Day 26-30
I was a little under the weather, so I didn't make a lot of progress. I picked up a $10 Project Dual G5 2.0GHz from Free Geek Minneapolis.
I picked up some perf-board from Micro Center. I pulled a couple PWM fan headers from a donor broken PC motherboard.
I did manage to solder up my LCS pump to PWM converter.
I had to rewire my fans as they are the DELL standard and not the "everbody not DELL" version. Dell on the left, "normal" on the right.
When I plugged it in to test it, only one of the fans came on. I rechecked my soldering and discovered I misplaced one wire, so I fixed that and reconnected everything. Still only 1 fan. So I wondered, did I somehow fry something in my probing and testing?
I searched a little more and found out that the PWM signal in a computer is typically the 25KHz range. The Arduino Pro Mini has pins that produce 490Hz and 980Hz. Depression set in, I didn't want to have to re-engineer the whole thing with some extravagant code. It can of course use counters to actually create the 25KHz, but other searches showed that the fans don't care, as long as the PWM percentage is correct for what your intended speed is. Days go by.
Here we are on Day 30, and I get back into it. I go back to retest the wiring harness and pin 2 is no longer registering any voltage. Again I think that I fried something. So I restart with the ASD CD and run the hardware tests. Intake fans not working properly. What? I look and sure enough, neither is running. FEAR: oh I fried something so bad! So dug deeper. I recall that I tried to remove the original wiiring harness from the motherboard, but then realized it also powered the intake fans. So I removed it again and saw that I had it 1 pin off. The pins/socket are keyed, but I must have pushed a pin through the plug. I aligned it correctly and seated it in. Now the intake fans were working.
Back to ASD CD and run the CPU and fans tests. All fans and CPUs pass, the PUMP fails (fans too fast). I have one of my PWM running full board and the other slowly so I will go back and do benchmarks to see if either changes speeds.
Nope... something is wrong. recheck my wiring and layout. Oh, I have my voltage divider resistors backwards. Must have been brain fog when I was not feeling well.
I will keep working on this tonight and update this post with progress...
[EDIT]
I got the resistors sorted correctly, but I am still getting a full 11.8V on the driver circuits, so I think the controller is getting the wrong Fan Speed readings from my PWM fans, or at least not what it expected. I then though I would try to turn down the max speed to 50-60% to see if that would be acceptable to the G5Q, but the Arduino Pro Mini 5V stopped communicating and I could not update it. Project for later or next year. I still have to 3D print a fan cowl to cover the air coolers. It is fully working under load, so I am going to leave it as is.
Pre-Benchmark temps: High 30s
During load: low 70s
1 minute post test completion, back down to low 30s