Profiling on populated or unpopulated board?

Hi,

Has anyone found any significant differences on profiling using populated and unpopulated boards? Is it worth sacrificing a board to the profile gods?

Joe

reply »

| Hi, | | Has anyone found any significant differences on profiling using populated and unpopulated boards? Is it worth sacrificing a board to the profile gods? | | Joe | What's the cost of not doing it. In the same breath, why not wait for a scrap board and to the profilling with it. There's more, but what are your needs? Currently, mine are 12-30 layer boards with perimeter BGA's ($200 a pop on a $3500 finished PCB) that are tough to part with, but profiling with TC's placed at the device's edges just doesn't always get it especially when internal planes are not near consistent causing one corner/side to heat too much and bridge and the other to not reflow at all.

And on it goes,

Earl Moon

reply »

Joe,

When developing profiles I recommend using a board that is populated the same way as the reflow will see in actual production, no more components, no less.

My company manufactures ovens and part of our QC Procedure is to make sure the thermal energy transfer is uniform across the width of the belt. In order to verify we run profiles using a bare board so that the thermal mass of the test board is uniform. These profiles are saved into the computer so that the customer can see the uniformity of the oven. When our customers have tried to use these profiles for their populated production boards they usually find the temps a little low (all depending upon how populated their boards are). If you have large components or a crowded PCB I would definately recommend profiling with populated board.

I agree with Earl, try to find a scrapped board to make your test fixture out of.

Brian

| Hi, | | Has anyone found any significant differences on profiling using populated and unpopulated boards? Is it worth sacrificing a board to the profile gods? | | Joe |

reply »