Zalman CNPS9500 LED

[Operandi]

LINK

 

[JEDIYoda]

This review and the SPCR review seem to go hand in hand except for what the SPCR review said about the noise!!

As I stated in other posts I had the feeling the zalman would be one of the best coolers but possibly not the best mainly IMO due to the size of the fan.
It would be interesting to see what a reworked 9500 with a 120 mm fan would be capable of!
Who knows that might happen.

Then again there can be no arguing IMO what the last patagraph of the Frosty tech review stated...which was even stated in the SPCR review in a different way!!

SPCR--
It's safe to say that the Zalman 9500 is the most efficient of all heatsinks that use a 92mm fan. The basic design is brilliant, but it cannot completely escape the limitation of fan size.

Frosty Tech--
Ultimately, the Zalman CNPS9500 LED is a very good low noise heatsink for any AMD Athlon64 (socket 754/939/940) processor. The heatsink is foremost quiet, and offers among the best balance of noise and thermal levels that we've seen to date. There are cooler heatsinks which are much louder, a very few heatsinks which are much quieter, but no heatsinks which are quieter and cooler.

Again....it all boils down to personnal preference....
I myself ordered a xp120....
I also ordered the Zalman 9500....
I might also order the Scythe Ninja.....
Then again for me part of the fun is trying different products as well as having a nice looking PC that runs cool and somewhat quiet.......

Still a Zalman boy

Anyways just my 2 cookies worth!

 

[BonzaiDuck]

At least this review was more honest. They said the CNPS-9500 was a good "low-noise" cooler, but did not perform as well as some others which were higher in the dBA ratings. This again confirms my comparison between the SPCR, Hartman.de and one other review.

Here, it is a matter not only of preference, but of philosophy.

Noise is a nuisance that can be dealt with in other ways besides limiting your CPU cooling potential to low noise (low CFM) fans. It can be dealt with even without the extensive use of noise-deadening pads, which may compromise cooling solutions that are dependent on the use of a case's heatsink potential. Any third-grader can design a tape-on (I'd probably use velcro) foam-board "cage" for a CPU fan that blocks transmission of sound to the case side-panel while allowing full-airflow.

This forum gives many people ideas in various categories: (1) what to buy, (2) how to modify, (3) how what you buy can be modified to be better.

As a matter of "philosophy," I would think if you are going to build the computer yourself, you shouldn't mind making DIY homegrown improvements, and if you are going to over-clock your computer, you shouldn't mind doing either.

 

[Operandi]

Originally posted by: BonzaiDuck
At least this review was more honest. They said the CNPS-9500 was a good "low-noise" cooler, but did not perform as well as some others which were higher in the dBA ratings. This again confirms my comparison between the SPCR, Hartman.de and one other review.

They also say the Zalman is unmatched in terms of noise to performance. If your system really is below the noise floor of the 9500 a fan swap should bring the noise in level in with the quietest XP-120s with similar or even better performance. LED fans bug me anyway.

 

[BonzaiDuck]

I agree about LED fans. I'm only now starting to "dress up" my MOJO, after experimenting with a side-panel blow-hole that is no longer needed. I won't replace a good conventional fan to compromise either CFM or low noise for the glitz, even so.

The Hartware.de review ran the CNPS-9500's fan to the max and compared it to a similarly revved XP120. The XP120 still showed a slightly lower load temp in this maximum range. Since the SI-120 trumps the XP-120 by a few more degrees Celsius (and certainly by 0.027 improvement [reduction] in thermal resistance, it looks like ThermalRight has proven for itself Zalman's own boast of matching or trumping some water-cooling kits.

It may be that the maximum cooling (minimum thermal resistance) is unimportant to those seeking the quiet life. But if you're going to over-clock or run a processor leaking more than 100W of thermal energy, you'd like the CPU fan to at least spin up at some temperature and push enough air to bring the thermal resistance of a cooler to its minimum, and if that minimum is lower than our topic of concern -- well --

I was thinking I might swap out my XP120 for the CNPS-9500 if Zalman proved its advertised hype about the water-cooling. But the reviews of the SI-120 put it in the range of three Swiftech wc kits. Unless I go forward with my Rube Goldberg five-radiator double-loop ice-chest evaporative cooler idea, I'm going to install an SI-120 anyway. There is a method to my madness, anticipating an upgrade to my brother's old P4-1.8A Ghz (400Mhz FSB) system at Xmas to double his processor speed and FSB. He'll get the XP120, and he won't need to over-clock. I get the "core return" for a file-server replacement of an old P3.

Anyway, if I went with the Zalman, I'd have to redesign my mobo duct. The review data just in tips me in the other direction . . .

 

[GalvanizedYankee]

A 7000al-cu can be had easily for $28 now. Next year I might buy a 9500 for<$40

Concerning the fan. There is a tutorial at spcr on cutting the housing off a fan of choice
and installing it on a Zalman CPU heatsink.

I really don't need the max cooling capacity. The build quality of Zalman's products is what
matters to me. Weight is also a concern.



GY

 

[Corporate Thug]

http://www.quietpcusa.com/acb/showdetl.cfm?&DID=8&Product_ID=266&CATID=3

 

[BonzaiDuck]

OK. Here it is. I believe in scientific objectivity. There may be some factors here that reverse the scores, but the available information seems sufficient to calculate a very likely thermal resistance value for the Zalman.

