TRUE Copper vs. ZALMAN CNPS9700 LED mini test-Pic heavy!

[SamHughe]

I recently purchased and installed Thermalright TRUE Copper Edition so I wanna share my experience with this particular heatsink compared to my old heatsink (Zalman CNPS9700 LED).
First let’s get the know the contenders little bit better.
ZALMAN CNPS9700 LED
Fan: 110mm, 1250-2800 RPM, 19.5-35 dBA
Material: Copper
Dimensions: 90 x 124 x 142mm
Weight Heatsink: 690g, Total (including fan): 764g
Price $61.99 @ newegg

TRUE -Cu
Fan: N/A (test fan: 120mm 840rpm, 8.7 dBA)
Material: 100% copper construction
Dimensions: 133 x 156 x 38 mm
Weight: 1900g (Son of a…!)
Price: $109 @ Frozen CPU

TRUE-Cu came in a nice box, securely packaged.
http://img528.imageshack.us/img528/3407/00018df7.jpg
http://img255.imageshack.us/img255/6135/00014mt2.jpg

Couple of pics with my crappy point-and-shoot camera.
http://img185.imageshack.us/img185/863/00005dq0.jpg
http://img521.imageshack.us/img521/8199/00003pw9.jpg
http://img404.imageshack.us/img404/2092/00004fp5.jpg

One of the things about TRUE-Cu is it comes with a horrible base! I guess Thermalright figured most users would lap this heatsink anyway so they didn’t bother.
http://img185.imageshack.us/img185/855/00012nv9.jpg

On the other hand, look at Zalman! And this is right after I pulled it from the PC.
http://img255.imageshack.us/img255/6113/00truecopper014bn2.jpg

This is how it looks after I worked on TRUE-Cu couple of minutes (600-1500-2000 grits)
http://img521.imageshack.us/img521/1181/00truecopper004wq3.jpg

The system, before I installed TRUE-Cu looked like this. It’s an old school MM-UFO case. Putting pictures here to give you an idea about the airflow and such.
http://img404.imageshack.us/img404/4646/00truecopper011cv3.jpg
http://img185.imageshack.us/img185/9448/00truecopper012iq1.jpg

The 775 mobo kit made of metal and overall the manufacturing quality is superb (compared to Zalman’s plastic kit). TRUE-Cu doesn't come with a 1366 kit :mad:
http://img521.imageshack.us/img521/6164/00truecopper017ax7.jpg

Yes! You are seeing right. 800rpm, 33.5 cfm, 8.7dBA. That’s all I’m going to use to cool my Q6700. Let’s throw a real challenge and see if TRUE-Cu is up to the hype.
http://img404.imageshack.us/img404/9408/00truecopper019gl9.jpg

This is how I put my paste AS5). A line in the middle, little dots on the corners. I’m letting you know so you can make total sense of the results.
http://img185.imageshack.us/img185/9723/00truecopper021em9.jpg

Here’s how it looks installed in the system
http://img255.imageshack.us/img255/9725/00truecopper023qs1.jpg
http://img528.imageshack.us/img528/9757/00truecopper031pn9.jpg
Now this thing is crazy heavy. Notice the bow?
http://img135.imageshack.us/img135/5969/00truecopper035nk4.jpg

Now the test:

The test Setup:
Q6700 @ stock 2.66
ABIT IN9 32X-MAX
8GB Muskin DDR 800
eVGA 8800GTX SLI
4x VelociRaptors (Raid 0) + 1 150gbRaptor JBOD (Page File)
MM-UFO
Corsair HX1000W
X-FI Titanium
Vista Ultimate 64

Measurement software: Core Temp 0.99.3

Cold start: Cold start temps were measured immediately after computer was booted.
During Zalman tests the ambient temp was 83F/28.3C
During TRUE-Cu tests the ambient temp was 79F/26.1C

Light use: Light use temps were measured after computer was on at least 3 hours. Typical desktop tasks like web surfing, word processing, music play with Windows Media Player were performed (no virus scan, hd defrag, or movie encoding/decoding)

Heavy use: Heavy use temps were measured after at least 2 hours of Fallout 3 gameplay. Game settings were 1920x1200, full AA, AAF, everything max. At about two hours in gameplay, the game was alt-tabbed and the tempetures reported by CoreTemp was recorded for both heatsinks.

