The TO-220 packaged VR1 regulator and Q1 transistor in the Commodore 64 motherboard (reference designators cited here from the Rev. B breadbin) are mounted in a way that suggests they were supposed to be tied down to the PCB to thermally couple the metal tab of the package to the PCB.

The legs are bent, there is a significant area of metal from the PCB exposed with no solder mask below the package, and holes are drilled to match the hole in the package's metal tab.

Yet, these components are not mounted onto these areas with thermal paste and screws, just bent over, barely or almost touching the PCB, as seen on the images from assy 250425 below:

VR1 Q1

Commodore were renowned for their cost reduction measures - is this likely to be an instance of those, having found the cooling to be sufficient without screwing the packages down, or is there a likely engineering reason, e.g. to get more airflow around the packages?

  • 3
    I'm pretty certain (though can't prove) it's a cost reduction measure. Screws cost money, screwing down the parts costs time in production. Realizing they could do without, Commodore cut these costs. Oct 20, 2020 at 20:17
  • 1
    Agreed, particularly given the counter example to the right in the top picture with a heat sink and case screwed to it.
    – Jon Custer
    Oct 20, 2020 at 21:17
  • 3
    Whether or not a linear regulator requires heat sinking will depend upon how much current is actually drawn through it. Laying out a board on the presumption that heat sinking might be required will often cost essentially nothing if it turns out that the part works fine without a heat sink.
    – supercat
    Oct 20, 2020 at 21:40
  • 1
    I think you're right. This component is designed so it can be attached to a heatsink if needed, but if not, I assume it can be used without one. Oct 21, 2020 at 6:26

1 Answer 1


Some points that came in as comments support that this is likely due to a "just in case" design that would allow for heat sinking the packages through the motherboard if the need should arise.

Presumably the additional cost of drilling the unused holes is minimal, whereas unnecessarily screwing down the packages, probably adding manual steps to the manufacturing process, would add significantly to the cost.

  • The cost difference doesn't even have to be "significant". If you build millions of units, even a single cent shaved off the total cost will save you tens of thousands of dollars. Volume production - and engineering for volume production - is surreal sometimes. Nov 12, 2020 at 23:23
  • Drilling the extra hole is probably less costly than producing a new board design that doesn't include the hole. Nov 14, 2020 at 4:49

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .