Class-D modulators and EMI interference

[SSassen]

I've been simulating several self-oscillating modulators (but mostly UcD) in LTspice and have noticed that simulation and practical application are far apart in many cases. One of the things I noticed is that EMI interference from the switching stage can really affect normal operation of the amplifier causing for excessive noise, squealing and crackling noises at the output due to EMI interference, which are not present in the simulation. A simple R/C filter in the feedback loop usually clears this up, but is not desirable in some cases as it significantly raises THD or causes for erroneous operation of the modulator. This is especially the case with higher order UcD modulators as it creates an extra pole in the feedback loop. I'm wondering how others have tackled these issues? I realize that my use of perforated board and through-hole components for the prototypes is not ideal but will a good double-sided layout and use of SMD components clear this all up? Any suggestions or comments are most welcome.

Best regards,

Sander Sassen
http://www.hardwareanalysis.com

[classd4sure]

Hi,

Yep.

Best example are the Hypex modules.

Proper component selection, placement, minimal lead length, minimal loop area, snubbing where required.. proper decoupling.. shielding.. what else.

I almost wouldn't worry about it that much with perf board though like you say, but there's still alot of choices you can make that will affect it greatly.

Definatly worth researching.

Of course there's still a number of other areas where the simulators fail to be realistic as well.

Regards,
Chris