[sdiy] (simple) musique mechanique

[Robert Pietrusko]

Hi List,

I'm going to try to stuff it all into one email. :)
I'll try not to forget anything.

First off, Ken, yes I *love* all of the old automated
instruments.... orchestrions and such.  musee
mechanique stuff.  Goes back to riding merry-go-rounds
as a kid.  There is a whole community that still deals
in building those kinds of instruments, pneumatics and
all, which I really respect. some of their websites
are just insane.


as far as mimicking a human drummer. I think that is
really interesting but well beyond my knowledge.
However, making Hurky-Jerky Robot Music?  I'm really
into doing that! It is certainly nothing new but I
just really like to build stuff and it's rewarding to
hear it firing off like an uzi. :)

As far as the operation :  the "simple"  in the
subject line is not self-deprecation but rather
descriptive.  I've never worked with actuators before
so I wanted to make this as simple as possible. Sorry
I didn't take better pictures.  this probably shows
the mechanism the best :
http://www.enfason.com/cgi-bin/i/batonmaton/autoDrum041.JPG


I'm using a linear pull-type 12v solenoid with a force
of 128 oz.  at 1/8".  You'll notice on the solenoid
arm is a compression spring.  basically,  the solenoid
is constantly pulling on the stick and is thus
compressing the spring.  when my logic goes high in my
circuit (i.e. 'hit' ) the circuit cuts the current
from the solenoid and the restoring force of the
spring smacks the drum stick into the drum head,
bounces off, the solenoid turns back on and once again
compresses the spring.  I did some dynamics
simulations in matlab to determine the location and
geometry of all of the components, for maximum power
delivered to the head, and to compute the spring
constant and length. 

Also,  the amount of time the solenoid is off is
constant and was calculated to be less than the
natural period of the spring so as to prevent
oscillations, but long enough to allow the tip to meet
the head and reflect off from natural force.  

PWM  :  I agree that being able to control the
velocity would be great and I thought about that for a
bit, however PWM in this project is just a
practicality.   The solenoid does not have an infinite
'on' time, and therefore needs to be constantly
switched on and off.  enter PWM.  I found a duty cycle
that would maximize the solenoid's 'off' time but
still deliver enough current to overcome the force of
the spring and hold it in compression.  Velocity
control is an ultimate goal but for now, I had to
start somewhere.

I chose this method because it seemed very simple and
like Tim said, I liked the idea of not plowing a stick
into the head of a drum with a solenoid. So instead, I
let a spring's natural motion do all the work.  Also,
this minimizes any delay in the control-- there is no
'cocking' that needs to be done because it is normally
cocked.  I was worried that the result would sound
whimpy but it is actually quite loud.  


If anyone has that Nuts & Volts article Michael
mentioned I'd love to read it. It would be a huge help
as I haven't touched a PIC in about 2 years, much to
my chagrin.  Time to print out that 300 page pdf PIC
document.  Say, does anyone have a friend at kinkos?
;)


I'll keep you guys posted.  and by 'posted' I mean
I'll be asking midi PIC questions.  Consider
yourselves warned!

-bobby.