Review: Zalman 7000-Cu / AlCu CPU Heatsink

Cooling
Viewing page 3 of 3 pages. Previous 1 2 3

TESTING

The integrated fan in the Zalman 7000 series coolers does not allow the use of our standard heatsink testing method of a Panaflo 80L (80mm low speed fan) at 12, 7 and 5 volts. (This procedure allows heatsinks to be compared directly without the variable of different fan airflow / noise.) The integrated fan and Fan Mate 1 were used for the testing.

Key Components in P4 HS Test platform

Intel P4-1.8A Northwood core - Nominal power is ~50W; may increase to 62.4W if speed throttling doesn't stop it first. 67° C rated maximum junction temp. The plug will be pulled if any HS lets the temp go much above 60° C. For ° C/W calculations, 50W is assumed.

Intel D845PEBT2 motherboard - Intel 845PE Chipset; on-die thermal diode monitoring

Any VGA card (AGP)

256 MB DDRAM - PC2100 generic

Any hard drive (in Smart Drive from Silicon Acoustics)

DigiDoc5 w/ thermal sensors

Any Good PSU

Arctic Silver 3 Thermal Compound

Two-level Plywood board with foam damping feet. Motherboard on top; most other components below. Eases HS changes and setup.

CPUBurn is the stress program used to load the CPU to 100%. It heats up the CPU a bit more than just about any other utility tried so far (including Prime95).

The heatsinks were cleaned and installed in turn on the test system as per the manufacturers' and Arctic Silver's instructions. The system was allowed to cool between tests for ~5 minutes with a large AC domestic room cooling fan running at full speed, the airflow directed over the entire test setup.

TEST RESULTS

  • All temperatures in degrees Celsius.
  • Diode: reading from P4-1.8A CPU thermal diode, recorded by Intel Active Monitor..
  • Room ambient is the average of temps measured at several points 2~4 feet distance around the test platform.

A. With P4 -1.8A (50W) -- Room ambient: 26° C

7000-Cu or -Al/CU
° C, Idle
° C, CPUBurn, 20+ mins
CPU
Temp rise
CPU
Temp rise
° C/W
5V (1440 rpm)
30
4
41
15
0.3
7V (1840 rpm)
30
4
41
15
0.3
10V (2450 rpm)
29
3
39
13
0.26
12V (2700 rpm)
29
3
38
12
0.24

At this ambient temperature, the heat dissipated by the P4-1.8A is easily handled by either 7000 model in exactly the same way. It appears that copper's natural heat transfer efficiency advantage is not showing up because both coolers are working well within their limits.

The performance at 5V is very good, as it the noise. The measured 1440 RPM at 5V is about the same speed as a Panaflo 80L at ~9.5V, and the noise level is similar. With its larger diameter, the 7000's built-in fan moves more air than the reference Panaflo 80L at the same speed. At this setting, the 7000 equals the ° C/W reached by the Thermaright SLK900U + Panaflo 80L at 12V and is definitely quieter.

The character of the fan noise is benign, much preferable to the separate 80mm and 92mm fans Zalman has used thus far. It is mostly turbulence noise with a touch of buzziness at the sub-7V settings. This is somewhat surprising because the separate fans are sleeve bearing models while the embedded fan in the 7000 is a ball bearing design, which usually means a bit more noise.

Surprisingly, the performance at 7V is the same as at 5V. At 7V it is clearly noisier, slightly noisier than a Panaflo 80L at 12V. Naturally, as there is no gain in performance, the 5V setting is preferred here.

The performance at 10V (about the maximum clockwise setting on the Fan Mate 1), cooling performance moves into the champion class with 0.2 ° C/W. Cooling is improved by a degree with the move to 12V, achieved by direct connection to the motherboard header.

At above ~7V, the noise level becomes too high for SPCR, however.

