We had a linear amplifier requirement to deliver +30 dBm (1 watt) peak power to a 50 ohm load over a frequency range of 10 kHz to 40 MHz. The approach was to make a push-pull, class-A, transformer coupled amplifier and eventually, after some effort, things worked out quite well.
On the way to our success though, we did have some trouble with the transformer's ferrite core going into magnetic saturation at the low frequency end, at the 10 kHz end, of the frequency range. When we saw that happen, the waveform distortion had a particular contour which we will now look at for a moment.
When all was well, a nice and clean sinusoidal output could be driven that was just fine. However, if we had a transformer with an inadequate ferrite core that would reach its flux density limit, the sinusoidal output would simply drop to zero at the trailing ends of each half cycle of output as in the sketches below.
The transformers were supplied to us by a vendor, so we asked him to change to using a ferrite core that would support the required flux density at the low frequency end. It was done and everything from there on worked just fine.
If you see waveform distortion that looks something like this in an amplifier of your own creation, you'll know from this example what the cause is and what you'll need to do about it.