Zalman CNPS7700-AlCu Heatsink/Fan

Table of Contents

The new Zalman 7700 heatsink/fan is a natural extension of the radial design of the successful 7000 series to the higher thermal demands of the most recent processors — particularly the fastest Intel Prescott core P4s with their >100W output. An integrated 120mm diameter fan is employed instead of a 92mm fan. How does the biggest and best from Zalman compare against its precessors, and against the Thermalright XP-120 with a 120mm fan of one’s choice?

Jan 5, 2005 by Mike Chin with Sean Boyd and Jordan Menu

Product
Zalman CNPS7700-AlCu heatsink/fan for socket 478 / 754 / 940 / 939 / LGA775
Supplier
Sharka Computers
Price
~US$45.00

The Zalman 7000 series radial heatsink / fans have been something of a reference standard for SPCR since they were introduced over a year and a half ago. Anyone who follows CPU heatsink development in the PC world knows that’s an unusually long time for a heatsink to stay in production, never mind stay on top. In many ways, the 7000 series represents the most efficient realization of Zalman’s original concept of combining multiple thin metal plates into a heatsink rather than the conventional approach of extruding, soldering or skivving the heatsink. All of the kinetic energy of a 92mm fan is employed for cooling by embedding it without a frame in a large radial heatsink.

The new 7700 series from Zalman is a natural extension of the 7000 concept to the higher thermal demands of the most recent processors — particularly the fastest Intel Prescott core P4s with their >100W output. An integrated 120mm diameter fan is employed instead of a 92mm fan.The dimension of the HS have naturally been increased to accommodate the larger fan. The photo below shows the difference.


7700 vs. 7000

We’re examining the aluminum / copper hybrid version of the 7700, specifically because its mass is a relatively modest 600 grams. The all-copper 7700-Cu tilts the scale at 916 grams, which we think is too high and best avoided unless the extra cooling capacity is deemed absolutely necessary. A high mass heatsink stresses a motherboard during installation and in normal use (in a tower style case) much more than lighter ones, and it does not take that much flexing to damage a multi-layer motherboard. This is especially true for socket 478 with either the 7700-Cu or 7000-Cu, which use the stock plastic HS retention frame and put high stress on the four holes in that frame. It may not be as serious an issue for K8 or LGA-775, for which both the 7700 and 7000 series have more secure and direct hardware with through-the-board support.

As with the 7000 series, the 7700 is a cross-platform device with compatibility for all the newer CPU socket types: The Intel 478 and 775, and the AMD 754, 940 and 939. The older Socket A and 370 are ignored; this heatsink would be overkill for any CPU in that form factor, and its size makes it highly incompatible.


All the hardware for the major CPU types as well as the new Fan Mate 2.

The 7700-AlCu comes with a complete set of hardware, including a revised version of their fan speed controller that allows it to be mounted outside the PC case for convenient access.


In order to make the above possible…


…they made the now detachable lead very long, with both input and output on the same connector.
The wire is meant to be split (into input / output) as needed.


The control knob is bigger and easier to use, and there is a mounting screw eyelet on one end.

ZALMAN’S SPECIFICATIONS

Zalman cites a °C/W performance difference of .02~.04 (depending on fan speed setting) between the lighter hybrid and the heavy all-copper versions. Whether the difference is worthwhile really depends on your application, but in practice this kind of difference can be easily swamped by differences in case airflow and other component choices.

Specifications CNPS7700-AlCu CNPS7700-Cu
Dimensions (mm) 136(L) x 136(W) x 67(H)
Dissipation Area (cm2) 3,268
Base Material Copper & Aluminum Copper
Weight (grams) 600 918
Thermal Resistance (°C/W) Silent Mode 0.28 Silent Mode 0.24
Normal Mode 0.21 Normal Mode 0.19
Integrated Fan 120 mm diameter, 0.35A, 12VDC
Fan speed 1,000 ~ 2,000 rpm
Fan noise 20 dBA min; 32 dBA max, ±10%

It’s also interesting to look at the differences between the 7700-alcu and the smaller 7000-AlCu

