non os measurement

[Bernhard]

This is about a partially filtered non os DAC.

First, music signal from an oversampling CD player, 31 kHz analyzer bandwith.
It can be seen the CD bandwith limitation of 22,5 kHz.



Now the transformer + cap filtered non os DAC.
The empty space above 22,5 kHz is filled with harmonics.
Non os gives back a little of what the CD format has taken away.



Pink noise from CD with oversampling, 83 kHz analyzer bandwith.



Pink noise from CD with non os DAC.
It can be seen, at 80 kHz HF is pretty much down.
That should not be a problem for a good amplifier.

[Bernhard]

1 kHz full scale sine non os.
There is the rest of the HF visible.



FFT of the sine, analyzer bandwith 10 kHz, 94dB vertical scale.
The k2 harmonic is caused by the passive I/V.



Same with 100 kHz analyzer bandwith, the 44,1kHz and 88,2 kHz are split, jitter

[Werner]

Originally posted by Bernhard
The empty space above 22,5 kHz is filled with harmonics.
Non os gives back a little of what the CD format has taken away.

These aren't harmonics, and thus these aren't what 'the CD format has taken away'.

Originally posted by Bernhard
Same with 100 kHz analyzer bandwith, the 44,1kHz and 88,2 kHz are split, jitter

Baseband frequency f creates images at every N*fs+f and N*Fs-f,
so 1kHz has images at
44.1kHz-1kHz
44.1kHz+1kHz
88.2kHz-1kHz
88.2kHz+1kHz
...

[Bernhard]

Originally posted by Werner

Baseband frequency f creates images at every N*fs+f and N*Fs-f,
so 1kHz has images at
44.1kHz-1kHz
44.1kHz+1kHz
88.2kHz-1kHz
88.2kHz+1kHz
...

Thanks

Still, there is something above 22 kHz and maybe it makes a difference. Perhaps.