It took a little tinkering to finally get a decent setup. One problem with pump testing using a reservoir is you will have static pressure that changes as the reservoir emptys. Also this reservoirs added pressure can "help" the pump along if the outlet of the test loop doesn't match and it lower. A closed loop will not have that problem, but I also needed something that bleed really fast and has absolutely minimal inlet pressure losses, so a custom reservoir with 5/8" barbs and radius inlets was key.
Here is the loop on my test bench. And no this isn't in my bathroom sink, I have a long bench in my home office next to my computer where I spend alot of time tinkering..
Pump> Pressure test T > Flow Meter > Globe Valve > Reservoir >
Seems to work really well, with as minimal a pressure drop as possible. Also because my manometer is adjustable to zero, I can remove the static pressure that is on the testing T from the loop above..
So first up, how about a maximum pressure head test. No it's not the 11' noted in specs everywhere, closer to 13'..:yessir:
And how about a maximum flow rate, note this still has 30+ inches of pressure due to the loop and flow rate meter, so it's not the pump maximum where pressure = 0, it's the maximum I can do with the pressure drop within the loop.
And to get the full picture, here is the systematic test results incrementally that were needed to develop the full curve. The end results were slightly lower than Laing's own graph, but it has the same curvature, and it's higher than swiftech's.
Nothing too spectacular here, it's about what I expected for the stock top, more than anything I now have a good baseline to compare a custom top to. The DDC3.2 with Petra top according to Petra's testing is still superior at 12v in the higher pressure side of the curve (The D5 doesn't take the lead until 2.5+ GPM range). I'll post some more curves to compare things later...
But what really had my curiosity was going over 12V. This is where things get interesting...
I chose to compare results of adjusting voltage while testing maximum flow rate. This produces the most extreme current draw from the pump and highest flow rate.
First up 12.0V nothing too special.
Then I noticed something right off. While slowly turning up the voltage, the maximum and plateau begins right at about 13.1V. Anywhere from 13.1V to 24V produced the exact same performance. It's strange, I can actually turn it up a bit and see the flow rate go up, and a moment later it back down to the 13.1V performance.
I've heard there was circuitry in the pump, and apparently this was clearly working. Not sure if the circuit adjust by maximum RPM or something else, but you can very clearly see it react to the jump in voltage as it slows the pump right back down.
So this is the Limiter test...
THE MAXIMUM CURRENT DRAW
Now....how about 24V volts....
At first I was thinking heat dump was tied to Amperage, it's not.
As Fairydust pointed out below. Wattage is Voltage x Amperage which is about 28 to 29 watts for anything at or above 13.1V.
So in the end, nothing suprising here other than I can say the limit on gains is a dismal 13.1V. Probably not really worth it to run any sort of special PSU on the D5's then..
After getting some great advise and looking around, I decided to record voltage and amperage relative to flow rate to make a "Watts used vs Flow Rate" curve.
This is the result:
Stock top D5 at 24V
And the direct comparison of 12V vs 24V. With this, I can say I see no reason at all to run at anything over 12V. The performance between 12V and 24V is exactly the same where it matters in water cooling. It's not until beyond 2.5GPM that you even start seeing any gain with the 24V curve. Also 24V is typically using about an extra 1 watt of power.
Last Update, Testing Complete
Here is a summary graph of all 5 settings tested along with a 12V and 24V on setting 5. I guess the one thing that became more obvious for me was how "Variable" the power consumption is.
Setting one may only draw 3 watts, while setting 5 could be nearly 10X that amount. It's probably worthwhile on light heat loads or smaller radiator setups to try some of the lower settings. The lower heat dump of these lower settings may be more beneficial than the higher flow rates in very special cases.
Also it's very obvious, and I guess this is consistent with other pumps, but power consumption go up as flow rates are increase (less restriction).
I'll plan on adding these into the estimator.