The TAS5630BDKD package is specified as 3.10 to 3.60 mm high. Somehow I thought it was 3.40 to 3.60mm when I started my design. http://robot-and-machines-design.com/Files/Packages/HSSOP-44%20%28DKD%29.pdf shows those dimensions so they might have been in an old revision. Parts mounted on prototype boards have been about 3.35 mm high.
The heat sink used on the TAS5615DKD2EVM had a 2.2+-.01 mm high built-in standoff.
The heat sink used on the TAS5630DKD2EVM has a built-in standoff specified as 3.2+-0.01 mm high, for a maximum of 3.21 mm. That is 0.11 mm (4.3 mils) higher than the minimum specified height of the package. Consequently, the heat sink might not touch the top pad of the package. Am I missing something here?
I might have a similar potential problem with my own design which uses standard parts with minimum machining. I used a 30x30 mm fan blowing on a commercial quarter-brick heat sink, a bare copper plate, and four standoffs on 1.86x1.03-inch centers which match openings in the heat sink and copper plate. This has worked on prototypes, but I am reconsidering whether to use 1/8-inch standoffs, 3 mm standoffs, or another design for production. 1/8+-.005-inch standoffs are 3.048 to 3.302 mm high, potentially 0.102 mm (4.0 mils) higher than the package. 3+-0.13 mm standoffs are 2.87 to 3.13 mm high, potentially 0.03 mm (0.12 mils) higher than the package. Both the package and standoffs are typically more accurate than their worst-case specifications; but the worst-case scenario isn't good, and even typical heights do not allow much pressure on the top pad for good thermal contact to the copper plate.
I also wonder whether I will be bending the 1/16-inch G10 fiberglass board too much with the opposite worst-case situation: 3.6 mm package height and 2.87 mm standoff height.
Comments or advice?
