A new DAC idea

Status
Not open for further replies.
Jump to Latest
#102
SY said:
Just curious: in what EE school was Nyquist-Shannon not taught?
Click to expand...

If you don't know the infinite spectrum is the output of the dac Nyquist Shannon is for
aliasing and it becomes important on ADC that you need a right analog filter on input to limit the band pass of your signal noting have to do with dac section
just pay 40000 dollars for year the university for know nothing,,,, USA ohi ohi....
 
#105
gumo73 said:
I try to tooking about the problem solving on this idea .... but you toking about nothing
Click to expand...

I understand the topic well enough to tell you that it doesn't work.

Now you obviously don't. But instead of demonstrating something concrete you produce weird and confused sentences and throw out insults or ad hominem.

So I will make a very minimal request:
Generate some white noise at 44.1*n kHz, low pass filter it to <22.05 kHz, now take every n-th sample (to arrive at a 44.1 kHz rate) and
post an image
a) of the original signal and
b) of the difference (or error) to a NOS DAC output of the 44.1 kHz signal (which to make it simple is just repeating the last sample n-1 times) without additional low pass filters

You have Matlab, this should take you ~10 minutes.


PS: This will be largely a demonstration for yourself.
 
Last edited:
#106
Nos response

I don't need matlab for answer , if you have a stair-step on output of your dac
the impulse response is a rect and is frequency response is a sinc and you have
a complete deletion of any signal around fc in this case 44.1khz .
But i think this idea is interesting for real settling time of the dac output if is possible to
increase it the median of vout of any steps is too close to right value .
🙂
 
#108
gumo73 said:
I don't need matlab for answer , if you have a stair-step on output of your dac
the impulse response is a rect and is frequency response is a sinc and you have
a complete deletion of any signal around fc in this case 44.1khz .
Click to expand...

So you understand how the difference between these signal looks like (but can't or won't produce a simple graph showing it) and that adding a sine doesn't help.

Great. Finally some progress.


gumo73 said:
But i think this idea is interesting for real settling time of the dac output if is possible to
increase it the median of vout of any steps is too close to right value .
🙂
Click to expand...
Forget settling time, you are just confusing yourself with it, and the above sentence is incomprehensible anyway.
 
#109
On psychical filter RC if you want increase run up time without change RC values you can
apply a 2 level stair step .
This trick i think works with the dac output of a stair step its output need time to reach
the end value , the additional signal tangles the output stage as I/V I think ....🙂
 
Last edited:
Status
Not open for further replies.
Top Bottom