[sdiy] Reading holes (voids) on piano paper rolls.. using Photo-transistors and PIC micro..

[Spiros Makris]

If you REALLY wanted to do it the most analog way possible, you can use
OTAs to make variable filters/time counters and achieve the 1/13 tuning
range you want (you can get a few orders of magnitude from these circuits).
Other than an interesting design exercise, I don't see any real value in
going down that path, though.
I think it's optimal to focus on how you can get an accurate representation
of the paper holes in the microcontroller, and then use software to
interpret them. I would also use the microcontroller to control the speed
of the roll with a PID, helping you achieve higher tempo stability and also
use that feedback into your code. If you can measure the shaft speed, then
you can easily calculate the on/off times your algorithm has to look for to
always work flawlessly.
Really interesting project, keep us posted!

On Fri, 11 Nov 2022 at 07:53, Donald Tillman <don at till.com> wrote:

> Interesting problem.
>
> Try setting up the light source and phototransistor for an aperture width
> of exactly the distance between sequential holes.  Ie., one hole plus one
> space.
>
>   -- Don
> --
> Donald Tillman, Palo Alto, California
> https://www.till.com
>
> > On Nov 10, 2022, at 5:56 PM, Jean-Pierre Desrochers <jpdesroc at oricom.ca>
> wrote:
> >
> > Hi all,
> >
> > My Player piano reader project is going fine so far..
> > Some little robotics stuff (stepper motors, sensors, etc..) going on
> now..
> > I’m now in the process of reading the small holes scrolling
> > with the music paper rolls ‘sandwiched’ between Infrared leds and 88 x
> Photo transistors (see the image below).
> > The actual music data appears on the paper as parallel lines of small
> successive holes
> > shown in blue below. Because of the need of paper sturdyness for long
> duration notes the original paper punches
> > made small successive holes in the same note line instead of a steady
> cut in paper
> > that could eventualy be damaged over time..
> > That is shown by the blues dots below (successive holes).
> > Now I need to read these constant note ‘ignoring’  all the occuring
> small holes
> > during these long  duration notes..
> > The second line shows what the reading Photo transistor ‘sees’
> > and the third line shows an ideal reading stored in the micro.
> > So far there are 2 ways I found to get close of the ideal reading :
> > #1  Place a capacitor in parallel with each Photo transistors to get rid
> of the ‘pulsating holes’ on a continuous note
> > #2  Use software with separate counters to ‘skip’ the incoming small
> pulses just enough not to miss any ‘real’ short notes.
> > My challenge here is that the paper roll speed will be a variable one
> > which could be varied from  1feet/min  up to  13feet/min.
> > That’s a  1 to 13 ratio !!
> > So method #1 using caps in parallel with the photo-transistors (fixed RC
> time)  will not be feasable (I think).
> > Is there another way you could think of to get rid of these small pulses
> > when necessary ?
> > Thanks !
>
> _______________________________________________
> Synth-diy mailing list
> Synth-diy at synth-diy.org
> http://synth-diy.org/mailman/listinfo/synth-diy
> Selling or trading? Use marketplace at synth-diy.org
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://synth-diy.org/pipermail/synth-diy/attachments/20221111/ccc96668/attachment.htm>