THD reducing by using oversampling - is it possible?

[safwatonline]

THD

Does the oversampling help in reducing the THD like it helps in reducing the Noise?

 

[spring1860]

Re: THD

I don't think so

 

[safwatonline]

THD

hmmm, could you elaborate?!

 

[sankudey]

Re: THD

THD should be related to distortion. This should not have relation with over sampling, as the distortion appears some where else , most probably be fore sampling.

sankudey

 

[gszczesz]

THD

Sankudey has a point,

Distortion is a non-linear effect usually caused by amplifiers, mixers, etc... THD power should increase faster then the input signal, so if the input signal is increased by 1dB, then THD power should increase by more then that.

Now an A/D's sampled signal has "distortion", but I'm used to it referred to as sampling noise. It's power is generally flat and is a function of the LSB, or the biggest difference between two adjacent steps. It always thought it's classified as a noise, and therefore not part of THD. The THD component would be caused by the sample-hold anmplifiers.

 

[asuprash]

Re: THD

THD is distortion due to presence of harmonics and is only related to the
non-linearity of the input-output characteristic.

Oversampling does not have any effect on non-linearity. At best it can produce a
better reproduction of the fundamental but will not have any other effect

 

[safwatonline]

THD

ok, let me state my thoughts, having oversampled data means that the Distortion is distributed on the band 0 to fs/2 and having my bandwidth as a fraction of fs/2 (Oversampling) then i wont take all the distortion with me (i.e. my share of distortion decreased) which should mean that the Distortion benefits from oversampling just like noise.
however, i am not quite sure of how correct my thoughts are so any comment is welcomed here.
thanks.
Safwat