Specifications CNPS7700-AlCu CNPS7000-AlCu
Dimensions (mm) 136(L) x 136(W) x 67(H) 109(L) x 109(W) x 62(H)
Dissipation Area (cm2) 3,268 3170
Base Material Copper & Aluminum
Weight (grams) 600 445
Thermal Resistance (°C/W) Silent Mode 0.28 Silent Mode 0.29
Normal Mode 0.21 Normal Mode 0.22
Integrated Fan 120 mm diameter, 0.35A, 12VDC 90 mm diameter, 0.23A, 12VDC
Fan speed 1,000 ~ 2,000 rpm 1,350 ~ 2,600 rpm
Fan noise 20 ~ 32 dBA, ±10% 18 ~27.5 dBA, ±10%

It’s curious to note that:

  • The total cooling area is increased by only about 3%. This means the bulk of the gains in performance comes from the increased airflow of the bigger fan
  • The performance is improved by only 0.01°C/W.
  • The noise penalty for the tiny performance gain is +2 dBA at minimum noise / cooling and +7 dBA at maximum.

In short, judging by Zalman’s own specifications, the 7700-AlCu has a tough lead to follow in the 7000-AlCu

How about the differences between the all-copper versions of these HSF?

Specifications CNPS7700-Cu CNPS7000-Cu
Dimensions (mm) 136(L) x 136(W) x 67(H) 109(L) x 109(W) x 62(H)
Dissipation Area (cm2) 3,268 3170
Base Material Copper
Weight (grams) 918 775
Thermal Resistance (°C/W) Silent Mode 0.24 Silent Mode 0.27
Normal Mode 0.19 Normal Mode 0.20
Integrated Fan 120 mm diameter, 0.35A, 12VDC 90 mm diameter, 0.23A, 12VDC
Fan speed 1,000 ~ 2,000 rpm 1,350 ~ 2,600 rpm
Fan noise 20 ~ 32 dBA, ±10% 18 ~27.5 dBA, ±10%

Here, the performance improvement at the low speed setting is a bit more significant at 0.03°C/W, with a noise penalty of +2 dBA. But in our view, the mass of 775 grams is already high; 918 grams really is best avoided.

Finally, here is a final interesting comparison: According to Zalman specs, the 7000-Cu outperforms the bigger, noisier but lighter 7700-AlCu. Just by 0.01°C/W in cooling but by 2~7 dBA in noise.

INSTALLATION

Zalman retained the same basic structure and design of the 7000 series: A large number of flat thin pieces of metal are bolted very tightly together. One portion of the joined edges is polished to serve as the base, and the other pieces bent to form a near-complete radial arrangement of fins. Two openings in the fin array allow space for the joining bolts, the central captive steel spring clip and the integrated fan. It is a unique design for which Zalman have a couple of Korean patents, and patents are pending in 20 other countries, including the US, EU and Japan.


Hard to tell it apart from a 7000.

The photo below shows clearly that the profile of the 7700 was designed to clear tall motherboard components around the CPU socket. The motherboard compatibility is probably not much worse than the smaller 7000 because the greater diameter flare happens at a good height above the motherboard.


As usual for Zalman, a superbly polished base.

The central rocker-style steel tension bar remains at the heart of the mounting system. It is a logical extension of the simple clip mechanism that has been used for CPUs for well over a decade. Where Zalman has been innovative is in developing interchangeable mating ends and mechanical interfaces for the various processor form factors.

For the Socket 478 platform, two metal bars or “yokes” are employed. These drawings from the excellent manual illustrates it well.

For the socket 775, a retention bracket system that goes above and below the CPU socket is employed. The retention brackets must be installed before the motherboard is mounted in a case. Again, drawings from the manual:

The setup for the AMD 754, 940 and 939 processors are all the same, as the same heatsink retention system is employed. A replacement backplate is supplied, but it is identical, as far as we can tell, to the ones that come stock on recent A64 motherboards, so replacement may not be necessary.

Here are photos of the 7700-AlCu shown on mounting mechanisms for the various platforms.

TESTING

The core of the test system is similar to that used in the past, but not identical.

The motherboard has been changed because the venerable Intel D845PEBT2 motherboard used for over two years finally petered out. It survived more than 200 HSF mount/dismount procedures, and probably 50 CPU changes; it’s a miracle the board lasted as long as it did. The replacement is another Intel 845 chipset motherboard which has seen only a dozen or so HSF changes. Its CPU thermal monitoring has been calibrated to be within 1°C of the Intel board it replaced.

