Browse Prior Art Database

Use of a mechanical probe to measure the bond line thickness of grease between the heat sink and die of CPU packages

IP.com Disclosure Number: IPCOM000004756D
Publication Date: 2001-May-01
Document File: 9 page(s) / 76K

Publishing Venue

The IP.com Prior Art Database

Abstract

Disclosed is a method for the use of a mechanical probe to measure the bond line thickness (BLT) of grease between the heat sink and die of mobile computer organic land grid array (OLGA) packages. Benefits include faster results without destroying the chip and module board .

This text was extracted from a Microsoft Word 97 document.
This is the abbreviated version, containing approximately 35% of the total text.

Use of a mechanical probe to measure the bond line thickness of grease between the heat sink and die of CPU packages

Disclosed is a method for the use of a mechanical probe to measure the bond line thickness (BLT) of grease between the heat sink and die of mobile computer organic land grid array (OLGA) packages. Benefits include faster results without destroying the chip and module board .

Mobile computers contain CPU modules that generate a lot of heat. This heat must be removed from the chip for it to function properly. To facilitate the heat transfer, a thermal transfer plate (TTP) is placed on top of the die and attached to the board. To avoid excessive stress on the chip during attachment, a gap is left between the bottom of the TPP and the top surface of the die. This gap is filled with grease that conducts the heat away from the die toward the TTP. Figure 1 illustrates the location of the grease in relation to the stack-up that includes the TTP, the OLGA package, and the board. The thickness of the gap that the grease fills is called the bond line thickness (BLT).

The BLT is a critical dimension and must be measured accurately. A BLT that is too high results in an unacceptable thermal resistance and may result in too high a chip temperature. A BLT that is too low results in grease pump-out. When thermo-mechanical strains occur on the TTP due to temperature cycling of the CPU module, the grease moves from the center of the die to its edges. The net effect of this pump out is an increase in the thermal resistance between the die and the TTP.

The conventional method of measuring the BLT is to cross-section the CPU module board through the center of the die, polish the board, and measure the BLT under an optical microscope (see Figure 1). The cross-sectioning technique however has two major drawbacks. It is too slow, taking about 24 hours due to the required cure time for the potting compound plus the polishing time. This method also is destructive. These drawbacks bar this technique from being used as an in-line measurement tool for process control.

Description of the mechanical probe BLT measurement technique

The disclosed mechanical probe technique entails a mechanical probe pushing down on top of the TTP in a location near the center of the die and measuring the force on the probe and the downward travel distance of the probe, as illustrated in Figure 2. The resultant force-deflection curve is shown in Figure 3.

Three regions are identifiable in the force-deflection curve. In Region I, the probe has not made contact with the top surface of the TTP, and no force is registered on the probe. In Region II, the probe has contacted the top of the TTP and is pushing down on it and some force is registered. This region extends until the bottom surface of the TTP contacts the top surface of the die. At this point, a kink is generated in the force-deflection curve because force is exerted when the TTP touches the die surface. The BLT calculation is straigh...