Author Topic: Class-D Audio output measurement  (Read 1503 times)

0 Members and 1 Guest are viewing this topic.

Offline ChoseeTopic starter

  • Newbie
  • Posts: 3
  • Country: 00
Class-D Audio output measurement
« on: December 08, 2021, 12:23:59 pm »
Hello EEVblog people!
I'm trying to characterize the audio performance such as SNR, THD etc. of a filterless class-D amplifier. I have read application notes on how to do this, where you low pass filter the high frequency voltage waveforms and analyze it with an audio analyzer. I also have a simulation environment up where I can do these things with an FFT. The paradox is, that is get different results when analyzing the voltage waveform vs. the current waveform in the speaker inductor(coil). I'm not sure what is most correct here. It seems that in the literature you analyze the voltage, however isn't it the current that actually moves the speaker?

What are your thoughts on this? Is it the current in the speaker that defines the 'sound' or the differential voltage across it?

Regards

Chosee
 

Offline PKTKS

  • Super Contributor
  • ***
  • Posts: 1766
  • Country: br
Re: Class-D Audio output measurement
« Reply #1 on: December 08, 2021, 01:27:43 pm »
depends...

All these "modern"  power and space saving gizmos...
they do not care about results ... real world results..

These results are called HiFi and linearity and response
are extensively discussed in audiophile forums...

Tubes are the TOP linear devices for audio for very good reasons..

The second alternative is electrical  results..
devices made to be small and power saving..


The first type of measures require a proper chamber
where the audio waves will be inspected.. cf standards

The second requires power meters and voltage/current spectrums

Very different and impossible to compare methods and results..
they are opposite to each other.

Paul
« Last Edit: December 09, 2021, 11:59:42 am by PKTKS »
 

Offline TimFox

  • Super Contributor
  • ***
  • Posts: 9000
  • Country: us
  • Retired, now restoring antique test equipment
Re: Class-D Audio output measurement
« Reply #2 on: December 08, 2021, 02:46:53 pm »
Actually, modernloud speakers are voltage-driven devices.  This is why the feedback on power amplifiers used to drive loudspeakers gives a low output impedance (constant voltage) rather than a high output impedance (constant current).
Since the impedance is a function of frequency, the current through the load when driven by a constant voltage will vary with frequency.
« Last Edit: December 08, 2021, 02:50:38 pm by TimFox »
 
The following users thanked this post: duckduck

Offline ChoseeTopic starter

  • Newbie
  • Posts: 3
  • Country: 00
Re: Class-D Audio output measurement
« Reply #3 on: December 09, 2021, 09:16:28 am »
Thanks for the replies!

It seems to me that you just have to choose whether you are analyzing the current or the voltage or just both. It just bothers me that i.e you get a SNR of 100 dB if you analyze the current and only 80 dB when you analyze the voltage...

 

Online nctnico

  • Super Contributor
  • ***
  • Posts: 28429
  • Country: nl
    • NCT Developments
Re: Class-D Audio output measurement
« Reply #4 on: December 09, 2021, 10:19:05 am »
Thanks for the replies!

It seems to me that you just have to choose whether you are analyzing the current or the voltage or just both. It just bothers me that i.e you get a SNR of 100 dB if you analyze the current and only 80 dB when you analyze the voltage...
That is likely because the current gets lower at higher frequencies (speakers are inductive!) so the higher harmonics get smaller compared to the fundemental frequency.
There are small lies, big lies and then there is what is on the screen of your oscilloscope.
 

Offline Circlotron

  • Super Contributor
  • ***
  • Posts: 3362
  • Country: au
Re: Class-D Audio output measurement
« Reply #5 on: December 09, 2021, 10:58:16 am »
however isn't it the current that actually moves the speaker?
The current provides the force that is applied to the speaker cone. How much the cone actually moves because of this force depends on a variety of factors.

Article on driving speakers with current rather than voltage.

https://sound-au.com/articles/current-drive.htm
 
The following users thanked this post: duckduck

Offline Marco

  • Super Contributor
  • ***
  • Posts: 7043
  • Country: nl
Re: Class-D Audio output measurement
« Reply #6 on: December 09, 2021, 11:05:18 am »
I'm not sure what is most correct here. It seems that in the literature you analyze the voltage, however isn't it the current that actually moves the speaker?

For HiFi it's easier to live in the idealized world before voltages gets dumped into nasty nonlinear mechanical contraptions, just measure it with a high power resistor :)
 

Offline SandyCox

  • Regular Contributor
  • *
  • Posts: 141
  • Country: gb
Re: Class-D Audio output measurement
« Reply #7 on: December 09, 2021, 05:36:16 pm »
For a two-level filterless class-d amplifier the THD of the output voltage is always more than 100%. You can read about this in the book by Holmes and Lipo on pulse-width modulation. You can easily prove it by making use of Parseval's identity.
« Last Edit: December 09, 2021, 05:38:47 pm by SandyCox »
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf