Digital Distortion Compensation for Measurement Setup

[phofman]

vovi: Even in this setup you could measure very low DUT distortions, but you have to count on the DUT having a very flat FR in the distortions band. If the DUT rotates the distortion generated by the DAC side, it will not be properly subtracted by the corresponding ADC distortion and the actual DUT distortion (the one we want to measure) will be vector-added to the residual fake distortion.

That's why splitting the distortions and thus generating a clean sine signal which behind the DUT will contain distortions introduced by the DUT only.

How long does your manual compensation hold valid and to what extent? IME a major portion is basically fixed and repeatable, only minor deviations throughout time, due to temperature I guess.

 

[vovi]

phofman, Thank you, I understand you
IME a major portion is basically fixed and repeatable, only minor deviations throughout time, yes due to temperature.

 

[vovi]

phofman :

If the DUT rotates fase the distortion generated by the DAC side, it will not be properly subtracted by the corresponding ADC distortion and the actual DUT distortion (the one we want to measure) will be vector-added to the residual fake distortion.

Yes, but will RTA-RAW show this DUT phase deviation relative to the calibration value? Can we see this and take it into account?

 

[staccatiss]

Hi,
I would like to try out the E.Sokol ©generator but I cannot find it. Do you perhaps have a link to thje software?


Best regards,

the final? for harmonic compensation: @REW beta14 RTA + E.Sokol ©generator with harmonics & phases @ CS4398 + AK5394 = The MOD 3.0 + UX3 usb card

 

[vovi]

@staccatiss
http://forum.vegalab.ru/attachment.php?attachmentid=402898&d=1629452056

 

[staccatiss]

Thank you!

 

[phofman]

Yes, but will RTA-RAW show this DUT phase deviation relative to the calibration value? Can we see this and take it into account?

The RTA will show parameters of the incoming signal. If the ADC distortions are not properly compensated by the pre-distorted signal, they will vector sum with the DUT distortions to create new distortions. IMO it is not possible to determine each component of the vector sum then. The best option is keeping the DAC signal as clean as possible.

 

[vovi]

phofman:

they will vector sum with the DUT distortions to create new distortions. IMO it is not possible to determine each component of the vector sum then. The best option is keeping the DAC signal as clean as possible.

Thank you very much for the clarification.

 

[vovi]

the final for harmonic compensation - THD=-146dB =-148dB

 

[phofman]

IMO the coherent averaging has wiped out the fluctuating-phase remains of the harmonics. What happens with the regular (magnitude) averaging? What is time stability of the compensation?

 

[vovi]

phofman

What happens with the regular (magnitude) averaging? What is time stability of the compensation?

ARTA for the signal -3dB, date-time and correction of air temperature changes

 

[phofman]

Thanks, that looks pretty stable. H2 from -114dB compensated to -157dB and 12 hours later still at -146dB, IIUC.

 

[vovi]

H2 from -114dB compensated to -157dB and 12 hours later still at -146dB, IIUC.

this is the influence of temperature !

 

[vovi]

@phofman
2) ARTA for a signal of -3 dB, date-time and correction of air temperature changes using an air conditioner, for cooling AK5394A? - https://www.diyaudio.com/forums/attachments/equipment-and-tools/500062d1440342797-diy-audio-analyzer-ak5397-ak5394a-ak4490-audio-analyzer-ak5397_compared-ak5394a_distortion-versus-level_150823-png ??