Velocity sensitive keys via DINX

[skunks]

There is a nice chip E510 for $20 that does a job I'd like to implement on core32+dinx:

http://www.cedos.com/e510_midi_chip.htm (see datasheet)

Why not using this chip? Because anyway core32 module is used and, I

think, it won't take all of it's resources to scan ~128 keys.

Then you may ask WHY CHANGEOVER type of contacts?

1. This type of contacts if the most convenient for implementing

mechanics on pianos, accordions, organs. Mechanical calibration is

simple. There is one pin for each key switching from one common bar

bus to another (so called NO and NC buses). Time between switches

is mapped to note velocity.

2. You have to connect as many wires as many keys you have + 2 (NO

bar and NC bar).

That's an advantage comparing to industrial

contact matrixes that has twice as many contacts.

Switch over contacts from my accordion (yet 1 bar missing, sorry) and scan matrix from Roland synth:

3. Anyway you have to use diodes for the matrix, so why don't use

multiplexors (or shift registers like on DINX) instead ? :) The prices

for those industrial keymatrix parts are $155 for PCB + $5 * 4 for

rubber contacts = $175 (prices found here

http://www.midi-store.com/Fatar-keyboard-parts-p-1-c-360.html )

Maybe PCB can be etched by yourself, so one could buy only rubber

contacts, but they don't fit into every music instrument that one

would like to midify.

Changeover type of contacts are universal :)

So I have one idea how to scan changeover contacts via DINX module.

Ground of the contacts in DINX module will be no longer Ground. It

will be connected to NO input bar or NC input bar during key movement.

So in order to distinguish between those 2 bars I have to connect both

of them to digital outputs of stm32 chip.

Scanning process will be made in two stages:

1. Set NO bar to low, NC bar to high, scan all the contacts.

2. Set NO bar to high, NC bar to low, scan all the contacts.

This way I'll get to know that contacts are pressed to NO bar, what

contacts are pressed to NC bar and what contacts are in intermediate

state (not pressed to any bar).

Am I thinking in a right direction? Can I replace GND in contacts in

DINX module with stm32's outputs?

[TK.]

Did you know this example project?

(the topic name might be confusing, therefore you probably haven't found it - I will change it in future)

Best Regards, Thorsten.

[carlosyounes]

There is a nice chip E510 for $20 that does a job I'd like to implement on core32+dinx:

http://www.cedos.com/e510_midi_chip.htm (see datasheet)

Why not using this chip? Because anyway core32 module is used and, I

think, it won't take all of it's resources to scan ~128 keys.

Then you may ask WHY CHANGEOVER type of contacts?

1. This type of contacts if the most convenient for implementing

mechanics on pianos, accordions, organs. Mechanical calibration is

simple. There is one pin for each key switching from one common bar

bus to another (so called NO and NC buses). Time between switches

is mapped to note velocity.

2. You have to connect as many wires as many keys you have + 2 (NO

bar and NC bar).

That's an advantage comparing to industrial

contact matrixes that has twice as many contacts.

Switch over contacts from my accordion (yet 1 bar missing, sorry) and scan matrix from Roland synth:

3. Anyway you have to use diodes for the matrix, so why don't use

multiplexors (or shift registers like on DINX) instead ? :) The prices

for those industrial keymatrix parts are $155 for PCB + $5 * 4 for

rubber contacts = $175 (prices found here

http://www.midi-store.com/Fatar-keyboard-parts-p-1-c-360.html )

Maybe PCB can be etched by yourself, so one could buy only rubber

contacts, but they don't fit into every music instrument that one

would like to midify.

Changeover type of contacts are universal :)

So I have one idea how to scan changeover contacts via DINX module.

Ground of the contacts in DINX module will be no longer Ground. It

will be connected to NO input bar or NC input bar during key movement.

So in order to distinguish between those 2 bars I have to connect both

of them to digital outputs of stm32 chip.

Scanning process will be made in two stages:

1. Set NO bar to low, NC bar to high, scan all the contacts.

2. Set NO bar to high, NC bar to low, scan all the contacts.

This way I'll get to know that contacts are pressed to NO bar, what

contacts are pressed to NC bar and what contacts are in intermediate

state (not pressed to any bar).

Am I thinking in a right direction? Can I replace GND in contacts in

DINX module with stm32's outputs?

Where can I purchase this wonderful ci, was what I needed for my first prototype of a keyboard?

Thanks

[skunks]

at http://www.cedos.com/e510_midi_chip.htm