How to do measurement on Class-D amplifier

[Pano]

How low do you think the carrier has to be, to not use a LPF before the AP?

 

[fumac]

How low do you think the carrier has to be, to not use a LPF before the AP?

hello panomaniac
i think about this question for a long time
hard to get the answer.

because i don't know the max input voltage of min range of AP
our mhz class-d amp just output a 70mV(rms) carrier frequency at the speaker pin,much lower than others.
but i don't think this is the only reason for that we needn't use the extension filter to cornet to ap
because this signal also will harm to the THD or IMD to the input stage of AP system.

my new thinking is :
because our MCD is running at the 1mhz or higher(the last picture is running at 2mhz 100wrms)
and the AP bandwidth is 20~500k (they call >500k)
so , perhaps AP have nothing happen when you input a 1mhz signal to it , 1mhz is far away from 500k
(perhaps this is the right answer)

if this is right, then:
when the class-d amp runing under 500k, then need the AExx filter . even you get the very low carrier frequency level at the output pin.

i know , you are doing with class-t amp, and i think you need the AExx filter,
because class-T is running at 300k~1.5mhz(from ta2022 data sheet)
so when T is running at under 500k , will harm to AP.

perhaps yes

rg
fumac

 

[42n82rst]

Worthy topic may need Refocus?

*Not that I know the answers, and *Not that I want to beat a dead horse (or a topic): I think besides the enjoyment of DIY, the issue of cost is the next driver for taking on a project that leads to 100% final success.
NOTE: "Refocus" suggestion is coming from and old-skool Hardware design/test EE.
We may not be able to get near the fidelity/resolution of an LabView or APx5?? analyzers, but most of the required :test equipment: do NOT have to be stand-alone used hw, as described; instead they can take the form of plug-in modules + software (Win? Linux? Mac?).