I base the estimate on previous reviews I've read which benchtest the Zalman CNPS-7000-Cu at a thermal resistance value of 0.21C/W. The AlCu model proves at about 0.23C/W or 0.24C/W, as I recollect.

The table of comparisons in the FrostyTech.com Frosty Tech Review shows the following "Increase over Room Ambient" for the CPU temperature in Celsius for both the CNPS-9500-LED and the CNPS-7000a-Cu coolers. Among several observations where fans are set to either "Low" or "High," we focus attention on each respective cooler's "High" setting, knowing that the "High" setting for the CNPS-7000-Cu yields a thermal resistance of 0.21.

The thermal power, constant for all tests, is 125W. Even though the table shows "increase over room ambient" in Celsius, it is not the room ambient that is relevant to thermal resistance, but the idle and load CPU values, or the values at the surface of each respective heatsink where it contacts the testing device. Further, even if room ambient is not controlled here, since we are dealing with differences between load and room ambient temperatures in either case rather than absolute load temperature values, we can subtract cooler A's increase-over-ambient from cooler B's increase-over-ambient to show what the difference in load temperatures would be if we used a controlled experiment -- therefore, controlling for room ambient is not crucial to the calculation.

First, find the delta-T or difference between idle and load temperatures for the Zalman CNPS-7000-Cu:

TR(z7000) = 0.21 C/W = delta_T / 125W

and delta_T[z7000] = 0.21C/W * 125W = 26.25 C degrees

The difference between the increases over room ambient for the two coolers is about 6.5C degrees, therefore the Thermal Resistance under load with fan at "High" for the Zalman CNPS-9500-LED is:

TR[z9500] = (delta_T[z7000] - 6.5C) / 125W
TR[z9500] = (26.25C - 6.5C) / 125W = 0.158 C/W
Now . . . . maybe there is some source of error, or maybe the CFM with the fan on "high" for the CNPS-9500 does not produce the very minimum thermal resistance for which the unit is capable. Somebody may think of something, but the issue is simply how reasonable the alternative explanation, and the probability that it is the operative explanation.

But it seems that the exclusion of the ThermalRight XP-120 but more significantly the exclusion of the ThermalRight SI-120 reduces the chance that somebody will see the Zalman CNPS-9500 as "second best."

We have Citarella's (?) article at OverClockers.com for August 27, 2005, showing a thermal resistance of 0.14 C/W for the ThermalRight SI-120, and the results of as review for the SI-120 at the German site Hartware.com is completely consistent with the differences between the thermal resistance of the SI-120 and its older sibling the XP-120 (0.167 C/W).

The ThermalRight SI-120 is the more effective cooler by a difference of 0.018 C/W in thermal resistance. The Zalman 9500 is actually outside the range of the least effective 120mm radiator water-cooling kit made by Swiftech and sold by SidewinderComputers.com, which shows thermal resistance of 0.15 C/W. The ThermalRight SI-120 is actually more effective than that Swiftech cooler, and for a heatpipe cooler gives a very good showing just above the middle of the range for the three Swiftech water-coolers which range from 0.125 C/W to 0.15 C/W.

While that's good enough for me, I already ordered my SI-120 based on information available to me last week, and I expect to see FedEx on my doorstep tomorrow.

 

[JEDIYoda]

My XP120 arrived today!!!

 

[BonzaiDuck]

Yo, JediYoda!

I have an earlier P4P800 (standard) mobo.

Who manufactures your PC3500 memory modules?

I will replace my P4P800 mobo with a P4P800 SE. Except for a different on-board sound system, the P4P800 SE is identical to the P4P800-E, less the (Promise Tech?) onboard RAID controller in addition to ICH5R.

I notice that the BIOS now contains memory options for DDR500 modules and higher.

Have you over-clocked your system? I'd be interested in the manufacturer info and your choice of settings for memory.

 

[JEDIYoda]

I am very happy with my OCZ Extreme Performance Series PC-3500!! I never did like heat spreaders but I love the memory!

Yes I have o/c`d this particular system!

I have had it stable at 3.8!!!
But it ran very very hot under load or while playing Quake4...etc....
Thats another reason I have so many fans in my case...lol

Right now I am running stock until I get this xp120 installed!!

I am thinking about doing way with all my case fans and replacing them with 120mm!
That is getitng rid of the 80mm!!

Anywyas I getting ready to build an AMD rig specifically for gaming.

Thus possibly the xp120 will be transfered at a later date!

I`m still getting the zalman 9500...even though I am almost positive the xp120 will be my choice for the gaming rig!

Unless I go water cooling......

Have a nice day!!

 

[BonzaiDuck]

Any of those coolers should work fine, especially with an AMD processor, but certainly with a 3.2EE.

I took a close look at the Aspire X Dreamer case. On the "up" side, it has a marvelous rear-panel 120mm fan-hole. Back to that in a minute.

The front-side is a mixed bag. there is room between the chassis and drive-cage for -- possibly -- either a 92mm or 120mm fan. You may want to "Dremel some extra ventilation" there, and as you say, swap out the 80s for either one (possibly two) 92s or a 120.

Back to the "up" side. You can duct the motherboard with Lexan or foam-art-board. Since this case has much "glitz," make the prototype of foam-art-board (2' x 3' @ $6), then use the pieces after testing to trace (scratch) the cut-outs from a piece of 1/8" Lexan. That saves your "glitz" with the LED lights etc. You can secure either the duct-parts made of foam-board or Lexan with little tabs of Velcro -- so you can pry them apart with a credit-card for the occasional maintenance or upgrade removal.