🚀 go-pugleaf

	I was interested in the messages last week about mice "dying", because
one of our rodents started exhibiting the same behaviour (only moving horizon-
tally & not vertically) the day before.  The messages suggested to me that the
cause is the loss of pulses on one of the four outputs (two for each axis), and
blocking the light-path to one of the phototransistors on the good axis
duplicated the behaviour on this channel too. 

	A little bit of investigation showed that each phototransistor is
connected between +5V and a resistor to ground; the signal developed across the
resistor is then fed to one section of an LM339 quad-comparator.  The reference
for the comparator was 0.20V, and the output should switch between 0 and 5V as
the input passes through 0.2V.  A multimeter showed that one signal from the
faulty axis was only just coming up to 0.2 while the other three were reaching
up to 0.3 to 0.4.  Obviously the circuit characteristics had changed enough
with time that the output pulses were no longer being reliably generated.  The
quick fix was just to increase the sensing resistor from 470 to 820 ohms, since
for small signals the phototransistor should develop a given current flow;
passing this through a higher resistance gives a proportionately higher
voltage. 

	I hope this info might be of use to anyone with a dying mouse and a
little bit of electronics equipment (hint: SolderWick or a SolderPuller are
highly recommended when unsoldering the resistor, as the circuit board is
easily damaged). 

		idr