Hi!  I'm a new member in France, and I've joined MMD because I've
MIDI-fied my player piano with a device of my own.

I'm a retired engineer, and I have been interested in pipe organs
since I was a child.  My job was related to computers, and my
hobbies concern musical electronics and microprocessors (I built
a microcomputer with a 6809 microprocessor!).

I discovered the MMD Archives by searching on the Internet and from
the French association "Perforons la musique".  I wrote an article
describing my player piano midification, which will be published
(in French) in the next issue of "cahier de perforons".

My player piano (a French "Klein") gives me room enough to install
a stack with electromagnet valves, on a piece of wood where holes come
out (sealed with a paper tape) connected to pouches and bleeds.

I've placed the valve seats on a stack.  Each seat is a brass model
porthole (internal diameter 4 millimeters).  Solenoids and pallets came
from small surplus 48-volt relays.  The arms of the pallets [the arma-
ture = the moving piece of the relay] are removed.  One side of the
pallet is covered with leather (to reduce noise), the other side with
(old) plastic insulating tape (for air tightness).  Centering is assumed
by non-magnetic pins in the square holes of the pallets.  (The
centering teeth on the solenoid core are eliminated).

The electronics consists of:

 - 3 identical printed circuit boards (4 of them max.), each with 32
outputs to control the valves, which may be controlled by the parallel
port of a PC.  (There delay problems with a Turbo Basic program running
under MS-DOS).

 - a small printed board with an AVR microcontroller (which converts
MIDI input to parallel output).

To adjust the valves I used a BF pulse generator and an electronic
pressure gauge connected to an oscilloscope.  The valve in test
was connected to a small cavity and a bleed to simulate pouch well.
I measured delays of 10 (on) and 30 to 40 (off) milliseconds with 30V
control and a gap of 0.3 to 0.4 mm.

It's been working for three months.  Meanwhile, I encountered an open
valve due to a small particle in it (or a centering problem ?).  Easy
to repair, but ...

Some questions remain about conception of the system, MIDI coding of
the Themodist, and use of a sequencer program (Cakewalk) to control a
player piano.  They will come later when I have more time.

Regards from France,

Jean-Pierre Jouandet

 [ Bonjour Jean-Pierre, et bienvenu a MMD !  Yes, I also used the
 [ 6809 microprocessor -- how time flies.  :)   Please tell us about
 [ your additional plans to make a music roll reader, too.  This
 [ topic is gaining interest at MMD.  -- Robbie