NZXT Sentry LX

Test Setup

Intel E6550 O/C 2.8 GHz @1.4V *
Cooler master Hyper212 CPU cooler
Gigabyte GA-X38-DQ6
4G Reaper X PC8500 (5-5-5-18 @1066Mhz @2.1V)
4G Reaper X PC8000 (5-5-5-18 @1066Mhz @2.1V)
NZXT PP-800 PSU
Diamond HD4870 512
ATI HD4870 512 [CrossfireX]
WD 640G Caviar (black)
Samsung DVD burner
NZXT Panzerbox Prototype
NZXT Sentry LX fan controller

* In order to increase the system temperature we increased the clock speed of the E6550 from 2.33 GHz to 2.8 GHz. We also increased the core voltage from 1.35V to 1.4V. The two HD4870's are running at stock speeds with auto fan set in the CCC.

Automatic Mode Tests

First off we will be recording temperatures and noise levels using the Sentry LX's default Auto mode. Using this mode the fan speeds are adjusted automatically by the Sentry LX using feedback from the temperature sensors. For these tests the temperature alarm is set to the default of 70C. In addition to running the Futuremark's Vantage and 3DMark6 benchmarks we will also be running Everest Stress test. Along with the benchmark programs we will also be testing the Sentry LX under 72 hours of normal everyday usage. The Futuremark benchmarks will be run once at default settings and the Everest Stress test for one hour intervals. Just as a side note the date and time are not accurate in the final pictures as we reset the Sentry LX several times during the tests.

Manual Mode - Full Speed

This series of tests is almost identical to the ones used in the Auto mode testing with the exclusion of the normal use scenario. We felt that running the system fans at full speed would only be required while running benchmarks or during heavy gaming at high resolutions. The temperature alarm was again left at the default of 70C and the rest of the system settings remain the same as the first series of tests.

Manual Mode - Half Speed

Due to the high noise level of the two large fans in the Panzerbox Prototype we decided to program a manual configuration for when the system was idle. All the fans were cut to 50 percent of their rated full RPM and we also enabled SpeedStep in Vista's CPU power management. On the E6550 during idle periods the multiplier drops from 7 to 6 lowering power consumption and heat as shown in the screenshots below. The system was tested under these conditions for an idle period of 48 hours and the results are also included below. As an additional precaution we set the threshold of the temperature alarm to 50C.

Thermal Analysis

When we selected the tests and testing modes we expected the half speed setting to show the highest recorded temperatures and obviously the full speed setting to show the lowest. To be honest we were quite surprised with the temperatures recorded during the half speed test and that prompts the question, Do we necessarily need all of our system fans running at full speed all the time to provide suitable cooling? The answer to that as shown in this review is of course NO!

The second observation we made is how closely the Auto setting was to the Manual full speed in managing system temperatures. In fact in some tests running the fans at full speed provided negligible differences in the recorded temperatures. This goes to show that even during system intensive testing the system fans don't always have to be running full speed to provide adequate system cooling. To give you an overall feel for the difference in temperature ranges we are going to show the Auto setting during normal use compared to the Manual low power setting below:

Acoustical Analysis

Using the Samson C01U USB Studio Microphone we previously reviewed we recorded sound levels at intervals during the tests. Since the Manual settings use fixed values we will show the highest dB levels for those as well as minimum and maximum for the Auto mode which fluctuates. For all tests we will be testing sound levels from the room level noise using Sound Forge. Let us take a look at these and then move onto some final words on the NZXT Sentry LX fan controller!