I did a little test today, hooked a 10A 240V variac to a SB-220 I just finished converting to 6M for a customer.
Without the variac and at 2065W DC input key down the sag from 2995V is 413V or 8.7% regulation. With the variac at 240V it is unchanged as expected. At 220V the sag from 2745V and 1840W DC input is 441V or 8.4% regulation due to less xfmr and variac loading. I didnt go any lower as the filament xfmr is off half of the primary winding. Thats not too bad for a variac of just about equal (2400W vs 2150W) rating to the SB-220 when the 150W filament is factored in. I was rather suprised.
The SB-220 transformer is rather wimpy while the variac is CCS.
Carl,
I'm having some problems trying to follow your math. I'm not seeing the same % regulation of the overall plate supply voltage.
Unless I'm wrong, I always thought % regulation was; no load V - full load V (sag) divided by the no load V times 100.
From your numbers; I see in case one, 413V/2995V x 100 = 13.79%
and in case two, 441V/2745V x 100 = 16.07%
My results show a poorer regulation with the variac set at 220vac, which is what would be expected. Your results are showing a better regulation. In addition, I'm not seeing how you arrived at those overall lower % numbers.
Maybe you define % regulation in some different manner.
I'm going to check a Handbook to see how the ARRL defines % regulation.
Fred