was using would have some speed variations that the circuit couldn't adapt to quick enough. I wish I had time to finish it, but the early results were promising.

Answer 2

Well what you're asking is a bit complicated....

Maybe make a system that will use a microcontroller to control a large number of fans based on data it receives from a large number of temperature sensors and some complex (maybe user-defined) algorithm? Say ventilation system for several rooms? Maybe add a minimal airflow requirement for a specific room? To further complicate things, you could add output say to a TV or a computer monitor, make it controllable over Internet or via a mobile phone.

Answer 3

Consider the problem worthy a good patent: acoustic noise from fans is a plague. One component of noise is fluctuation of torgue caused by cogging and unevenness of air density. To beat the problem in scientific way, think of maintaining constant torque instead of constant velocity. Which is replacing voltage regulator with current regulator with bandwidth much higher than normal curcuit with a cap.

Prooving the solution must also involve some test setup with microphones, audio and spectrum analysis. And again. This problem was never solved. In the whole human history you will be first.

Another idea, but already published, is making a subwoofer using fan. This might require a very advanced motor (say 3 phase DC with PWM controller). If budget allows, or some clever trick will let a cheap fan to be used. It may drive a lot of interest of hobbyist community.