3w amplifier for portable stereo speaker

[Zero999]

Did you test a range of ICs at different temperatures?

Data sheets often give the worst case scenario so it's not surprising your tests show better performance than the manufacturer's specifications. I've found this to be the case with other ICs: for example, I noticed an oscillator built with th 74HC04 would work down to 1V but I wouldn't expect all ICs to perform the same, across the rated temperature range.

[Shredhead]

That's a good point Hero.  When testing amplifiers of any kind I record the maximum output when the amp or chip is relatively cool or at idle temp.  I test amps that are rated in the kW's too so it's generally not a great idea to push them with sine waves for an extended period of time.  I run the sines for up to 30 seconds tops usually to see if there is a drop in output due to heat or PSU sag. 

I didn't test this chip anywhere near it's thermal limits, maybe it would show much different readings if I did.  Real world content wouldn't be as stressful as sine testing though so I think you can count on my numbers for playing dynamic source material. 

[Audioguru]

1% distortion from an audio amplifier is pretty bad. The peaks are squashed because it is actually clipping but the clipping is not severe. The graph of distortion vs output power shows the distortion suddenly increasing as the output power is increased.

[Shredhead]

1% distortion from an audio amplifier is pretty bad. The peaks are squashed because it is actually clipping but the clipping is not severe. The graph of distortion vs output power shows the distortion suddenly increasing as the output power is increased.

In my experience with testing many amp topologies at different power ratings, when the peaks are squashed or cut off this indicates higher levels than 1% THD.  1% THD is not "pretty bad" in fact it is unlikely that you would detect 1% THD at 1kHz.  Here is an article where a blind test was conducted in the detection of THD:

http://www.axiomaudio.com/distortion

In this article an interesting conclusion was stated:

"For detecting distortion at levels of less than 10%, the test frequencies had to be greater than 500 Hz. At 40 Hz, listeners accepted 100% distortion before they complained. The noise test tones had to reach 8,000 Hz and above before 1% distortion became audible, such is the masking effect of music. Anecdotal reports of listeners' ability to hear low frequency distortion with music programming are unsupported by the Axiom tests, at least until the distortion meets or exceeds the actual music playback level."

In the vast majority of cases where an amplifier is reaching it's max output at 1% THD into a dynamic load, the drivers in the system are producing substantially more THD than the amplifier anyway.