Electronics Question

[mumblecrunch]

I'm building a stirplate for yeast starters and I cannibalized a robust fan out of a Dell server to base it on. I also got a couple cheap PWM motor controller units off of ebay after reading that people have had trouble dialing in speeds in the right range with potentiometers.

One of the controllers is pretty much the same as this one while the other is more like this one.

They work pretty much identically when hooked up to the same power supply (9V and 12V adapters with appropriate current ratings). But they both exhibit the same behavior: they basically go from 0-100 twice across the travel of the knob from full off to full on. So you start turning and it gets faster, then about halfway through it starts back again at zero. Any ideas why they would do this? It's not that it's really a problem or anything, I'd just like to understand the reason for the behavior.

[Jayme]

Makes no sense to me! Would have to probe it with an oscilloscope to figure out what's going on there.

[AllanMar]

is it a 4 wire fan?

[mumblecrunch]

is it a 4 wire fan?

It is 4-wire. It's a Nidec M35172-35, which is one of many variants that go under the TA350DC model number.

Here's a picture of the fan (which is pretty much this one):

IMG_3957.JPG

I wired it up again tonight. First, I just plugged the fan power leads (red and black wires) into the Motor+ and Motor- slots on the PWM controller. Then, taking into account that it's 4-wire, so there's a PWM control lead, I hooked it up like the picture below, with the red lead sharing the Power+ slot with the feed from the adapter, the black in the Motor- slot, and the blue in the Motor+ slot (that means the yellow wire must be the one-wire tachometer):

IMG_3956.JPG

The behavior is not as stark as I originally posted. The fan starts up and then around halfway through the knob travel there's a reduction in power which builds back up and then again at about 3/4 of the way through the knob's travel it does the same thing again. It was similar (but maybe not identical) when I wasn't using the PWM control line as when I was.

In digging up information about the fan it supports thermal control. Not sure how that works, but it almost sounds as if there are some smarts built into the fan circuitry. Could there be three "phases" to it's operation based on the incoming pulse widths?? Given that both PWM controllers I have exhibit the same behavior, I tend to think it's not the controllers. Also, there's a surface mount LED on the PWM controller board that just gets steadily brighter, which suggests that as far as it's concerned, it's just steadily changing the pulses.

I'm a little worried this fan might spin too fast at minimum power to work effectively as a stir plate driver (of course I won't know until I glue a magnet to it and try it. Would it be safe to feed it lower voltage (say 9V instead of the 10 it wants)? My adapter is a 12V 1A and the fan is specced at 12V 0.55A. And could that be accomplished with a suitable resistor instead of trying to dig up a 9V adapter?

Oh, and apropos of nothing, I hate stranded copper wire. HATE IT. That is all.

[Jayme]

I think the issue is you're feeding the power source to the fan a varied voltage, and internal pwm controller a varied voltage also. I think you should be able to feed the voltage supply lead on the fan a constant source, and vary the pwm line with your controller.

[mumblecrunch]

A bunch of reading has led me to the conclusion that the PWM Controllers I have are not suitable for 4-wire fans. The controllers I have are designed for DC motors. The PWM output provides pulses of 12V power. 4-wire fans want steady power on the + and - lines and a logic signal on the PWM Controller line.

You can control a 4-wire fan with voltage pulses, but it's apparently not good for the fans.

I could build something like this, but then I'd have a different problem: that's a 5V device designed to control computer fans. It assumes that the fan is being powered by the 5V or 12V feed available in the computer. So in order for that controller to be useful I'd need to find a power supply capable of providing 5V and 12V and that is also small enough to fit in a stir plate enclosure.

I think I will have better luck trying to find a 3-wire fan. They can be PWM modulated using voltage like a DC motor. You apparently can't get the same level of control, but it's still better than a 2-wire fan with a potentiometer (or PWM for that matter: I plugged a 2-wire fan into the controller and it went from "not moving at all" to "blazing speed" in about 1/10 of the knob travel).

[Jayme]

Not sure why you think it would be better than a 2 wire fan - usually the 3rd wire is just a tach sample. Anyway, sounds like you are ultimately on the right track now.

[essenmein]

Probably a bit late but 4 wire fans already have a speed controller built in. I'm not sure if they are picky on control PWM frequency, but you could try connection red and black to 12V (or what ever its rated for) and your speed controller output to the PWM line, that way the variable PWM from the ebay controller drives the command of the fan (you may need a pull down resistor to make it work here), changing its speed. Or just tie the PWM line to +12V and run the pwm controller on pos and neg or just get a regular 2 or 3 wire fan.