For a recent project it was decided to try to use hot air to solder just a part to a PCB. This is a quick post just to document how it was attempted - it is not "best practice".
Although reflow soldering was not really mandatory for the particular PCB, it was decided to try to attempt it for the large TO-263 device that needed heat sinking. This is non-trivial due to the amount of metal that needs heating, and the holes which will enable capillary action to occur for flux and solder. This was my first attempt at such a procedure, which was to heat from the underside using a preheater, and heat from the top using a hot air tool.
The strategy was not completely successful (I suspect the part works, but the whole procedure could have been executed much better – it is not hard to melt solder, but far harder to do it at the correct profile) and the three main lessons learnt were (a) to practice more, (b) perhaps use heat sink compound with the thermocouple, and (c) to set the preheater to a much higher setting to get it to heat the board quicker.
The setup is shown here (you can see the amount of solder paste that was used; due to lack of stencil, it was directly placed on the large pad of the TO-263 device):
The thermocouple was pressed against the board (using a ‘helping hands’ tool) but sadly didn’t play a useful role during this attempt due to the single point of contact. I had initially pressed it into a hole but it had moved. Kapton tape and heat sink compound will be attempted next time. The final result is shown here:
A bit more detail:
Underside (cleaned of flux):
Here is a short (70 second) video showing the steps taken:
So, although the part is soldered down and may work, the next attempt will try to correct the procedure!