TIM application on direct Heatpipes

[CallMeJoe]

I got my Kingwin RVT-12025, and will be mounting it this weekend. I was looking at the base, and wondered if I should do anything different in my TIM application. I favor the "dollop" method in general, but am unsure about this instance. There are small voids in the base between the heatpipes and the aluminium insert. I am debating whether to try to fill them before mounting the HSF, or just place my dollop of TIM on the (lapped) IHS and let it try to spread up into the voids. I am also considering which TIM to use; have my choice of AS5, AS Ceramique, or Shin Etsu G-751 (my most likely selection).

Any thoughts?

 

[Zepper]

A review I read of one model of open pipe HSF or other said they got better results by filling the voids and then applying on the contact areas as normal.

.bh.

 

[DerwenArtos12]

I would use the spread method if it were me. Just use some kind of plastic or cardboard card to make sure it's spread evenly and thinly. Don't want any air bubbles.

 

[CallMeJoe]

Originally posted by: DerwenArtos12
I would use the spread method if it were me. Just use some kind of plastic or cardboard card to make sure it's spread evenly and thinly. Don't want any air bubbles.

I used the spread method with the Athlon XP's small exposed core, but with the larger C2D IHS feel it would be too easy to trap tiny air pockets under the heatsink, no matter how carefully you spread the TIM. I am currently inclined to spread the Shin Etsu across the base to fill the voids, removing the excess with a flat edge, then applying a dollop to the IHS before seating the heatsink.

 

[Idontcare]

Originally posted by: CallMeJoe

Originally posted by: DerwenArtos12
I would use the spread method if it were me. Just use some kind of plastic or cardboard card to make sure it's spread evenly and thinly. Don't want any air bubbles.

I used the spread method with the Athlon XP's small exposed core, but with the larger C2D IHS feel it would be too easy to trap tiny air pockets under the heatsink, no matter how carefully you spread the TIM. I am currently inclined to spread the Shin Etsu across the base to fill the voids, removing the excess with a flat edge, then applying a dollop to the IHS before seating the heatsink.

What you are describing as doing here to your HSF is called "tinting" and in fact is the manufacturers recommended method of applying Artic Silver's Ceramique.

Tinting is basically the process of adding TIM to either the HSF or IHS (or both), and then wiping as much of the excess off as possible with a lint free cloth or paper. (I use coffee filters, works great)

Make sure that you remove the excess TIM from the surface by wiping "across the grain" so to speak. If all your scratches or machine marks run north-to-south then you want to remove the excess TIM by wiping East-to-West for example. This ensures you leave the TIM down inside the machine grooves and scratchs.

When done properly you will be looking at a surface that looks like it is almost clean but slightly tarnised...i.e. tinted.

Next you dollop or line some TIM onto the IHS just as you would have had you never tinted the IHS and/or HSF, and assemble the HSF as per usual.

I tint my HSF and IHS surfaces regardless the thermal paste, it works with TX-2, AS5, ceramique, etc. Does it actually help reduce temps or increase overclockability? Have no idea, if it does then all I can say is that it doesn't do it to the extent that I have noticed. Why do it then? Because you can.

 

[CallMeJoe]

Originally posted by: Idontcare
What you are describing as doing here to your HSF is called "tinting" and in fact is the manufacturers recommended method of applying Artic Silver's Ceramique.

I am familiar with tinting, and practice it on conventional heatsink bases. The Direct Heatpipe heatsinks are a little different, in that they have small voids between the heatpipes and the parts of the aluminium base that run between the 'pipes. My idea is to fill those voids with TIM before mounting the heatsink over the main dollop.

 

[Idontcare]

I'm getting lost by the adjective "direct" in your use of it with "direct heatpipes".

Is a Tuniq Tower 120 a "direct heatpipe" to you? What is an indirect heatpipe?

If you are talking about squeezing some thermal paste into the area where the heatpipes make contact to the HSF base then I understand what you are saying, and I think it makes a lot of sense to do this.

Anything you do to increase the contact area between the HSF base and the heatpipes will improve thermal conduction and reduce the overall thermal resistance.

 

[Rubycon]

Is that smooth? :Q

 

[CallMeJoe]

Originally posted by: Idontcare
I'm getting lost by the adjective "direct" in your use of it with "direct heatpipes".

If you follow the link to the Kingwin RVT-12025 and check out the photos, you will see the exposed heatpipes in the heatsink's base; what Kingwin calls "Heatpipe Direct Touch". You will also see, if you look very closely, the small voids between the heatpipes and the parts of the aluminium base protruding between the 'pipes.

 

[Idontcare]

Ah, now I am totally with you.

Oh my! That is pretty dam cool. Direct touch as in heatpipes directly contact the IHS.

So to answer my other question, a Tuniq Tower 120 would be an "indirect heatpipe" as it's heatpipes do not directly make contact with the IHS.

So yeah, Rubycon's question is next. Just how smooth can they actually make it? The "voids" you are referring too should certainly pose a thermal transfer problem and filling them in manually with TIM should be done (I would think, obviously I have no experience to support this perception).

How the heck would you lap a direct heatpipe HSF? If you wanted to or needed to?

 

[CallMeJoe]

Checking with my straight-edge, the RVT-12025 base appears to be quite flat and smooth, with the exception of the voids previously noted. The very fine lines visible on the base are about what I would expect from a surface finished with ~800 grit paper, so I wouldn't think further lapping would be needed. Obviously, if you felt the need for a mirror finish, you would have to be extremely careful lapping the base so as not to weaken the sides of the heatpipes.

 

[The Boston Dangler]

Xigmatek is the OEM. they are supplying Kingwin and OCZ. www.frostytech.com thinks very highly of the HDT design, and they claim the contact area is very smooth. to me, tinting the HS sounds like a very good idea, but only experiments can confirm. i'll be picking up the HDT-SD964 soon.

 

[DerwenArtos12]

Speaking of Xigmatek, I can't seem to find it anywhere but, I love the design of their battle axe VGA cooler. It's got four HDT 6mm heatpipes. It's downright massive too, with a pair of 92mm, 2000rpm 23Db PWM fans. Looks similar to the S1 and S2 but with HDT and fans are included.

 

[CallMeJoe]

Originally posted by: DerwenArtos12
Speaking of Xigmatek, I can't seem to find it anywhere but, I love the design of their battle axe VGA cooler. It's got four HDT 6mm heatpipes. It's downright massive too, with a pair of 92mm, 2000rpm 23Db PWM fans. Looks similar to the S1 and S2 but with HDT and fans are included.

You could look for Xigmatek at Newegg. They started stocking after I purchased the Kingwin from Directron.

 

[DerwenArtos12]

Originally posted by: CallMeJoe

Originally posted by: DerwenArtos12
Speaking of Xigmatek, I can't seem to find it anywhere but, I love the design of their battle axe VGA cooler. It's got four HDT 6mm heatpipes. It's downright massive too, with a pair of 92mm, 2000rpm 23Db PWM fans. Looks similar to the S1 and S2 but with HDT and fans are included.

You could look for Xigmatek at Newegg. They started stocking after I purchased the Kingwin from Directron.

Yeah, they don't have the VGA sink nor did they have or could they tell me an ETA for it. I'm waiting to hear back from SVC and jab-tech. Someone's got to be able to get one.