[solved] Button Matrix Latency

[Marxon]

Hi MidiBoxers,

first off all i want to thank TK for his

amazing MidiBox project.

Spending all the time since over ten years

and all this for free! Thats really impressive!

Ucapps was thereason for me to learn more about

electronics and to live the d.i.y. way of life.

THANK YOU TK!

Ok finally here is my question:

Does the button matrix latency only increase by adding rows

regardless the number of collums?

Or does adding colums affect the latency too?

I made a 7x4 matrix layout and wired them:

7 rows (d-out lines) and

4 colums (d-in lines).

Now i wonder if it would be better to wire them "4x7 style"?

But i guess 4x7 is better because all buttons in a row

can be scanned together so there will be only 4 "row scans", right?

Thanks for your advice!

Best regards

Marxon

Edited March 17, 2013 by Marxon

[TK.]

The shift registers are scanned each 250 uS, means: with 4 rows the worst-case latency is 1 mS, with 8 rows 2 mS, with 16 rows 4 mS

So - regardless of the row configuration, the latency is ignorable, you won't notice it.

 

The number of columns don't increase the latency, because they are read during a single scan.

 

Best Regards, Thorsten.

[Marxon]

Thanks for the info TK!

Best regards

Marxon