Subscribe to EDN

Guy Dagan on heatsink planarity

February 2, 2010

Guy Dagan over at Cool Innovations saw my blog about using pressure-sensing film to evaluate your heatsink. This really struck a chord with him since Cool Innovations has found they need to lap their heat sinks flat for best performance, but sometimes the IC packages are not flat and that hurts your thermal flow. Here is a nice little write-up Guy sent:

I think that your readers may benefit from an expended discussion on the issue of heat sink/device contact. This issue has immense implication on heat sink performance, but is rarely covered. At Cool Innovations we have adopted the lapping process as the default base surface finish for all our heat sinks. We researched a wide variety of processes and found out that lapping is the most appropriate one. We have been lapping for over five years and find it to be a very valuable process. I believe that introducing your readers to the lapping process may be of interest to you. You may use the following in any way you wish.

The contact between a heat sink and the device it is mounted on is a function of the flatness and smoothness of both the heat sink and the package it is placed on. While the surface quality of the device is dependent on the type of package employed and the manufacturer, the contact surface of the heat sink can be controlled by using the appropriate finishing processes.

Flat and smooth heat sinks are of immense importance for two reasons: Improved flatness will result in less of a gap at the ends or center of the contact area, and improved smoothness will result in less microscopic voids, both of which need to be bridged with thermal interface material. Very poor flatness can also result in heat sink fall off due to delamination of thermal adhesive tape or thermal adhesive glue.

Smoothness and flatness qualities are dependent on the heat sink manufacturing process. Heat sinks that are extruded develop microscopic irregularities and can develop a curvature depending on the quality of the extrusion die and the stock material. Many heat sinks are face milled or fly cut flat, however, smoothness is limited by the cutter which will leave radial marks across the contact area which are microscopic peaks and valleys. Further compounding the problem, if the heat sink is low profile or thin base, machined heat sinks can warp due to clamping pressures or stress relief in the material itself.

The lapping process, on the other hand overcomes the shortcomings of the other base finishes. Lapping involves rubbing the heat sinks against a special flat hard table and using a fine grit abrasive solution to remove any irregularities on the base of the heat sink. There are several advantages to lapping. First, the flatness is near perfect. Lapping tables are manufactured to highest flatness and are replaced regularly as they wear down, resulting in a flatness of 0.001"/" or better. Second, the surface finish is very smooth, by controlling the particle size in the abrasive solution, roughness of the part can be kept at or below 16 RMS. Thirdly, the consistency of the process is outstanding as each heat sink is lapped individually. This process results in parts that require less thermal interface material reducing thermal contact resistances. The end result, is better and more consistent performance. Both in the prototype stage, and in the production stage, as quality does not diminish regardless of number of parts produced.

Be sure to think about planarity and flatness when you have to move a lot of heat from our modern IC packaging. That is a critical factor with some chips dissipating 60 watts or more.

Posted by Paul Rako on February 2, 2010 | Comments (2)

February 16, 2010
In response to: Guy Dagan on heatsink planarity
William Ketel commented:

Lapping does not need to be "very expensive", it can be fully automated, and the actual lappiung equipment is neither complex nor very expensive. But, lapping does take a bit of time. Probably, for dsposable consumer junk, lapping is not a good idea, but for items of value it does make sense. And, of course there is still a need for some thermally conductive adhesive. It turns out that the more thermally conductive the adhesive is, the less needs to be spent on surface smoothing. So like many things, it winds up being a trade-off.


February 2, 2010
In response to: Guy Dagan on heatsink planarity
ponguy commented:

Lapping your heatsinks sound like a very expensive and inpractical solution. There are some very good phase change thermal pads that allow good thermal performce on less than perfectly smooth surfaces. The idea is right, but the solution is a little off the mark.

POST A COMMENT
Display Name
captcha

Before submitting this form, please type the characters displayed above. Note the letters are case sensitive:

Advertisement
Advertisement
Advertisement
About EDN   |   Site Map   |   Contact Us   |   Subscription   |   RSS
© 2012 UBM Electronics. All rights reserved.
Use of this Web site is subject to its Terms of Use | Privacy Policy

Please visit these other UBM Canon sites

UBM Canon | Design News | Test & Measurement World | Packaging Digest | EDN | Qmed | Pharmalive | Appliance Magazine | Plastics Today | Powder Bulk Solids | Canon Trade Shows