When testing with 8 or 4 ohm resistive loads (8 ohms shown), the tips of the sine waves (both + and -) get a noise or instability on them as I get within a few volts of the clipping voltage. It triggers the Clipping LED on the EVM before the signal actually clips. I'm using a 50v, 10a linear supply. Here is the scope view. Is this normal or how can I reduce this effect?
Thanks for the ideas, Adam. I rechecked the supply voltage and now show it as the green line on the scope. It does dip to about 49v at the peaks but as you can see, there is at least 4 volts of headroom both above and below the peaks when this effect starts. By my calcs, with 45v-5v = 40v peak + and - going to the 8 ohm load (28v rms), it's really closer to 100w going to the load when the fuzziness appears. Also below is a blowup of the fuzzy area at 2x vertical scale and 100us/div. Looks like perhaps a frequency of maybe 60kHz - not the switching frequency for sure. I may be able to add more L-C filtering to the output to reduce it. One other thought is that I'm using a bank of wire-wound resistors as the load - may not be the most desirable load impedance?
