Seeking advise on miniature pneumatic actuators and solenoid valve for a percussion player...

Putting the pneumatics in the circuit introduces delays that might be
hard to compensate, since  they depend on manifold pressure. Why do
you feel they are necessary? Why not hit the surface (or the far end
of a lever with a striker on the other end) directly with the
solenoid?

Yes, I think you are right. Just a solenoid (an electro magnet with
long enough stroke length) would do the trick.
Probably a PWM ignal may also control the applied force. What do you
think?
How to select solenoid? (Probably the selection parameters size
diameter/length, stroke length, operating voltage/current,...)
How to find out how much force required? (I think answer is the do
some experiment?)
Is there any application that you are aware of for the force control
with solenoid?

Roberto

IIRC, Hanon's piano finger exercises are often played around 108
beats/minute, 4 quarters per beat.  A Czerny exercise might go twice
that.  120*8 keys/minute = 16 Hz.  Yes, 20Hz is reasonable.  On the
percussion side, I'd guess that's near the max speed a snare drummer
could perform.


As ZWS noted, pneumatics are not known for speed or precision...


Pneumatic force is proportional to the supply pressure.  Solenoid force
is proportional to the drive voltage.  No good ideas for supply.


If you have trouble finding "the right parts", then get a single, cheap
actuator moving first...  If you have some servos modified for
continuous rotation or similar motors handy, you might try attaching a
spoked striking mechanism and spinning them at different speeds...

- Daniel

108 beast/min < 2 Hz, twice 108 beats/min < 4Hz.
It is impossible to get 16 Hz using *ONE FINGER* (You can try :-)
BTW, 20Hz accepted as continual sound, not as sequence of beats.

Correct; I was quoting exercises where the pianist plays successive
notes with different fingers.  For "single-finger" speeds, Hanon's
exercise 51 has the player play sixteenth-note octaves at 40-84 beats
(quarters) per minute.  That's 2.6 to 5.6 Hz per finger.  The trick is
not to move the fingers much; the wrist bounces much faster.


Yes, I've heard 20Hz quoted as the lower threshold of "sound", though
most people's ears have a higher frequency cutoff.  But I still maintain
that any competent pianist can strike 12 or more keys in succession
within a second.  Horowitz did much better.  Since the snare drummer
uses two sticks, a combined rate of 2*5 Hz still seems easily achievable.

- Daniel

Not yet... 5 Hz is pretty enough.

Hello,

In addition to the force applied by the striker to the percussion
instrument's skin, there are more parameters to select and control;

Stirker Material:
The striker material is important. Hitting to a drum, to a tabla, a
darbuka etc by hand or with a wooden striker produce different
sounds.
Human player use different sections of their hands. They also chance
the shape of their hands.
How to do this on a robotic player?

Duration of Contact:
The duration contact of the striker and with the instrument's skin
produce different sounds.
How to control this on a robotic player?

Location Of Strike:
Location where the striker hits to the instrument surface produce
different sounds.
How to move the striker to different locations over the instrument on
a robotic player?

Any comments?

Roberto

 People are scared of electric solenoids ,

 so they opt for pneumatic .

  But at 20 hz , solenoid is far more

 controllable and effecient and lower

 cost .

  The trick to solenoids , is using switch

 more techniques .   Even a stepper motor

 H bridge can be digital driven !

 It actually simplifies circuitry .

  Its fortunate , solenoid has high inductance.

   Simply use current feedback to tell the

 300khz driver when to on/off , and the inductance

 limits the current rise , to allow enuf time for the

 feedback to get to the controller ..

 PWM'ing is another digital method that needs

 inductance to create the analog result .
      --------------------------
   The LM494 controls power supplies , but it
 needs to control dead band ( obsolete ) .
 More modern ,, is to use an oscillating ,, 2 transistor
 half H bridge . Like used on all elctronic ballast , flour' lites .
First PC power supplies used LM494,
 now they oscillate , and i just bought a new PC
 p.s. that has only one NPN driver !

  BTW My 93C66 serial EEPROMs are blank
 and i need to know if they will boot .
  Has anyone ever tried to boot ?
 1) no address input
 2) apply power
 3) clock out data ,starting at
 4)  address 0000



__________________________-