Every now and again, you get asked about a part that’s been around for ages. There’s not usually any surprises left and any ‘gotchas’ are usually well documented in the datasheet. So when I got a question via one of our catalogue customers about an industry standard voltage monitor chip behaving erratically, it came as a bit of a surprise.
The problem came, they explained, when their end-customer potted their PCB containing the device with a potting compound, an epoxy resin for which they kindly provided me with the data sheet. According to their data sheet, the epoxy was suitable for potting transformers and other electronic components.
I did some reading up on potting compounds and from this research a likely cause of the customer’s issue began to emerge. The problem with epoxy resins is that the glass transition temperature (Tg – the temperature at which the resin goes from being soft to being hard) is relatively high.
The Coefficient of Thermal Expansion (CTE) of the potting compounds is higher than that of the board, so the potting shrinks more than the board as it cools. For through-hole boards this is not a problem as the component leads provide ‘strain relief’ to alleviate this mismatch in CTE.
However, with surface mount boards, as the ambient temperature is below the Tg of the potting compound, the compound becomes harder and pushes against the surface mount component until in some cases it breaks the solder connection. For this reason, epoxy resin is not recommended for surface mount boards.
For potting surface mount boards, low Tg potting compounds like silicone and urethane, which remain flexible and deform rather than moving the component, are recommended. Our customer also supplied a silicone resin for potting via their catalogue and a quick check of the data sheet showed it was recommended for potting surface mount boards.
Their end customer hasn’t come back with any more questions, so I presume that this was the problem and that the proposed solution is working just fine.