Strange but removing the try/except part of the second script (leaving only the processing) removed the 2 minute lag at the end of each subroutine.