How mismatched? Officially, matching impedances is for providing maximum power transfer; it occurs when the source Z is equal to the load Z.
But I think this is not what you're interested in.
I read this as you (the OP) asking about using a transformer in a circuit in a way that disregards the nominal source/load it was designed for. If that’s what you’re asking, the results vary for different transformers, but there are some general patterns.
Generally speaking, using, say, an input transformer specified for 150/600 source impedance and 10k load impedance in a circuit where the source impedance, load impedance or both are significantly higher or lower than the transformer was designed for will result in a shift in frequency response. Depending on the specific conditions, you may see a drop off in low end response or a drop off in high end response, and/or a rise in one or the other.
But this isn’t always the case. Some transformers are more sensitive to such misuse than others. Some may show little change in response from higher or lower than nominal source or load impedance, particularly some 1:1 transformers. In my experience, however, most will show at least some shift in response when these parameters are shifted up or down. For some, the shift will be drastic.
It’s easy to test this out yourself. Do a series of sweeps with a function generator on a transformer with different series resistor values on the input (between the generator and the primary of the transformer under test, taking into consideration the output impedance of your generator) and different parallel resistors across the secondary. Watch how the response changes on the bode plot with each sweep with the different resistor values on the input or the output.
If this wasn’t what you were asking about, please clarify.
February 01, 2006, 01:59:19 PM