> O'Meara's formula: M1=1/SQRT(1+SQR(1/n-n)/2) where M1 is band-span coef. and n is transformer ratio. maximal bandwidth have 1:1 transformers.
Am I doing something stupid? When I figure for a 1:1 transformer, I need the square root of zero? And the result is "1"?
Anyway, a practical man would say:
For iron-core transformers such as are used for power and audio, the ratio of effective inductance to leakage inductance is fairly constant and "about 1,000". Maybe 500 for very simple windings, maybe 10,000 for rich iron and layered windings. And the inductance is square of turns.
But the capacitance is mostly related to winding size, not number of turns.
So when you increase the number of turns, you increase the inductance and the leakage inductance, but the capacitance hardly changes. This means that a high inductance winding will have a low resonant frequency where the leakage inductance and the capacitance are equal.
In practice, 600 ohm windings can be simple, 10K windings show resonance, and 100K windings have to be carefully layered to meet hi-fi treble goals.
As far as I know, a 1:1 100K:100K transformer won't have a wide response. But a 1:10 6 ohm to 600 ohm winding probably will have wide response even with simple winding.
