Pro/Forums - View Single Post - How much pump is enough? How much is too much?

Cathar · 11-04-2004, 04:30 PM

Plotted a graph and worked out some figures for some in-series pump solutions in order to explore the "ideal" pumping scenarios I touched on above.

2 x Eheim 1048 => 6W of in-line heat
Swiftech MCP600 @ 13.8v => 11W of in-line heat
Iwaki MD-20RZ @ 50Hz => 25W of in-line heat
2 x Swiftech MCP600 @ 12v => 17W of in-line heat
2 x Swiftech MCP600 @ 13.8v => 23W of in-line heat

The PQ curves work out as follows:

Flow, radiator C/W, and block C/W works out as follows:

2 x Eheim 1048, 5.6 LPM, 0.042, 0.169
Swiftech MCP600 @ 13.8v, 6.8LPM, 0.041, 0.164
Iwaki MD-20RZ @ 50Hz, 7.5 LPM, 0.040, 0.162
2 x Swiftech MCP600 @ 12v, 7.2 LPM, 0.041, 0.163
2 x Swiftech MCP600 @ 13.8v, 8.7 LPM, 0.040, 0.159

...and our predicted CPU temps are:

2 x Eheim 1048, WB Delta = 16.9C, Water Delta = 106 * 0.042 = 4.5C, CPU Temp = 21.4C above ambient
Swiftech MCP600@13.8v, WB Delta = 16.4C, Water Delta = 111 * 0.041 = 4.5C, CPU Temp = 20.9C above ambient
Iwaki MD-20RZ @ 50Hz, WB Delta = 16.2C, Water Delta = 125 * 0.040 = 5.0C, CPU Temp = 21.2C above ambient
2 x Swiftech MCP600 @ 12v, WB Delta = 16.3C, Water Delta = 117 * 0.041 = 4.8C, CPU Temp = 21.1C above ambient
2 x Swiftech MCP600 @ 13.8v, WB Delta = 15.9C, Water Delta = 123 * 0.040 = 4.9C, CPU Temp = 20.8C above ambient

Summary

As expected, the 2 x Eheim 1048's pretty much exactly match a single Swiftech MCP600, and provides the best all-round solution when silence is required.

The single MCP600@13.8v shines through out of the blue here as the most promising single pump solution. It is only bested by the dual 13.8v MCP600 solution, and even then it's only just. The MCP600 pump would require a single 2A rated 13.8v PSU to give this sort of performance. Jaycar sells a suitable Switch-mode wall unit here for $60. They also offer a cheaper, somewhat border-line solution here for $30.

The 50Hz Iwaki MD-20RZ comes extremely close to matching the 60Hz Iwaki MD-20RZ as the best all-round single pump solution.

The 2 x MCP600's @ 12.0v, as described above, provide an excellent alternative to the Iwaki MD-20RZ solutions.

The 2 x MCP600's @ 13.8v pretty much exactly model the "perfect pump" scenario that I was describing a few posts back, and offers a significant gain over even the 60Hz Iwaki MD-20RZ. Now 2 such pumps would be drawing around 33W of power, or about 2.4A @ 13.8v. You would probably want to use at least a 4A rated 13.8v PSU to power the pumps up from standstill.

I still hold great confidence in saying that the "ideal" water-cooling pump, that is a realistic and feasibly built possibility (based on the MCP600), would be:

6-8mH2O (20-26') of peak pressure head
12-13LPM (3.2-3.4 US GPM) peak flow rate
25-30W nominal power draw (when in operation)
12VDC
Separated Magnetic Armature design (~16-20W of heat added to the water)
Centrifugal pump

Any pump which has specs stronger than this starts to take strides backwards in terms of offering better performance.

11-04-2004, 04:30 PM	#2
Cathar Thermophile Join Date: Sep 2002 Location: Melbourne, Australia Posts: 2,538	Additional pumping scenarios Plotted a graph and worked out some figures for some in-series pump solutions in order to explore the "ideal" pumping scenarios I touched on above. 2 x Eheim 1048 => 6W of in-line heat Swiftech MCP600 @ 13.8v => 11W of in-line heat Iwaki MD-20RZ @ 50Hz => 25W of in-line heat 2 x Swiftech MCP600 @ 12v => 17W of in-line heat 2 x Swiftech MCP600 @ 13.8v => 23W of in-line heat The PQ curves work out as follows: Flow, radiator C/W, and block C/W works out as follows: 2 x Eheim 1048, 5.6 LPM, 0.042, 0.169 Swiftech MCP600 @ 13.8v, 6.8LPM, 0.041, 0.164 Iwaki MD-20RZ @ 50Hz, 7.5 LPM, 0.040, 0.162 2 x Swiftech MCP600 @ 12v, 7.2 LPM, 0.041, 0.163 2 x Swiftech MCP600 @ 13.8v, 8.7 LPM, 0.040, 0.159 ...and our predicted CPU temps are: 2 x Eheim 1048, WB Delta = 16.9C, Water Delta = 106 * 0.042 = 4.5C, CPU Temp = 21.4C above ambient Swiftech MCP600@13.8v, WB Delta = 16.4C, Water Delta = 111 * 0.041 = 4.5C, CPU Temp = 20.9C above ambient Iwaki MD-20RZ @ 50Hz, WB Delta = 16.2C, Water Delta = 125 * 0.040 = 5.0C, CPU Temp = 21.2C above ambient 2 x Swiftech MCP600 @ 12v, WB Delta = 16.3C, Water Delta = 117 * 0.041 = 4.8C, CPU Temp = 21.1C above ambient 2 x Swiftech MCP600 @ 13.8v, WB Delta = 15.9C, Water Delta = 123 * 0.040 = 4.9C, CPU Temp = 20.8C above ambient Summary As expected, the 2 x Eheim 1048's pretty much exactly match a single Swiftech MCP600, and provides the best all-round solution when silence is required. The single MCP600@13.8v shines through out of the blue here as the most promising single pump solution. It is only bested by the dual 13.8v MCP600 solution, and even then it's only just. The MCP600 pump would require a single 2A rated 13.8v PSU to give this sort of performance. Jaycar sells a suitable Switch-mode wall unit here for $60. They also offer a cheaper, somewhat border-line solution here for $30. The 50Hz Iwaki MD-20RZ comes extremely close to matching the 60Hz Iwaki MD-20RZ as the best all-round single pump solution. The 2 x MCP600's @ 12.0v, as described above, provide an excellent alternative to the Iwaki MD-20RZ solutions. The 2 x MCP600's @ 13.8v pretty much exactly model the "perfect pump" scenario that I was describing a few posts back, and offers a significant gain over even the 60Hz Iwaki MD-20RZ. Now 2 such pumps would be drawing around 33W of power, or about 2.4A @ 13.8v. You would probably want to use at least a 4A rated 13.8v PSU to power the pumps up from standstill. I still hold great confidence in saying that the "ideal" water-cooling pump, that is a realistic and feasibly built possibility (based on the MCP600), would be: 6-8mH2O (20-26') of peak pressure head 12-13LPM (3.2-3.4 US GPM) peak flow rate 25-30W nominal power draw (when in operation) 12VDC Separated Magnetic Armature design (~16-20W of heat added to the water) Centrifugal pump Any pump which has specs stronger than this starts to take strides backwards in terms of offering better performance.