> Ok, I think we are getting confused here.  Manchester encoding is not
> based on pulse width, but on the direction of the level transition in
> the middle of each bit.  There are synchronous schemes that do depend
> on pulse width, and which are also fairly immune to speed mismatches
> between encoder and decoder, but they are not Manchester.

Okay, I think we're dealing with essentially isomorphic codings here; the
two versions may be transposed by xor'ing each bit (on the "unencoded"
side) with the previous.  I've seen both forms referred to as "manchester"
coding.  Also, for the terminology, one scheme is described as always
having a transition at the START of each bit, and MAYBE one in the middle,
while the other MAYBE has one at the start and ALWAYS has one in the middle.
I personally find it easier to use the former interpretation since it uses
an actual event to locate the start of a bit (instead of having to wait
for the transition in the middle to determine when the start of the bit
actually occured).

If people use PLL's to sync on "the middle" of bits, all the more power
to them; even with Manchester coding as you describe it, though, I'd think
it easier and more reliable to read things as long and short pulses, and
then try to interpret the long/short pulse strings.