Originally Posted by repo_man
I don't know what the RPMs are. I still haven't gotten the header plug put on the wire. Will do that tomorrow. I did get the voltmeter out today and checked the voltage. 12.20-12.21v to the pump. No fluctuation or anything, nice and steady.
Will update tomorrow with RPMs and loop situation (I'm going to tear it down and clean it thoroughly).
Martins lab did a review/post/test on over-volting the pump too see if flow rates could be bias'd one way or the other by putting more juice to it, and the answer was a flat Nope, looks like any additional volts going to the thing just poured off as heat/watts, so that said, the understanding is, that it can be undervolted to lower the speed, but overvolting seems to switch it to a different resistor/resisting element internally so it just heats up.
May of been built like that intentionally so that it would always "do" what it says it does on the box, no matter what dirty signal / crappy psu/12v service one puts it on. Or, that it is some sort of reverse current cutout designed to lower the applied voltage after the pump gets up to speed, (i wanna say the thing can juice upto 19 volts from a standstill to get going.)
If you dont have the 3 pin mobo header, then your pump is the 3.1? There is a mod that can be done to ramp up the pump to a 3.2 spec (solder mod) which makes it the same spec as the 3.2 (head pressure), afaik thats the only difference between the two, (that and the impeller is orange/white and not blue/orange) that and the mobo header.
Check your psu as well, even though the thing is the only device on the molex does not mean that internally to the PSU that is is not in series with another physical plug.
I learned this the hard when when I started running SLi, that even though it was on a different service/plug off the psu, that in fact, internally they were wired together in the psu. Had to switch to a SLi certified psu to fix my issues.
Lights/diodes/fans anything with variable speeds could cause the resistance to go up and down, changing the overall dynamic of the current in the line, causing the pump to throttle, causing something else to change, causing the pump to throttle again... food for thought. Got a spare old 200 watt psu lying around you could toss it on too see?