Potted Eheim Pump Housing

Date Log; January 11, 2006

Results

At its simplest the primary performance measure of any pump is a combination of its head rate and flow rate. For our purposes head rate is considered to be an unimpeded measurement of the pumps ability to sustain a column of water. Flow rate is considered as the number of liters pushed by the pump through a given size of tubing measured against time.

Secondary performance issues relate to heat added by the pump as part of carrying out its pumping function. Pumps rated with a high operating wattage introduce more heat into the coolant fluid. Pumps immersed in the coolant, as is the case here, add all of their heat to the system so this effect has to be taken into consideration as part of their performance.

I often talk to people who opt for a more powerful pump. They hope to gain higher flow rates and improved head to assist the cooling loop only to find that they have added far more heat than the gain that is available to them from the increased pumping ability.

The third and final consideration is resistance. That is the pumping effect on the cross sectional area of the tubing used in combination with the system components used during testing. For the purpose of our testing we used 1/2" tubing as this is currently the most commonly used form of tubing.

Test Bed

The results here should reflect all of the considerations discussed above that is Head, Flow rate and Temperature gain added by the pump. To do this we need to have the ability to alter the flow rate and increase the work done by the pump to measure what if any effects occur.

Head;

Most pump manufacturers produce a pump performance graph indicating the maximum capable head of the pump so apart from stating this value we have very little application for measurement of this value. The Eheim Compact 600 has a head value of 1.3M.

Flow rate;

Unimpeded by restriction the Eheim Compact 600 is rated at 600 l/h

To measure this value we have constructed a loop contain a CPU block and two GPU blocks coupled to a dual pass radiator. Loop configuration was

Pump>Radiator>CPU block>GPU block one>GPU block two>Pump

Using this configuration the Eheim Compact 600 was able to sustain 3.85 l/min. ( 231 l/h)
Removing the second GPU block raised the value to a very creditable 4.11 l/min (246.6 l/h)

Temperature gain;

The Eheim Compact 600 is stated to have an 11 Watt power consumption.

We measured the reservoir temperature value with no heat source attached to any of the blocks and no fans running on the radiator. The heat added to the water would have to come from changes in ambient over time and the pumps own addition to the cooling loop. During the course of the testing we measured the gain until a stable temperature was reached for a period of two hours the gain temperature added to the water was an amount of 1.5 degrees.

Longevity;

The final testing addresses the longevity/durability of the pump and housing. This has already commenced on our own server which operates twenty four/seven so it will get a good work out during the next couple of months of high ambient summer temps. Given the reliability of Eheim pumps I would not wait around for a pump failure anytime soon.

Conclusions;

I am very satisfied with the potted Eheim project. The Eheim Compact range of pumps when used in combination with a reservoir adds a very neat finished look to any cooling loop setup. No more fiddling with small reservoirs in confined places. Checking the coolant condition and fluid level is a breeze. The footprint inside the case lends this application to many smaller cases which may be overwhelmed by the use of many components within the confined space available. Of course we have relied on the 240 volt version of this pump to test our Potted project something which does not appeal to some people who prefer the freedom of 12 volt application. For those 12 volt aficionados not to worry enter the German company Alphacool with their very extensive range of potted Eheim 12 volt pumps in 600 and 1000 l/h.

We managed to obtain a couple of these for testing and so far the results very closely follow the 240 volt versions. We should have more information about these versions at a later date when testing is competed.