impedance problem

[udip]

Hi everybody!!!


I'm confused about opamps ideal source impedance, and
hope you can help me fix things.

I think that every opamp should idealy look behind and
see low impedance for best noise figures...
but than I look around and see that this is not the case.

My question is - why? In all opamps datasheets the noise vs impedance graph shows that the lowest noise is achieved
with the lowest source impedance possible, so
where is my mistake?

I'm asking because i'm designing a transformer mic preamp,
and wonder what should be the best impedance of the
transformer's secondary for the opa627.

Thanks.

[ulysses]

If you look a little more closely at a good thorough op amp data sheet, you'll see that there are two different kinds of noise specified: Voltage noise and current noise. If you divide the voltage noise by the current noise, you will get a value in ohms (V=IR). This, in theory, is the ideal source impedance for the op amp. In the case of the OPA627, you'll see that the current noise is ridiculously low, which means that the "ideal" source impedance is ridiculously high (like 20 Megaohms). In practice, you'll end up choosing the transformer turns ratio based on how good of a transformer you want to use (one with an "ideally" high ratio would have terrible phase response, among other flaws). But it is normal for FEt-input devices, like tubes, to wish for a source impedance that's substantially higher than is practical. So this is one of the considerations in your set of compromises. It's a very low-noise amplifier, so even though you won't find its "ideal" you will not have a noisy preamp. You could use a 1:10 ratio transformer such as the Jensen JT-115 and you'll suffer more performance "degradation from the transformer's high ratio than from the less-than-ideal source impedance feeding the op amp. It'll also give you lots of free gain (almost 20dB).