Test Platform

  • Intel P4-2.8A The Thermal Design Power of this P4-2.8 (533 MHz bus) is 68.4 or 69.7W depending on the version. As the CPU is a demo model without normal markings, it’s not clear which version it is, so we’ll round the number off to ~69W. The Maximum Power, as calculated by CPUHeat & CPUMSR, is 79W.
  • AOpen AX4GE Max motherboard – Intel 845GE Chipset; built-in VGA. The on-die CPU thermal diode monitoring system reads 2°C too high, so all readings are compensated up by this amount.
  • Panaflo FBA08A12L1A 80mm DC fan
  • OCZ DDRAM PC-3700, 512 MB
  • Seagate Barracuda IV 40G 1-platter drive (in Smart Drive from Silicon Acoustics)
  • Seasonic Super Tornado 300 (Rev. A1)
  • Zalman Multi-Connector (ZM-MC1) and Fanmate2 voltage controller
  • Arctic Silver Ceramique Thermal Compound
  • Two-level plywood platform with foam damping feet. Motherboard on top; most other components below. Eases heatsink changes and setup.

Measurement & Analysis Tools

  • CPUBurn processor stress software
  • Motherboard Monitor 5 software to monitor CPU temperature
  • Enermax UC-A8FATR4 multifunction monitor/fan controller w/ thermal sensors
  • T.H.E. KP- 6M reference omni microphone
  • M-Audio Tampa mic preamp
  • M-Audio Firewire 410 external digital sound interface
  • B&K 2203 sound level meter
  • SPCR lab custom-built multi-channel variable DC power supply
  • Digital display multimeter
  • Digital display anemometer to measure fan airflow

The ambient temperature in the test lab was 21°C. Ambient
noise in the lab was ~20 dBA most of the time. When measuring quieter fans, all unnecessary systems were turned off so that the ambient dropped to ~18 dBA. Supplementary SPL measurements were also conducted in a quieter, larger room with a lower 16 dBA noise floor. Maximum load temperatures were recorded >20 minutes
into a CPU stress test with CPUBurn.

RESULTS

Zalman CNPS7700-AlCu
Fan Voltage (V)
Airflow (CFM)*
SPL (dBA/1m)
Load Temp (°C)
°C Rise from ambient
°C/W MP
°C/W TDP
12
32
38
39
18
0.23
0.26
9
25
32
41
20
0.25
0.29
7
18
28
43
22
0.28
0.32
5
12
22
46
25
0.32
0.36

*CFM (cubic feet per minute): Measured with the fan on the heatsink, with heatsink mounted on the CPU.
(+/- 10%)
SPL: Sound Pressure Level in dBA/1m measured with high accuracy B & K SLM at 1 meter
Load: CPUBurn for ~20 mins
°C Rise: Temperature rise above ambient at load
°C/W MP / TDP: Temperature rise per Watt, based on Maximum
Power (79W) or Thermal Design Power (69W) rating of CPU

* Why was airflow measured with the fan on the installed HS?

We wanted to make fair comparisons of this HSF against separate HS + fan combinations . Because the fan in the 7700 is frameless, when held in free air, its flow is not as directed as a normal box frame fan; hence its CFM measures very low. But both the 7700 fan and a framed fan can be measured mounted on the HS, which provides the best apples-to-apples comparison, and in-use conditions. From a performance point of view, the HS that cools best with lower airflow is the better one.

Performance at 12V is excellent, as expected. The noise measures higher than specified by Zalman, and it is louder than most SPCR readers would accept. This, too, is expected. Overall, the acoustic signature is benign. There is mostly whooshing wind turbulence and no bearing noise to speak of.

At the 9V fan drive level, the noise has dropped nicely but is still a bit loud. Performance is still excellent. Ditto at 7V, but it’s still not really quiet enough for SPCR.

It’s at under 6V that the fan noise comes down to a level most SPCR readers would accept. The 5V performance is still very good, and it is very quiet at this point. There is evidence of low frequency hum and a touch of buzzing, but inside a case with other noise producing components, it would be pretty hard to hear.

