[sdiy] My Player piano roll project... refining my design.. QRS test roll follow up..

Jean-Pierre Desrochers jpdesroc at oricom.ca
Tue Apr 11 21:38:59 CEST 2023

I just received a QRS test roll for testing all my 87 piano notes/sensors
and the test passed perfectly ! No missing note ! Even when tempo accelerated.
I’m VERY happy !




Here is the thing,


These old paper rolls have always had tendencies to 'undulate' over time..

I mean even if you align the paper let's say to the left from a known
paper hole position with the corresponding tracker bar hole
after a short while the paper alignment will slowly drift left or right
and you'll lose your 'alignment' and start loosing notes appearing
on the paper roll. So you cannot align then 'forget' what will happen.

But I found a very easy way to stabilize the paper position
all over the paper length.


But first let me describe my path so far..

My first attempts were to read the paper's far left/right edges

Using the tracker bar holes there for this purpose..

2 to the left and 2 to the right.

Well this approach worked very well for the original pneumatics 

auto stabilisation. But using these 4 holes with optic sensors

didn't work very well.. Here is why:

The 4 holes pneumatics workings were pretty 'forgiving' and tolerant
for any varying paper width, undulations and thickness (no paper crossing optics back then)
but could nevertheless detect any changes in the paper alignment
and act upon. Using optical sensors needed them to be at very precise
positions within the tracker bar to produce the same fast results
as the pneumatics. 

And that assuming the paper rolls width are ALL exactly the same
which is not the case.

Working Solution:


So.. Instead of checking BOTH left and right sides of the paper
why not use only one sensor.. let’s say one of the two far left ones.

And check the HI or LOW state of it while the paper roll is playing


                                    Transition zone

______________________ _|

  HI, No IR beam so                  |  LOW, IR beam passing so

   Paper over sensor                 |  edge too far right that let beam pass

   Push paper to the right        |  Push the paper to the left



The thing here is to try to keep the paper at the ‘transition’ zone
which is impossible  ->  but this keeps the centering motor working
all the time (almost.. since there is a slight delay between each changes)
and keep the paper aligned within less than 1mm precision !!!
Please note each paper holes have around 2mm in width so alignment is paramount !
Two sensors would not give this precision (unless there is a more elaborate design around them)
The only thing I have to do now is to adjust the ‘offset’ screw
at the left of the roll mechanism to align the paper holes
with the tracker holes just before playing the roll. That’s all.
But this time there is an auto-alignment working afterward !


Using this method shrunk a lot the centering software

and saved me around 15 piano rolls that were unreadable
and more ‘undulated’ and difficult to read than all the rest of rolls I had..
These rolls were almost ready for dumpster.. 
But using this easy technics I can now read them and saved them.


Too long email but explanations are always worthwhile.






> -----Message d'origine-----

> De : Roman Sowa <modular at go2.pl <mailto:modular at go2.pl> >

> Envoyé : 11 avril 2023 06:17

> À : Jean-Pierre Desrochers <jpdesroc at oricom.ca <mailto:jpdesroc at oricom.ca> >; Synth-diy at synth-diy.org <mailto:Synth-diy at synth-diy.org> 

> Objet : Re: [sdiy] My Player piano roll project... refining my design


> How precise are all rolls ever made? I mean I find it hard to believe that paper

> rolls made 100 years ago were so precisely made that paper edge was in +/-

> 0.1mm tolerance distance from the lowest or top note hole. What I'm saying is

> that maybe it's better to adjust paper position based on actual note sensors,

> and don't really care about the edges?

> OTOH it might be too complex, because you'd have to monitor the amplitudes

> of all sensors, multiplying circuits complexity a few times.


> As for paper edge location stability, I like Gerry's idea, it's the way to do it, but

> you may also go one step further, and use linear scanning sensor, the kind

> used in flatbed scanners, or industrial positioning systems. This will give you

> wide range of exact location info.

> You could even make simplified linear scanner on your own, just put a bunch

> of small phototransistors side by side, read them all by ADC and calculate

> measured paper edge location in the micro controlling stepper motors.


> Roman


> W dniu 2023-04-09 o 17:09, Jean-Pierre Desrochers pisze:

> > Hi all again,

> >

> > I’m still trying to refine my design and working on

> > the paper centering process circuit.

> >

> > So far I’m using the tracker bar edge’s holes to detect

> > left/right paper movements while playing the rolls.

> > Those who are at the paper’s edge sides.

> > The system works pretty well

> >

> > but still some rolls lacks of centering correction.

> > SO I want to do some tests using Photointerrupters - Slot Type – with

> > open collector Outputs.

> > Like those types:

> >

> > I’m trying to select the best accurate one to detect _paper edges_

> > not the overall paper cutting the IR beam..

> > Any idea of the best specifications to pick one ?

> >

> > *De :*Synth-diy < <mailto:synth-diy-bounces at synth-diy.org> synth-diy-bounces at synth-diy.org> *De la part de*

> > Jean-Pierre Desrochers

> > *Envoyé :* 23 février 2023 17:31

> > *À :*  <mailto:Synth-diy at synth-diy.org> Synth-diy at synth-diy.org

> > *Objet :* [sdiy] My Player piano roll project... a small video demo for

> > you..

> > *Importance :* Faible

> >

> > I’m not done yet with this project but this is promissing..

> >

> >  <https://www.youtube.com/watch?v=K7xXYqCB0AY> https://www.youtube.com/watch?v=K7xXYqCB0AY

> > < <https://www.youtube.com/watch?v=K7xXYqCB0AY> https://www.youtube.com/watch?v=K7xXYqCB0AY>

> >

> > I let the piano roll play its song while filming.. around 4:00 duration.

> >

> > Enjoy !

> >

> >

> > _______________________________________________

> > Synth-diy mailing list

> >  <mailto:Synth-diy at synth-diy.org> Synth-diy at synth-diy.org

> >  <http://synth-diy.org/mailman/listinfo/synth-diy> http://synth-diy.org/mailman/listinfo/synth-diy

> > Selling or trading? Use  <mailto:marketplace at synth-diy.org> marketplace at synth-diy.org

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://synth-diy.org/pipermail/synth-diy/attachments/20230411/915a1dd2/attachment.htm>
-------------- next part --------------
Synth-diy mailing list
Synth-diy at synth-diy.org
Selling or trading? Use marketplace at synth-diy.org

More information about the Synth-diy mailing list