Now here are the results
http://img404.imageshack.us/img404/6290/tablelx8.jpg
(Results are based on the temps on the hottest core –see the pics below)
Zalman:
Cold start: 33C, light use: 43C, heavy use 49C

TRUE-Cu:
Cold start: 31C, light use: 37C, heavy use 48C

Below: Zalman -> light use (sorry couldn’t get the SS for the heavy use but it was 49C)
http://img521.imageshack.us/img521/7460/idleactivewindowks5.jpg

TRUE-Cu -> Light use
http://img185.imageshack.us/img185/8156/trueidleactivewindowtk4.jpg

TRUE-CU-> Heavy use
http://img255.imageshack.us/img255/1738/trueload2hrsfalloutactipv5.jpg

Another thing I noticed about TRUE-Cu is how well and quickly it restores it’s cooling power after stress. The below SS was taken only 15 secs after the above indicating the fan is definitely not enough.

http://img528.imageshack.us/img528/8333/trueload10secslateractiqo8.jpg

My conclusions:
I think TRUE-Cu performed impressively considering the fact that there is only one very weak fan attached to it. Based on the performance I see so far, it is almost certain that with two decent fans in push-pull configuration this heatsink will be a killer overclocker.

On the negative side:
-The original base was horrible. I don’t think you can use this HS without lapping which is ridiculous given the price.

-Boy this thing is heavy! I always used big, heavy heatsinks and never give a rat’s ass about weight warnings but this one is almost excessive. You have to keep your mobo vertical if you wanna sleep right at night. I’m seriously considering changing the case for this.

-I don’t think these results enough to justify the $100+ price tag. Though I am sure with two high CFM fan configuration, you can overclock the :banana::banana::banana::banana: out of your cpu!

-Speaking of overclocking. I apologize not including any overclocked results. There seems to be something wrong with my system. I haven’t been able to raise my FSB even by 1 since I got q6700 (regardless of the heatsink). It’s my mobo, ram or the both. As soon as I solve that problem, I’ll update this post. Meanwhile any help will greatly be appreciated. I will also post results with two fans. Heck, I even got a 220CFM Delta GHE fan; I’ll slap it just for fun and let you guys know.

There you go. I tried to be as detailed as possible giving you all the factors involved in my testing process so you can draw your own conclusions from it. Any feedback (positive or negative) is welcome.

UPDATE*** UPDATE*** UPDATE*** UPDATE

This afternoon I installed two 1600rpm fans in push/pull configuration.
http://img243.imageshack.us/img243/3572/000turetest006qy8.jpg

I bet with these two fans the total weight hit the big 2.0!:eek:
http://img136.imageshack.us/img136/5877/000turetest001nd6.jpg

Here are the results:
Ambient temp: 80F/26.6C
Cold start: 32C
Light load: 33C
Heavy load: 43C
Updated comparison chart
http://img136.imageshack.us/img136/7715/updatedtesthg8.jpg

Update ***** Update*******Update


Allright, I run Prime95 for about 6 hrs last night with the dual 1600rpm fans in push/pull config. This is what I got:

http://img392.imageshack.us/img392/2042/primeload1smalltz5.jpg

Then I installed the Delta fan this morning and run another 6 hrs of Prime95

http://img392.imageshack.us/img392/5777/000deltatest003ws5.jpg

I couldn't do my usual testing because of well...this:

http://www.youtube.com/watch?v=RyMUddH0qzY

So I just setup Prime95 and got out of home came back 6 hrs later. So here's the result:

http://img262.imageshack.us/img262/14/truedeltaloadsmallxe8.jpg

I don't know why there is a 7-8C difference between core temp and real temp and I don't know why there is a temp difference between the cores? You guys be the judge.

That basically concludes this review. I put one 1600rpm fan back and keep the system that way until I figure out a way to run the cpu @ a stable 3.2ghz.