COMPARISONS

1) Against the Zalman CNPS7000-AlCu

Rather than retrieve and extrapolate data from earlier reviews, a fresh set of tests were run with the 7000 on this new test platform. The table below shows the results.

Zalman CNPS7000-AlCu
Fan Voltage (V)
Airflow (CFM)
SPL (dBA/1m)
Load Temp (°C)
°C Rise from ambient
°C/W MP
°C/W TDP
12
23
39
41
20
0.25
0.29
9
15
33
43
22
0.28
0.32
7
11
28
45
24
0.30
0.35
5
8
22
48
27
0.34
0.39

A side by side look at the most salient data of the 7000-AlCu and the 7700-AlCu:

Fan Voltage (V)
Airflow (CFM) SPL (dBA/1m) °C Rise from ambient
7700
7000
7700
7000
7700
7000
12
32
23
38
39
18
20
9
25
15
32
33
20
22
7
18
11
28
28
22
24
5
12
8
22
22
25
27

It’s surprising, but the smaller 7000 measures a tad higher with the fan at higher speeds and is essentially at the same loudness at slower speed. These findings contradict Zalman’s own specs. In listening, the smaller 7000 fans always sounds a bit higher pitched because it is spinning faster, and the larger 7700 fan has a touch more growl. Which you prefer is probably a tossup and dependent on your own hearing. Practically speaking, there’s not much acoustic difference.

In terms of cooling, the larger 7700 stays ahead by at least 2°C at the higher fan speeds, and increases the gap to 3°C at low fan speeds. This is a significant if not earth-shattering difference that will come in handy with hotter CPUs.

2) Against the Thermalright XP-120

We crowned the Thermalright XP-120 as the new king of coolers a few months ago. The XP-120 is the only other 120mm fan CPU cooler we know of and from a brand that’s been neck-and-neck with the Zalman for top HS honors for 2~3 years, so the comparison is a natural.

Again, rather than rely on previous data — ably collected in the XP-120 review by Ralf Hutter but on a different platform and a different CPU — a new set of tests was conducted with the XP-120 on this platform.

There are many 120mm fans that can be used with the XP-120. In order to keep this review from ballooning into an unwieldy tome, only one fan was used: The Globe S1202512L-3M, the quietest 120mm fan employed by Ralf Hutter in his XP-120 review. This sleeve-bearing fan is thermally controlled via a thermistor at the end of a long lead that emerges from the hub. The thermistor was shorted to take the thermistor out of the circuit so that the fan speed could be controlled directly.

The Globe fan was placed atop the XP-120, blowing down only. The voltage to the fan was carefully adjusted to give the same CFM values measured at 12, 9, 7 and 5V with the Zalman 7700. In this way we’re comparing the two heatsinks with very similar amounts of airflow.

Thermalright XP-120 w/ Globe S1202512L-3M fan
Fan Voltage (V)
Airflow (CFM)*
SPL (dBA/1m)
Load Temp (°C)
°C Rise from ambient
°C/W MP
°C/W TDP
8
32
27
38
17
0.22
0.25
6.8
25
23
40
19
0.24
0.28
5.4
18
20
42
21
0.27
0.30
4.5
12
18
45
24
0.30
0.35
12*
47
39
36
15
0.19
0.22

The airflow / noise ratio of the 7700 fan and the Globe fan are quite different. There is no way that the Zalman 7700 fan can match the maximum airflow of the Globe fan at 12V. (*The 12V results in the table above are listed at the bottom and in gray to show that there is no comparable data for the Zalman 7700-AlCu. Note that this noise level is about the same as with the 7700 at 12V, but the temperature is 3°C lower.)

Below, then, is a side-by-side comparison of the core data for the Zalman 7700AlCu and the Thermalright XP-120 with Globe 120mm low speed fan.

Airflow (CFM)* Fan Voltage (V) SPL (dBA/1m) °C Rise from ambient
7700
XP-120 / Globe
7700
XP-120 / Globe
7700
XP-120 / Globe
32
12
8
38
27
18
17
25
9
6.8
32
23
20
19
18
7
5.4
28
20
22
21
12
5
4.5
22
18
25
24

Airflow is the reference for the table above; the voltage for each fan was set so that the airflow would be the same for both HSF. The Globe fan requires less voltage to provide the same airflow level. At the same airflow level, it also makes much less noise. The difference ranges from over 10 dBA/1m at the high end to 4 dBA/1m at the low end. These are huge differences.

