Installation Positions and Performance
We skip our traditional testbed description in this review because this testing wasn't fully complete. We can only say that the testbed and our testing methods haven't changed and the room temperature was 26°C during the test.
Since the user manual doesn’t mention the best way to orient the cooler on the CPU, we decided to check this out on our own. We first installed the cooler with its heat pipes at a right angle to the memory modules.
We set the speed of the intake and exhaust fans installed on the heatsink at the maximum 1300 RPM and were astonished at the results: the temperature of the hottest CPU core was as high as 86°C after three cycles of the first LinX pass. The test itself aborted with an error message after less than 5 minutes.
Our further attempts led to the same results with minor variations in time. Recalling that the efficiency of down-draft coolers depends on the type of the system case, especially as the Phanteks PH-TC14CS is as tall as 160 mm (with one fan), we removed the side panel of our case and ran the test once again. Unfortunately, this only affected the duration of the test, but not the cooler’s efficiency:
The peak temperature of 91°C and the thermal throttling of the CPU mean that there’s something wrong about the contact between the cooler and the CPU. And that’s indeed what we saw when we took the cooler off:
One third of the cooler’s convex base did not touch the CPU heat-spreader, which explains the low efficiency of the Phanteks PH-TC14CS.
Then we reinstalled the cooler after turning it around by 90 degrees clockwise so that the ends of its heat pipes were up.
Surprisingly, we got the same results if the side panel was closed: the peak CPU temperature was 86°C and LinX aborted after 7.5 minutes.
When we removed the side panel, the system passed two 11-minute tests, the CPU being no hotter than 81°C.
It’s clear that the classic layout of our ATX system case isn’t a good match for the Phanteks PH-TC14CS which requires an open testbed or at least a chassis without a side panel.
When the speed of the fans was reduced from 1360 to 1000 RPM, the CPU was again as hot as 86°C and the test aborted with an error message. This orientation of the cooler produced an acceptable thermal grease imprint, though, so poor contact could not be the cause of the poor performance.
We reinstalled the cooler, checking out the pressure of its fastening mechanism and running some more tests, but couldn’t make our CPU stable at lower speeds of the fans. So, the Phanteks PH-TC14CS just can’t stand a comparison with any other cooler. Priced at $75-80, it is inferior even to the modest $40 Thermalright TRUE Spirit 140 whereas its senior cousin PH-TC14PE is 15°C more efficient under the same conditions, in a closed system case and with two quiet 135 mm fans at 900 RPM.
We also tried to run the Phanteks PH-TC14CS with only one fan (above or below the heatsink) but to no effect. The cooler refused to keep our CPU stable unless with two fans at the maximum speed and with the side panel of our system case removed.
It’s sad but this test session is incomplete just because the Phanteks PH-TC14CS turned out to be but one half of a super-cooler both from a technical standpoint and in actual performance. The small total effective cooling surface of its heatsink, the excessive height (for a down-draft cooler) with the top fan attached and, consequently, the high dependence on the type of the system case, the convex base and, possibly, the poor contact between the heat pipes and that base, the difficult installation procedure and limited compatibility – all these do not make the Phanteks PH-TC14CS such a good choice.
We can only hope that it can do better in a compact system case (with only the bottom fan installed) on a non-overclocked CPU. It may also be interesting to users who want to provide more cooling for their mainboard components and memory modules around the CPU socket. However, the high price of $75-80 makes it uncompetitive even in such limited usage scenarios. It seems to be really hard to justify an investment like that...