Hello ke.

> Sorry Paul B, but I have been silly enough to think that You must have
> to remember my problem !

  Well, there seem to be a number of problems.  Your printer has me a
bit baffled, I tend to put it in the "too hard" basket while I think it
over.  I had a problem in that your diagram with the three transistors
is scaled so big that I can't get a good look at it on my 17 inch
monitor, so I was going to print it out.

  Problem is that I now have here at home no less than *three* (second-
hand) laser printers that don't work and must be repaired.  OK, i've
sent it to my work now and I'll print it on the laser printer(s) there
which were bought new and always work (except when Windoze doesn't)!

  I'm going to need to know the printer connections - what each wire
connects to, what the sector patterns connected to each wire are, and
to remind me at what voltages you think the motor and electromagnets
should operate.

  What I can see from the photos so far is:

  There is a gap in the encoder wheel, which suggests that it must turn
one way only (clockwise, looking at the encoder).  That makes driving
the motor much easier if you do not have to reverse it.  You will want
only one transistor or FET (or Darlington pair) to drive it.

  The picture was taken with the encoder not in the "resting" position.
The outer encoder brush has been damaged, possibly by turning the
encoder the wrong way.  It needs re-forming very carefully with
tweezers.

  The spacing of the encoder segments would allow for 16 in each ring if
they were filled, but I presume there are exactly 12 segments (contacts)
in each.  And there are 12 characters on each wheel?

  The segments are "Gray" coded so you do not need to debounce them.  To
count a digit position, just wait until first one contact closes, then
the other.  When one closes, the other will always be open to avoid
any debounce confusion.

  I presume the electromagnets have to pull in to "lock" a wheel ready
for printing, and do not have to hold until the character is printed,
but they may have to do the opposite, i.e., hold from the start until
the character is lined up, then release until it is printed.  Do you
know which it is?

  There is a mechanism which closes a contact as soon as the gears start
turning, and opens as the hammer fires.  The hammer must fire last in
the sequence, and the paper is advanced during the time before this
happens.

  The connections appear to be:

 1} contact   | Which is the
 2} contact   |  common, and
 3} contact   |  which for
 4} contact   |  each purpose?
 5} Digit 1 Solenoid (Left-most)
 6} Digit 2 Solenoid
...
17} Digit 13 Solenoid (Right-most)
18} Solenoid common
19} Motor +ve
20} Motor -ve

  What component is under the PCB connecting to pins 18 and 19 (I cannot
see pin 20)?

  Do tell me again how much current the motor requires, and at what
voltage?  Your estimate of 7,5 (seven point five) milliseconds per
digit is noted.  That sounds rather fast for one of these printers to
me, as I figure it as printing not much more than one line (revolution)
per second but I'm sure we can figure it out (with a *lot* of work!).
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

> I try to get a LED to light the whole time after that I turn it on
> with a momentary switch (- on ). What who happenings now is, when I
> released the switch the LED turn off. ( IĞm only used bsf and bcf
> instructions )

  This was the question you asked more recently for which I suggested
you send the code and I would try and figure out what you were trying to
do.  That's a *lot* easier than getting the printer to run!
-- 
  Cheers,
        Paul B.