SUMMARY: An in-line pump will run cooler than a submersible; a little more effort to set up, but results are impressive.

I received a few emails on the subject of using a submersible pump rather than an in-line model. In-line pumps have the advantage of working outside a water container so water temps are not affected by the pump's heat and bulk is cut down considerably. One potential disadvantage is there is no expansion tank in a closed loop system; with the temps at which most systems work, water expansion may not be a major concern.

However, trapped air is a problem - I would not like to have any air pockets in the system, and bleeding a closed loop system is absolutely required. I guess you could put the whole thing under water and assemble it without any air pockets, but for my taste, I would like to have some kind of visual reservoir.

One fellow suggested the Eheim 1046 (mag drive) as a good in-line pump, so I bought one at Pet Warehouse for $46 plus S&H. I was thoroughly pleased with these folks - shipped out right away and emails to confirm etc. This pump can be used submerged as well - the pic below shows the inlet screen in the background. This pump is not small - it is 5.7" x 3" x 4.1" (145mm x 75mm x 103 mm) and uses 1/2" hose at the inlet and 3/8" hose at the outlet.

If you read the specs on pumps, you will see that some can be used either submerged or in-line, so the Eheim is not the only one. It comes with a two year warranty and you can buy parts for it, so it should last a long time. In use, it was the quietest pump I have used yet - almost silent. As you can see from the pic below, I sat the pump on a small piece of foam to dampen any noise. It comes with a mounting plate, so it can be bolted down easily.

I first ran my submersible pump (Beckett G90) and then substituted the Eheim. You will notice a "T" at the water block - I used this to add an "expansion tank" and as a means to add water and visually check it. It also allows air bubbles to rise so that there are no air pockets in the system.

Is it necessary? You don't have to do this, but that's me. I had to use a collar on the water jacket to adapt 1/2" tubing to 3/8" intake - I used a small piece of 3/8" tubing and slipped the 1/2" over it, then clamped. Works fine.

THE TEST

I ran my 600E at 800 MHz with Prime 95. I waited until temps stabilized and took 500 one-second samples using Motherboard Monitor. results are shown below:

Eheim In-line Pump: CPU 25.7 C, System 24 C;
Beckett Submersible: CPU 29.0 C, System 25 C.

Can't argue with results like this - 3.3 C lower. While there are some differences in rated gph's, actual flow rates are close so that between the two pumps, the impact on the temp difference should be negligible. The Eheim is barely warm to the touch while running, so the amount of heat added into the case is minor; in fact, case temps were lower with the Eheim.

CONCLUSION

In-line pumps can yield lower CPU temps compared to submersibles. Bleeding the system to avoid trapped air is required, but it looks like results are worth the effort. In-lines also are easier to mount inside an existing case due to smaller space requirements.