Originally Posted by Ricci
Great post Mike. The test you describe makes a lot of sense since anything involving speakers or musical content will be far too variable. It also sounds like it would be a bit tedious. Automating the process would be desirable if possible. What is the APx525 you mention? Perhaps the addition of a multitone test spanning a very wide freq range would be worthwhile too.
It's a new box from audio precision that has the awesome analog front end they're known for, but all of the data capture is done with your computer over USB. With the right modules, it's set up for doing Class D amplifier measurements along with a series of other tests. The amplifier load would of course need to be external to the unit though.
I think you're looking at about a $16k entry price though...real lab gear tends to be that way though. The only reason I mentioned this particular unit is because it has really cool automation features that would make this measurement a piece of cake.
Btw, I was thinking about what I proposed earlier and it'd probably be more meaningful to do smaller steps. Most amplifiers will do 6dB peaks, which would only be 6 iterations at 1dB steps. Regardless, the main intent is to look at how long each power level can be sustained, and then extrapolate that to the dynamic range requirements for specific source material.