Only a 'mentating' but:

That's an unusual cluster of errors.
EITHER the two middle bits are inverted OR they stay high when
preceded by more than one "1".
The latter seems more likely.

So, look for a hardware point lacking a pullup or pulldown resistor
(depending on what hardware you have in circuit.
The series of 1's drives the node high (or low) and it takes longer to
discharge to a 0 that the bit time.
This could be due to an open collector output but could also be caused
by eg writing to a register which is set as an input or writing to a
direction rather than a data register.

If you have an oscilloscope I'd not be surprised if you saw some
circuit point that had an exponentially decaying voltage rather than
the sharp edges that you expect.

A closer look at the limited example set provided shows that that
"rule" is perhaps not always followed (not enough data (even for
mentating :-) )), but

Note that characters are sent LSbit 1st so a look at any AC behaviour,
as above, needs to account for right to left reading.
Look for a series of bits or even an adjacent bit to influence one or
more to the LEFT of it.

________

I initially missed the meaning of the string you provided.
ie
dD ddddddffffffaybzaybzddeettyyzzayayaybzbzbzc{
means
First character transmitted results in 2nd character received.:

I assume dD is a typo?

dd
ff
ay
bz
ee
tt
yy
zz
c{

A whole alphabet sent in order with short pauses between each
character would be more useful.



=A0=A0 =A0 =A0 =A0 Russell

--=20
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist
.