B. With P4-2.53 (sSpec SL6EG: 61.5W) -- Room ambient: 25° C

This test was run to check the accuracy of the testing system, and also because the P4-1.8A may simply run too cool to be a useful gauge of real world heatsink performance as 800 MHz bus P4s move into the >3GHz range. The P4-2.53 (sSpec SL6EG) runs with a Vcore of 1.5V, a 533 mHz bus speed, and is rated at 61.5W. This CPU was dropped in place and run at stock presets in the standard Intel P4 motherboard / testing platform.

In theory, ° C/W should remain consistent for a HSF regardless of CPU heat and ambient temperature -- at least within a range. Intel's spec of 61.5W was used for ° C/W calculations with the P4-2.53.

CPUBurn >20m

7000-AlCu
7000-Cu
CPU
temp rise
° C/W
CPU
temp rise
° C/W
5V (1440 rpm)
44
19
0.31
44
19
0.31
7V (1840 rpm)
41
16
0.26
42
17
0.28
10V (2450 rpm)
40
15
0.24
41
16
0.26
12V (2700 rpm)
40
15
0.24
40
15
0.24

The ° C/W results were very close but not identical to those obtained on the lower power CPU. The numbers were a touch poorer all around, but this is probably well within the margin of error in the system.

The same can be said of the differences seen between the AlCu and Cu models. They are essentially indistinguishable in performance with the P4-2.53 in this ambient as well. Certainly, the Zalman 7000 HSF has no problem cooling a P4-2.53 even at minimal fan speed settings. One suspects that with the very hottest CPUs or with higher ambient temperatures in a system installed normally inside a case, the copper version would perform better that the AlCu, especially with the fan at low speed.

CONCLUSIONS

Real world conditions are tougher than those in the SPCR lab. When a system is installed in an enclosed case, the ambient temperature is sure to rise at least 5° C, more often 10° C or even more, depending on particulars. Many quiet PC enthusiasts run systems with no case fan or just one low airflow fan. In consideration of these factors, low noise enthusiasts are urged to add 10° C to the test results when trying to guesstimate what their temperatures would be. Even better would be to measure the in-case temperature near the CPU fan and use the C/W figures to calculate likely CPU temperatures with these HSF.

It is best to regard the test results not in an absolute way, but rather, as comparative guidelines. The result obtained with this heatsinks will vary greatly on the particulars for each system.

The Zalman CNPS7000-Cu and CNPS7000-AlCu both provide superb cooling performance even with the fan at a quiet 5V. With the worst-case 5V result of 0.31° C/W (with the P4-2.53), even a 82W P4-3.0 would see a temperature rise of just 25° C. This means that as long as the case ambient temperature can be held to 35C or lower, a safe max temp of 60° C under heavy CPU load can be maintained.

This is about the best performance at this noise level (equivalent to a Panaflo 80L at ~9V) of all HS tested thus far. If you carefully examine the results for the Thermalright SLK900, you'll see its temp rise with a Panaflo 80L at 9V would be about the same as for the Zalman 7000 at 5V.

While the noise level at 5V is higher than that achieved with a Panaflo 80L at under 9V, many will find it quiet enough. Despite the same rating of 20 dBA at 1 meter at 5V, the 7000 fan is subjectively much quieter than Zalman's stand-alone 80mm and 92mm fan offerings. If you seek the lowest noise, you'll have to find a way to slow the captive 7000 fan to lower than 5V, swap the captive fan for something quieter, or go with a HS that lets you choose your own fan.

It may not be a silent champ, but it is probably about as quiet as you can get while effectively cooling the hottest CPUs Intel is churning out these days. Given that it is as effective as the more expensive Cu version, and less stressful for the motherboard with its lower mass, the 7000-AlCu model is a bargain. Strongly recommended for power lovers who still want quiet.

The 7000 HSF and other fine Zalman products can be found at attractive prices at Sharka Corp.

* * * * *

Comment on this article in our Forums.



Previous 1 2 3

Cooling - Article Index
Help support this site, buy from one of our affiliate retailers!
Search: