Heya all,
Trying to build a 16 Rows X 20 Columns (16x20) spectrum analyzer using a MCU.
I've attached a simple drawing of what I intend to to.
I don't have a clue if its ok or not,
that's why I am here - to ask you, the experts.

Tell me your thoughts and notes, everything is welcome,
just tell me what you think.

on a side note,
Some say DSP would fit better,
but I think a MCU would suffice for a simple analyzer,
If you thing otherwise do tell your thoughts.

Thanks a bunch to all the awesome people around here.

 

That is not the way I would build a spectrum analyzer around an MCU.
I would digitize the waveform and then compute the Fast Fourier Transform (FFT).

 

I do heard a lot about using FFT as well.
But the question is -
Can I do it in the way I've suggested?
Will it work?

Because I really don't know much about Fourier Transforms,
especially when it comes into implanting it in a project.

 

In your drawing the active filters would serve no purpose because you are summing all the outputs together.

 

No,
you got it all wrong.

each time frame, I enable one channel (Channel = Led Row = Filter),
thus, each frame time I am taking care of one filters output by turn.

this time frame is obviously will be really small,
so that the human eye wont recognize the lag.

The peak detector will reset every time I've finished pulling the information I need from the current channel (in this case - initial amplitude indication).
At the led matrix - The suitable row will light up the necessary amount of LEDs.

Thats how generally the basic idea.

Please continue sharing your thoughts and ideas,
this is the first time I am doing such a 'grand' project, everything is welcome.

 

Duplicate thread here.

http://forum.allaboutcircuits.com/showthread.php?t=83124

 

Yes, it is the same subject, but there it went off-topic.
So I decided to open a new organized one.

Anyway,
More opinions?