[sdiy] Using a PIC micro as an Octal schmitt buffer with programmable threshold levels..

[brianw]

Interesting.

What is your goal? ... or what is your exact challenge?

Are you trying to design an adaptive circuit because, e.g., each paper roll ends up having different alignment problems? There are certainly many engineering challenges that require adaptive solutions.

... or are the light pipes that have more difficulties always the same ones? If you know that this is entirely a mechanical alignment problem with the reader - that doesn't change from one paper roll to the next - then it might be better to tune the circuit for each specific channel.

Brian


On Nov 1, 2024, at 10:37 AM, Jean-Pierre Desrochers wrote:
> Hi list,
>  
> Long post here..
>  
> Since I often asked help and suggestions on this forum for many issues,
> I thought: Here is an idea I'd like to share..
> Some months ago I talked about a project I was working on
> to read old piano paper rolls.
> https://www.youtube.com/watch?v=ZhH9kftuT_A&t=2s
> This project works well.. but with some paper holes reading issues I had to address.
> Let me explain:
>  
> The 88 paper holes are read using IR photo transistors with 100k pullup resistors
> with the help of 74HCS165 SPI buffers through a PIC micro.
> A wide Infrared LEDs band is in front of the scrolling paper that has holes
> punched in it (played notes). Each hole lets the IR beam reach an aligned Phototransistor
> that has a ‘light pipe’ to ‘conduct’ the light beam to it.
>  
> <image002.jpg>
>  
>  
> Some of the light pipes has more difficulties to get the light flow to its PT..  
> This arrangement causes some PT’s ‘ON’ low state to get too high in voltage values
> read by the 74HCS165 SPI readers. The PT OFF state does not cause any problem
> since  it goes high impedance up to 5vdc. Ok. But the ON low state values can go from 0.31v up to 1.7v !!
> Because of the very variable 74HCS165 threshold voltage values from chip to chip
> this sometimes causes reading problems and loss of some paper hole notes..
> I tried to use 470k pullup resistors instead of 100k but the system
> starts to get in an unreliable floating state when no notes detected.
>  
> Here is my idea..
> Why not use a small PIC micro to work as an octal schmitt input buffer
> that has programmed threshold levels at fixed specific values ?
> Here are the HCS165 threshold levels compared with my PIC threshold :
>  
> <image006.jpg>
> Now the PT’s could have up to 2.5v in their ON (low state) and still
> be read as a low state by the still in place HCS165 SPI readers..
> 88 notes means 11 x PIC16F1829.. 
> The use of 8 x ADC inputs, applying threshold values while reading
> to finaly output 8 x standard digital I/O.
> Voilà !