The Thermalright XP-120 makes better use of the airflow, besting the Zalman by a single degree. The difference is probably insignificant, but it is consistent across the range of airflow.

If the data comparison was changed so that the noise was the constant rather than the airflow, then we’d see the XP-120 running up to 5~6°C cooler than the Zalman at the same noise level. No matter how you look at it, the fan in the Z7700 is not as quiet as it could be.

3) Audio Recordings in MP3 Format

Zalman 7700 – 12V – 38 dBA/1m

Zalman 7700 – 9V – 32 dBA/1m

Zalman 7700 – 7V – 28 dBA/1m

Zalman 7700 – 5V – 22 dBA/1m

Zalman 7700 – Nexus 120 – 12V – 19 dBA/1m *

Thermalright XP-120 / Globe fan – 12V – 36 dBA/1m

Thermalright XP-120 / Globe fan – 8V – 27 dBA/1m

Thermalright XP-120 / Globe fan – 6.8V – 23 dBA/1m

Thermalright XP-120 / Globe fan – 5.4V – 20 dBA/1m

Thermalright XP-120 / Globe fan – 4.5V – 18 dBA/1m

Zalman 7000 – 12V – 39 dBA/1m

Zalman 7000 – 7V – 28 dBA/1m

Zalman 7000 – 5V – 22 dBA/1m

* See text box at bottom of page: 7700 Fan Swapping

HOW TO LISTEN & COMPARE

These recordings were made with a high
resolution studio quality digital recording system. The microphone was 3″ from
the edge of the fan frame at a 45° angle, facing the intake side of the fan to
avoid direct wind noise. The ambient noise during all recordings was 18 dBA or
lower. It is best to download the sound files to your computer before listening.

To set the volume to a realistic level (similar to the original), try playing this Nexus 92mm case fan @ 5V (17 dBA/1m)
recording and set the volume so that it is barely audible. Then don’t reset the volume and play the other sound files. Of course, tone controls or other effects should all be turned off or set to neutral. For full details on how to calibrate your sound system to get the most
valid listening comparison, please see the yellow text box entitled Listen to
the Fans
on page four of the article
SPCR’s Test / Sound Lab: A Short Tour.

CONCLUSIONS

The Zalman 7700-AlCu provides superior cooling performance than its predecessor, the 7000-AlCu, which still gives it a pretty good run for the money. The small but measurable cooling gain is obtained with no serious acoustic price, as the fans on these coolers measure almost identically for SPL. Given that the 7000 remains among the most popular of high performance, low noise HSF, a product that surpasses it has to be praised. They do sound a bit different, but whether one sounds louder than the other is difficult to judge. We suspect that with the Fan Mate set to minimum (5V) on both HSF, the 7700 would probably sound a touch louder for most listeners.

This noise level cannot be described as superbly quiet, however, even at the minimum 5V setting of the Fan Mate 2 voltage controller. We commented on this in our original 7000-AlCu 18 months ago, and we hoped that the 120mm fan would give us lower noise. The 7700 fan has not done that, though it has give us a slightly better noise-to-airflow ratio. It is possible that our Zalman HSF sample(s) were on the noisy side of their ±10% noise tolerance range, but we’ve had more than half a dozen 7000 series HSF through the lab, and they’ve all sounded pretty much the same.

The Thermalright XP-120, which is the only other 120mm fan CPU cooler at this time, is a tougher competitor to the 7700. Its multiple heatpipes and unusual design, combined with a separate conventional 120mm frame fan provide slightly better cooling than the 7700-AlCu with the same airflow, and at a much lower noise level. If we had set noise as the reference in our comparison chart (i.e., kept the noise constant rather than the airflow), the cooling advantage of the XP-120 would have been much greater than 2°C; it would have widened to as much as 4~5°C. However, that would simply have shown the superior noise-to-airflow ratio of the Globe fan over the built-in 120mm fan in the Zalman cooler.

