Zalman 9500

[astrophysics]

Just got one, boy is it huge!
My first gripe about it is, with my stock cooler it was idle @ 42C.
With the Zalman + AS5 it is @ 38C idle.
Lowest speed on both fans. However raising it to full on the zalman only produce 1-2C cooler.

My second gripe is I can twist the fan and it moves! It's definitely not going to fall of nor move on it's on but I am still able to put a tiny bit of force to it and move it,despite me screwing it as tight as it would go.

Now I know AS5 takes a few days to really break-in, but I know that will only constitute maybe 2C more cool. I am very disappointed I was expecting more difference from the Zalman compared to the stock AMD fan. Both were just as quiet, and 4-5C different doesn't seem worth the 60$ price tag.

So did I do something wrong here?!

Zalman pic for the curious

 

[RallyMaster]

Uh...is your computer dual core? Did it come with the heatpipe cooler? If so, you probably won't see much of a difference.

 

[JEDIYoda]

Originally posted by: RallyMaster
Uh...is your computer dual core? Did it come with the heatpipe cooler? If so, you probably won't see much of a difference.

I have a dualcore rigg and it has the 9500 on it.
The 9500 rocks!!
Compared to the stock heatsink thast came with the dual core!!

Also you could be facing a airflow uissue from inside your case.
As in do I have enough fans to move the air properly..
Does the air move from front to rear?
Are you using a side panel fan or a upper top of case fan?

Also note that some stock heatsinks rockand are just as good or almost as good as buying a seperate heatsink!

 

[astrophysics]

Theres a link to a picture of the system in my first post.
I have it opened case when I record the tempatures.
It is dual-core, and the original fan was the stock AMD fan that comes with retail.

The weird thing is, it is @ 38C but it doesnt feel the least bit warm. The fins or the copper heat pipes. The base is hard to tell since i am touching the top plate and not the actual base but from there it is cool too.

 

[soydios]

You want to make sure the airflow through the heatsink is towards the rear of the case, where the exhaust fan is. I would take a look at your load temperatures before you get all worried. The Zalman really shines over the stock heatsink when it has to cool a loaded dual-core.

Some questions:
- Did you run the CPU with the stock heatsink on? If the stock heatsink had a thermal pad, it would have contaminated the surface of the CPU's heat spreader, reducing the effectiveness of the Arctic Silver.
- Did you properly apply the Arctic Silver? You need a thin layer on the CPU. My personal technique is to cover the CPU with a layer thin enough that you can almost see the letters on the CPU heat spreader, and brush a tiny bit on the bottom of the heatsink. I clean both surfaces beforehand with ethanol and lint-free cloth. Then, put the heatsink on and twist a little bit in each direction to squeeze out any bubbles and make sure the contact is even.

 

[DerwenArtos12]

I've foudn that with as5, if I apply a thin layer, install the heatsink, let it run overnight, pull off the heatsink, wipe off the AS5 stuck to it then re-install the heatsink, I've seen an additional 2C drop just in that, I guess I'm bad with applying a "thin" layer but, I've found a way around it.

 

[astrophysics]

You may be right soydios. I did run the stock AMD fan on it with the thermal pad for the time being while ordering the Zalman.

I tried to clean it off as best I could, but you never know. And I'm sure I didn't get every bit. I did try my best to apply AS5 as thinly as possible. But it's definitely not as thin as you say where you can still read the information on the cpu.

My question now is, would the AS5 eventually thin itself out after use?
Also, this might be why the heatsink itself is only mildy warm when touching the pipes (not the fins). The heat probably isn't transferring that well.

 

[Bobthelost]

You did try to clean it off with soemthing like isopropanol didn't you?

The AS will thin/spread out over time, it needs 200 hours to settle down and cure. By thin we are talking about somthing so thin that you can more or less read the text on the heatsink through it?

 

[astrophysics]

More or less, I can't remember.
I used nail polish remover, I know I should get a pure substance, but I didn't have any.
I apply a drop no bigger than a grain of rice of AS5. Then I used a stick that came with my AS5 with a flat tip at the end to spread it around. It was uneven of course like if you try to put peanutbutter on a bread. But after that I would say there were a few thicker areas after spreading it.

But I did notice it dropped another 2C afterwards of running prime95/memtest (Can't be sure because of the weather). But I want to know, if I did apply it too thick, would it eventually thin the layer down a bit more to an optimal density?

 

[soydios]

Nail polish remover. Thermal pad. Lots of contaminating oils on the heatspreader right there. 70% isopropyl rubbing alcohol (cheap stuff from the supermarket is perfect, actually) is what I use.

If you screw the heatsink down as tight as your supposed to, it should spread the layer out. If you put heaps of thermal past on the heatspreader though, the heatsink will squeeze if off, and no matter what they say about AS5 not being electroconductive, it's got a very high silver content, and silver is the second-best conductor at room temperature, so I don't want any of it oozing out onto the motherboard around the CPU socket, because there's 60 Amps of power at 1.4V going through that socket.

The heatsink should only be mildly warm. Look at how huge it is!
My bet is that your load temperatures are at or around 50C. It's within nominal operating parameters for your CPU, it's just that you don't have nearly as much overclocking headroom as you could. Besides, 50C load is considered pretty good for a Pentium 4 Prescott core. =P

 

[Howard]

Originally posted by: soydios
If you put heaps of thermal past on the heatspreader though, the heatsink will squeeze if off

The temps would still be higher than necessary simply because the interface layer is thicker than it has to be.

and no matter what they say about AS5 not being electroconductive, it's got a very high silver content, and silver is the second-best conductor at room temperature

Actually, it's the best [electrical] conductor - of the metals, anyway.