I think the most impressive results so far are the ones with the 800rpm fan. This mini review should not be used to judge this heatsink's performance by any means. I'm sure TRUE-Cu will work wonders at the hands of some more overclock-savy members here in the xtreme systems. To me these results mean one thing: I can run my cpu with a 800rpm quite fan and get away with daily use and gaming at stock speeds.

[vengance_01]

Well lets raise the heat levels and see the differences.

[RADCOM]

Nice little home review thanks for sharing. You look ike you have a copper farm there :)

[speedfreak86]

I think that is way too much thermal grease. I think that both would perform better with less, however, I am happy that you are seeing good results.

[Simps]

Nice review! Congrats on the new TRUEcu!
If you can't o/c the cpu, you should try to raise the vcore, to raise the heat levels... It would be nice to see the difference between both coolers, with a quad @ 70-75C under load (running some prime95 small ffts) temps... You should try 1.45-1.5v on that cpu and see how both coolers perform...

[nullface]

I think that is way too much thermal grease. I think that both would perform better with less, however, I am happy that you are seeing good results.

My first thoug too, that is WAY to much!

[Slovnaft]

ya dood, pump the volts and use IBT for your load test, then we might see true cu pull away.

[Vapor]

Not with an S-Flex D you won't....the amount of power needed to overload the 9700 is well past what a TRUE/TRUECu can move with a single 800RPM fan. :-/

Good to see it beat the 9700 with such little airflow though :up:

[Spawne32]

"heavy use" should be prime95 for 4 hours, and thermalright bases always look like crap lol

[Simps]

Vapor,

What would be your choice, if you had to choose a Panaflo or a S-flex? (considering both with the same cmf / noise ratio and rest of specs)

[]'s
Simps

[Vapor]

Whichever gives the best performance at my desired noise level :shrug:

CFM/dBA noise ratio is useless number.

[illidan]

s-flex f (1600rpm) will do just fine :)

[SamHughe]

s-flex f (1600rpm) will do just fine :)

I've got two of those. I'll slap them on and run the same tests to see what kind of difference they'll make.

I will also try increasing vcore and running prime 95 but only after I replicate my current tests with different fans.

[Metroid]

..................

If is that possible, could you run all the cores at 100% load using both the Zalman 9700NT and the True Copper?

If you do that, we will be able to determine what is really about being a true copper cooler.

I do have a Zalman 9700 and I think the True Copper is ages better than the Zalman 9700 when all the 4 cores are at full load (100%).

I7 is coming and I'm about to jump in but I think my Zalman 9700 will not do a fine job i7 at 4.0 Ghz.

If is that possible for you to make such test. I would be very grateful :)

Thanks :)

[BigJoe]

Nice test :wave:
PS
that will be cool if you can test standard TRUE ;)

[celemine1Gig]

The best idea for using this heatsink 24/7 would probably be to drill holes (parallel to the backplate) into the mainboard tray and bolt it to it to eliminiate the extreme leverage on the mainboard itself. Someone with the sink should try that. ;)

[fatguy1992]

Nice test :)

I want to see what it could do with 2x 220CFM delta's :p: - Anyone tested this?

[xsbb]

Thanks for sharing the results with us!

Could you run IntelBurnTest and let us know how hot it gets?

http://www.xtremesystems.org/forums/showthread.php?t=197835

[Simps]

After some tests here with a regular TRUE, I am pretty much convinced that once you raise your CPU heat, the TRUEcu will beat the Zalman by at least, say some ~8C on full load temps... I just got this temps, with a regular TRUE, using a 65nm QX6700 B3 (that heats much more then your Q6700 G0), overclocked to 3.2GHz (400x8) @ 1.3V, under 2h prime95 small ffts. Ambient temps were high, arround 27-28C. Using real temp with 100C TjMAX.

http://img504.imageshack.us/img504/2604/temps1gi4.jpg

[SamHughe]

I updated the original post with the 2fan config results. I tried to replicate my original test conditions and criteria as much as possible. There is one more test I wanna do with my un-xtreme (average joe) testing methods which is the Delta Fan test. After that I'll try to replicate Simps' test using prime95 and real temp for the measurements.

