BPA LM3886 - is it ok to mix batches?

TI acquired National Semiconductor in 2011 so if there were any changes to the process the LM3886 is manufactured on it would have happened around then. That said, one doesn't just change the semiconductor process, in particular not on analog products, so odds are good that the chips from the early 2000s are the same as those from more recent years.

Due to part-to-part variation the ICs won't perform exactly the same. Chips from the same batch are more likely to perform similarly, but that's not a guarantee. Chips from the same wafer would be even more likely to perform similarly, but even that isn't guaranteed either and I don't think there's a way to identify which wafer within the batch a particular IC came from.

I know hardly anything about the BPA300 from Shine7 and the Shine7 website doesn't seem to exist anymore, so I can't tell if part-to-part variation would impact the performance of the design. I would expect it to not matter, but don't know for sure. It's mostly the DC offset of the individual ICs that can cause a bit of trouble and that's easy to deal with using ballast resistors on the outputs of the ICs. Those are typically 0.1-0.22 Ω in series with each IC output.

I wouldn't worry too much about mixing and matching. If anything try to keep chips from the older generations on the same board but don't lose sleep over it.

Tom
 
  • Like
Reactions: mrjayviper