Two common waveform problems with the ST heads are spikes or "ratcheting" from the non-skewed windings and the "harmonic hump" which is a distortion of the waveform caused by the harmonic winding. Both of these effects vary greatly from one unit to the next. Both are significant enough to cause some trouble on many ST heads with UPS units that are monitoring the AC waveform and frequency.
If you are going to the trouble of using an AVR, I would suggest using the mains for the excitation source, which will eliminate the harmonic hump. Alternately, Ronmar has found that rectifying the harmonic, then adding a large amount of capacitance (1000s of uF via electrolytic capacitors) got rid of his harmonic hump. You may be able to do that with some AVR designs but not those that use Triacs or SCR for switching of the excitation source. Check with the vendor before trying that.
The "ratcheting" with spikes on the waveform can be significantly reduced by adding up to 60 uF of motor run capacitors across the mains. (Across each leg to Neutral if the head is set up for 240V output with Neutral.) The amount of ratcheting varies greatly between ST units.
There is also a considerable amount of high frequency EMI on the ST mains from the rectifier diodes, regardless of the excitation via harmonic or mains. Few care about this, but for those that do, adding a 0.1 uF metal film capacitor (250V AC rated) in parallel to each diode in the bridge as a snubber will dramatically reduce the EMI.
With these fixes, the ST waveform is still more of a triangle than a voluptuous sine like a Marathon head makes, but works with most gear and is relatively "clean" AC (absence of high frequencies).