[Simps]

Nice man.
Can't wait for the results!

[fatguy1992]

What delta fan are u going 2 use?

[Ar3s]

didn't seem to break the motherboard but i understand your worries. I wanna get this the copper just looks great !

[SamHughe]

I updated the original post with some 220 CFM Delta fan results.

[fatguy1992]

WTF that delta fan didn't help?

Also the reason that there is a temp difference between real temp and core temps is cause real temps uses a Tj. max of 95C and core temp uses 100C. Thats what I was told, plz correct me if i'm wrong.

[Simps]

Thanks very much for sharing the results with us, and for keep updating it.
Your real temp is set for 95C TjMAX and your core temp is set for 100C TjMAX. That is why there is a 5C dif. between them. You can change that TjMAX in real temp, on the settings tab.

What is the real TjMAX for 65nm quads? No one know.
Intel stated that it is 90C LOL, that means your temps shoud be 5C lower on real temp and 10C lower on core temp. No one believe 90C is real. Everybody till present date, is using 100C TjMAX for 65nm quads, so I would say, at least for us to have a reference, you should use that too. That means your core temps are correct, and your real temp is 5C lower then it should be.

Anyway, thanks for sharing it, gotta love the TRUEcu.

BTW, what is your Q6700 Vcore on the test?

[crazydiamond]

man id be nervous w/ that thing hanging off my mobo :eek:

[fatguy1992]

If I had 1 i'd tension a rope/string between it and the top of my case - just to take some of the weight off.

It makes me feel a bit better about my IFX-14 (790g + 250g fan = 1040g).

[BababooeyHTJ]

You hung it? I would be so worried. I was worried to hang my True black with 2 120x30mm fans. How is the bowing?

[Metroid]

I updated the original post with some 220 CFM Delta fan results.

Amazing review, thanks very much.

[eternal_fantasy]

One of the things about TRUE-Cu is it comes with a horrible base! I guess Thermalright figured most users would lap this heatsink anyway so they didn’t bother.
http://img185.imageshack.us/img185/855/00012nv9.jpg

I just received my TRUE copper and it has a mirror polished base...! :eek:

http://gallery.me.com/jeffrey_lee/100024/_BMP2136/web.jpg?ver=12274055050001
http://gallery.me.com/jeffrey_lee/100024/_BMP2140/web.jpg?ver=12274055070001
http://gallery.me.com/jeffrey_lee/100024/_BMP2141/web.jpg?ver=12274055100001

Maybe it's a later batch that had the base polished? The base seems completely flat too.

Here it is installed-

http://gallery.me.com/jeffrey_lee/100024/_BMP2148/web.jpg?ver=12274055160001

For now will use it till a 1366 waterblock I like are found. :)

[ex2cib]

Thanks for sharing the results with us!

Could you run IntelBurnTest and let us know how hot it gets?

http://www.xtremesystems.org/forums/showthread.php?t=197835

^^^^^

run that :yepp:

[Vapor]

I just received my TRUE copper and it has a mirror polished base...! :eek:

....snip...


Maybe it's a later batch that had the base polished? The base seems completely flat too.

Here it is installed-

...snip....

For now will use it till a 1366 waterblock I like are found. :)Yours is SO much shinier than mine....all parts of it. Like someone took a bunch of elbow grease and Brasso to it. :eek:

Mine looks like SamHughe's

[Demo]

@Simps

Wow, your B3 runs cooler than my own idle G0.. i have it at 3.2Ghz also. idle is at 48-48-40-40


@SamHughe

If possible, can you test using only 1 of those 1,600 S-Flex ?

[Metroid]

I just received my TRUE copper and it has a mirror polished base...! :eek:

Maybe it's a later batch that had the base polished? The base seems completely flat too.

It looks like we have the same batch, it has a polished mirror base and by the looks I can say it is flat, however I do not think it is a 100% flat. I was wondering if 1200 grits upwards would make it flatter.