fan current draw issues need EE advice
 

I have been having problems because my fans draw too much current, especially at idle. I was thinking about making a fan controller to address my problems an hopefully the problems of others as well.

Basic premise is:
Controller: Arduino
Power: 12v switched, w/ volt reg for chip
Inputs:

• Fan 1 circuit
• Fan 2 circuit
• RPM

Basic idea is that when the fans turn on, it instead signals the controller to run them (via some big mosfets). But it only runs them at 100 percent duty cycle above 2500 RPM. Up to 1500 RPM 50 percent cycle, up to 2000 75 percent. This keeps everything happy at idle.

Questions-

1. Was going to sense RPM by splicing into the trigger wire on one of my COPs and feeding it to an opto-isolator. This shouldn't affect the wire I'm tapping in to, should it?
2. Was going to do something similar with the 12v fan power... Do I just connect the 12v line to the isolator and then to ground or am I missing something?
3. Can you run fans via PWM?
4. Am I missing some obvious reason this is a dumb idea?

I know nothing about Arduino but my fans are PWM controlled. I also had to run an external 35a relay because the factory one is only good to 10a

Sure, what you are describing could work. The Arduino Uno can only drive 5V, so you have to chosse switches that are logic level. Make sure you flyback the drain to VBATT, or alternatively get a FET with a large VDS voltage. I suspect a FET with high VDS at the current you want, logic level may be a little harder to find.

I would use a relay for the high side (12V), or allow the main ECU to continue controlling the 12V turn-on and splice your FETS into the low side (Return) on the fans.

Most Miata versions have a tach signal, I would use that instead of the coil triggers. I don't think you will need to opto isolate that, but you will need to level translate. Running a precision counter in an arduino requires interrupts. You can do it with polling if you don't mind a little error in the signal. Another option would be to put the TACH on the 20dB slope of either a low pass or high pass filter followed by a diode and secondary R C to ground. Then you can sample it directly with the ADC.

I just had an epiphany. I'm going to disconnect the fans from each other (I did the old ECU wire cross trick years back). I have a theory that having both running at the same time when the AC is off isn't really that important. I mean, so long as you are ducted properly, the fan only works when you're not moving.... Which means low thermal load.

If that doesn't work, I'll break out the soldering iron.

Edit- I'll put a divider between the fans so only one fan on doesn't suck air through the other hole instead of the radiator. Yay riveting.