Does your Zalman 9500 suck big arse?

[s1rrah]

...
...

(i've posted this to several other forums, and it's quite an older mod by myself, but a recent thread discussing the 9500 prompted me to post it here too)

...
...

test machine
---------------
asus p5B-deluxe motherboard
core 2 duo 6400 running @ 3.5ghz (stock is 2.13ghz)
1gig dual channel mushkin extreme ddr2-6400 (stock settings in bios)
zalman 9500 heatsink (testing two different 120mm fans)


test fan data
---------------

(in the pics, i have obviously cut and trimmed the man fan motor/hub out of the typically square assembly; this is necessary so you can attach it to the zalman mounting arms)


global win (aka "stubby")
--
model: BP1202512m
Airflow: 74.5 CFM MAX
Dimensions: 120 x 120 x 25 mm
Speed: 2500 RPM max
Noise Level: 45 dBA at max
--




silverstone
--
model: FM121
Airflow: 110.03CFM MAX
Dimension - 120 x 120 x 25mm
Speed - 2400rpm max
Noise Level: 39.5 dBA at max
--



The test
--------------------

Test is running two instances of dual prime95 (one for each processor core), for six hours and with the test setting "in-place large FFT's", the setting in dual prime95 that i've found generates the most heat in the shortest amount of time.

Temps are achieved via motherboard monitor and Asus's provided AI Suite tools.

I'm going to do the global win image and stats first as it's been tested dual prime95 for six hours already; i'll post the silverstone image since i just finished mounting it but won't have stats available until tomorrow.

here's a pic of the global win fan mounted and in the case:





globalwin - *RESULTS*
six hours dual prime95 - in-place large FFT's
core2 duo 6400 2.13ghz running @ 3.5 ghz
-------------------------------------------
cpu idle: 41-42
cpu load: 60-61
mb idle: 32
mb load: 35-36

(this is about a 8 or 9 degree decrease over the stock zalman fan)



and now for the silverstone fan, ....

this is a pic of the silverstone FM121 mounted on the zalman heatsink:



and another (fan blades 1/4 inch from RAM!):




Silverstone FM121 - *RESULTS*
six hours dual prime95 - in-place large FFT's
core2 duo 6400 2.13ghz running @ 3.5 ghz
-------------------------------------------
cpu idle: 41-42
cpu load: 58-60
mb idle: 32
mb load: 34-36

glad i tested the silverstone last since it beat the other fan by about an average of 2 degrees.

I'm going to leave it mounted.

the main noticeable improvement, however was in how incredibly fast the fan brought the temps down to normal idle temps after stopping the dual prime test ... just 3 or 4 seconds and it was back down to 42 or so at idle; anyone with a zalman 9500 should change they're fan immediately if they're interested in decent air cooled clocks.

now i have to see how much more speed i can get out of it before hitting 65 absolute max load temps.

here's a link with details on how to switch fans; note ... you will have to do some careful bending of the zalman mounting arms to get the 120mm fan to line up vertical with the fins ... get some needle nose pliers and take your time.


http://www.silentpcreview.com/article284-page1.html

do it.

[Chewbenator]

Is it possible to put another fan close behind in a push-pull config? Or, does the out-take fan on the case provide that funtion.

[s1rrah]

Is it possible to put another fan close behind in a push-pull config? Or, does the out-take fan on the case provide that funtion.

if you had the time and inclination you could modify the mounting base and fabricate another fan mounting arm to replicate the mount on the opposite side ... but in my opinion, the 9500 wouldn't really benefit from a push > pull > setup.

the push > pull > setup works well on coolers like the noctua and especially the infinity because they have a *very* dense fin structure that makes it hard for a single low RPM fan to push through; adding the pull > fan dramatically improves the flow on coolers with very dense (tightly packed) cooling fins; the infinity is an especially good example of a cooler with almost too dense of a fin assembly when using a single fan.

the 9500's fin assembly is quite 'breathy', ... in that they have rather decent space between the fins and air passes through the entire assembly quite easily.

[Omastar]

I was never a big believer in the 9500s and 9700s. Scythe Infinity, Tuniq Tower, CM GeminII stomp a mudhole in Zalman's ass in this arena, anyway.

[s1rrah]

I was never a big believer in the 9500s and 9700s. Scythe Infinity, Tuniq Tower, CM GeminII stomp a mudhole in Zalman's ass in this arena, anyway.

which is *exactly* why i'm now using an infinity with the pressure mod; it not only stomps a SINKhole in the stock 9500, it handily throttled the modified 9500 as well. here's a link to a review of both coolers using the same FM121 fan: Infinity OWNs the Modified Zalman

...

couldn't be more pleased with the infinity; won't upgrade until i see the test results of the new thermalright coolers (IFX-14, etc.).

...

[Chewbenator]

Thanks for the response and explanation. Looks like it provides some additional cooling for the memory being so close.

[Cupcake]

my zalman rock already

[verdun]

I mean, onto the heat pipes, or are they just press fit?

[bing]

Zalman HSFs aint heat sink, they're just made for decorating your CPU, get a decent one like Ultra 120, Tuniq or Infinity !

[Omastar]

Zalman HSFs aint heat sink, they're just made for decorating your CPU, get a decent one like Ultra 120, Tuniq or Infinity !

Well the Zalman 7xxx series was actually pretty good. Great for cool running single cores (had a 7700 for my A64 3200+, seen them used on Pentium Ms, as well).