What do you think of the feasibility of modding the miniNG to make it suitable for true/quality PWM control of 2x G1238B12BBZP-00.
I briefly looked at the miniNG and BigNG, and wow, they're pretty good! If I had a lot of fans to control, I'd buy them.
And then I looked at the datasheet for that fan you want. That's a crazy fan.
-It looks like it's 4-pin PWM.
-But at 5KHz. Now the earlier posts make sense about 25KHz vs. 5KHz.
-Similarly, the 'true/quality' you refer to.
So... the miniNG... 50W/channel PWM, 3-pin.
50W/channel, and the fan's specs are
Input 53.4 W
Voltage 12 V
Operating Voltage 9.6 - 13.8 V
Current Rating 4450 mA
Current Starting 6100 mA
53.4W / 12V = 4450mA
-I'd say this is iffy at best. It'll probably work, but the miniNG's parts could be slowly dying.
-If not much extra headroom was given in the miniNG, it could die pretty quick, but I doubt it.
-To be safe, I'd only run 1x G1238B12BBZP-00 per miniNG.
-But there's a slim chance the miniNG's power electronics can't handle it (strictly speaking, 53.4W > 50W), so could be looking at completely redoing the miniNG's power output. Not happening.
3-pin / 'true/quality'
-Looking at the bigNG and miniNG, they seem to only have 3-pin outputs (unless I really missed something).
-They have nice photos of the bigNG (Atmega 16), but I can't quite make out the chip on the miniNG (looks like Atmega 8). I've worked with these before, and it looks like they've exposed the debugging pins, so custom firmware changes might be possible unless they've locked the chip. But hopefully it wouldn't come to that.
So.. first thoughts on modding the miniNG:
-I'd make a custom 4-pin header as so:
12V - connect straight to input molex.
GND - connect straight to input molex.
RPM sense - connect to same pin on miniNG's 3-pin fan header.
PWM - connect to PWM output directly on the mcu. It would be a simple matter to find which pin it would be via the datasheet, then hooking up an oscilloscope / good AC voltmeter to confirm. If the signal gets too degraded by the fan, then run it through a buffer.
This would, in theory, give you what you want, since I think the miniNG's software (via a bigNG) can control the PWM duty cycle in small increments (the true/quality requirements).
Potential problems:
-miniNG's power output is too weak (described above). If this is the case, then find a different controller to mod.
-The fan's 5 KHz PWM freq doesn't work well. If this is the case, then the firmware would have to be modified to change the PWM frequency. That could be as simple as changing 1 byte, or it could be much worse. It might not even be possible if they've locked the mcu. Otherwise, it might be possible to construct a frequency divider. That would take some thinking though.