SUMMARY: Very even performance between 0.9 to 0.3 gpm – an interesting design.
The good guys at Thermal Transtech were nice enough to send a sample of their nPowerTek Heat Column Waterblock to test. This is a very interesting design – it incorporates a 25.4 mm heatpipe into the waterblock. If you look at the bottom of the waterblock
you can see that the heatpipe’s base contacts the CPU top, and the heatpipe is surrounded by a water jacket for cooling. The inlet/outlet is on the small side – 5 mm ID, 6 mm OD. Parts are included for Intel LGA775, P4 socket 939, 940 and 478 and AMD K8.
The nPowerTek Heat Column Waterblock was tested using the CPU Die Simulator and the P4 CPU Simulator on the Waterblock Test Rig. The small tubing size took its toll on pressure drop – I could not get more than 0.9 gpm through it, and as such these results are not directly comparable to others I have done at 1 gpm.
| Waterblock | C/W | Pressure Drop – psi | Pressure Drop – inches H2O |
| nPowerTek – Small Die, 0.9 gpm | 0.22 | 4.20 | 116.3 |
| nPowerTek – P4 Die, 0.9 gpm | 0.38 | 4.20 | 116.3 |
| nPowerTek – P4 Die, 0.5 gpm | 0.38 | 1.70 | 47.1 |
| nPowerTek – P4 Die, 0.3 gpm | 0.39 | 0.86 | 23.8 |
I’ve included results ranging from 0.9 to 0.3 gpm – note that, in contrast to other waterblocks, the nPowerTek Heatpipe Waterblock does not vary greatly – a mere 0.01 C/W between 0.3 and 0.9 gpm. In contrast, other waterblocks show more substantial changes over this range.
Test Results indicate that the nPowerTek Heat Column Waterblock’s pressure drop (or head loss) across the waterblock is very high at 0.9 gpm; flow resistance such as this means that nPowerTek will require a high power waterpump to attain high flow rates. However, it appears that the performance increase may not be worth the added expense – an interesting turn of events.
The P4 die test is included as a point of interest – I must caution that these results have no comparison to others, as the P4 die tester is new and I don’t have many results to compare against yet. It does appear that the P4 die tester is a very tough test, as the die temps are high; I will be placing a thermocouple on the case as another measurement point.
nPowerTek’s Heat Column Waterblock is a very interesting animal – it appears that it may perform very evenly over various flow rates, not a bad characteristic for use in low-flow watercooling systems.
Disclosure: Joe Citarella has a financial interest in a company developing thermosyphon products for electronic chip cooling.
Be the first to comment