The difference between the Globe fan and the integrated fan in the 7700 underline our point about the noise performance of the Zalman fans. Oddly, for a company that has staked its reputation on low noise cooling, Zalman fans have never been that quiet. Admittedly, this is related partly to Zalman’s use of ball-bearing fans, which is almost mandatory in high heat applications such as power supplies and heatsinks.

Sleeve bearings rely on oil lubrication that tends to dry up sooner when subject to higher temperatures. When they seize, they do so quietly, without warning. Ball bearing fans last longer in heat, and when they go, they start screeching, getting louder and louder as they get worse. This is a kind of built-in safety warning for the PC user to explore what is going on, much like the screeching of worn brakes. For these reasons, ball bearings are almost mandatory in stock fans on heatsinks; most system integrators and OEMs do not accept sleeve bearings in HS or PSU fans. But they ignore the fact that ball-bearing fans are more easily damaged by physical shock during shipping. In practice, a good sleeve bearing or modified sleeve bearing such as the Hypro bearing in Panaflo fans lasts a long time even on a hot heatsink. (I’ve never seen a failure of any Panaflo due to heat damage in nearly four years with well over 100 units passing through my hands.)

Getting back to the 7700-AlCu, while it may not be the quietest or highest performing HSF in the market today, it is right up at the top, and the street pricing is only marginally higher than for the 7000-AlCu, which has always been a good buy. The XP-120 + 120mm fan is definitely more expensive, and because of the very high tension of the mounting system, actually more difficult to install than a Zalman 7000 series HSF. The fan integration is something a lot of people prefer for simplicity. And if the performance is good enough, and it fits in your system but you’d prefer lower noise, you can always swap the fan, as some SPCR forum members have done on the 7000 series. (See text box below: 7700 Fan Swapping)

As with all recent Zalman HSF packages, we have nothing to fault with this one (except the use of the too-common, environmentally unfriendly plastic). The new Fan Mate 2 is a genuine improvement, and the mounting hardware for the various processor types is well thought out.

Pros

* Excellent cooling performance
* 22 dBA/1m min noise is very good
* Great accessories package
* Excellent cross-platform compatibility
* “Tried and true” radial design
* Weight reasonable
* Pretty easy to install
* Probably better compatibility than XP-120
* Well priced for performance & package
* CU (all-copper) model
probably performans better

Cons

* Still not as quiet as we’d like
* Outperformed by XP-120

* Probably too big for some systems
* CU model is too heavy

Much thanks to Sharka Computers
for the Zalman CNPS7700-AlCu sample.

* * *

7700 Fan Swapping

SPCR Forum members reported success in replacing the fan in Zalman 7000 series coolers for a quieter 92mm fan, such as Joseph Clemente did in this thread: Zalman 7000A AlCu modded w/ CHROME FAN! SPCR reviewer Ralf Hutter also modded a Z7000 — with a Nexus 92 fan. His technique involved replication of the screw mounts and a layer of silicone damping.

Regular readers know that the Nexus 120 has become a SPCR reference fan for its extremely quiet performance. Sean Boyd, who worked on the thermal testing, decided to see whether he could perform a fan transplant on the 7700. The answer is yes, he could.


A “Boydstein” creation, but far from being a monster.

It was a quick mod for which readily available materials were used, namely, double-sided adhesive tape. Two layers about 1/8″ thick in total were needed to raise the Nexus fan blades high enough above the fins. This is not ideal for any application in which the fan faces any direction other than straight up as shown in the photo above. You would not want to trust the tape to keep the fan secure on a vertically mounted motherboard. Sean is looking into high temperature double sided tape from 3M rated for high adhesion at 150°F long term.

One obvious acoustic benefit of the dual layer of double-sided tape was the mechanical decoupling effect. It was easy to hear that not only is the Nexus fan far quieter than the stock Zalman fan, in this application, it is quieter than when within its box frame. Even plugged straight into 12V, the SPL measured just 19 dBA/1m. The results of thermal testing with this mod:

MP3: Zalman 7700 – Nexus 120 – 12V – 19 dBA/1m

Fan @ 12V * 14 CFM airflow (measured on HS) * ~23°C rise from ambient * 19 dBA/1m

Comment on this article in our Forums.

Silent PC Review is reader-supported. When you buy through links on our site, we may earn an affiliate commission. Learn More

Leave a Comment

Your email address will not be published. Required fields are marked *