Problem(s) Encountered:

The dome is not following correctly, and it was immediately apparent that the cause is that the encoder is only registering changes in one direction of rotation.  We set the dome manually for half the night

Solutions:

Tried cycling power to the dome encoder interface (“buffer”) box.  This often helps when the dome setting becomes unreliable.  However in this case, with the encoder only counting in one direction (CCW if you’re interested), this did not resolve the issue.  Replaced chipset in the dome encoder interface box and verified that dome azimuth changed predictably in both directions.  Slewed and tracked the telescope around a bit with the dome slaved to it and sent the dome back home (Dome Home).  All worked well.  Not sure why this chipset fails with relative frequency, once or twice a year.  It is a bit concerning as a couple of the chips are no longer produced and subsequently are becoming more and more difficult (and expensive) to find.  As an aside having nothing to do with anything, I found that the External Computer switch (top-left on the TCS panel) was flipped up, ON.  I believe this is typically left down, in the off position, although it probably doesn’t matter for anything.