Variable impedence inputs: good or evil?

[johnS]

Recently we've seen a rash of mic preamps that feature variable impedence inputs. Basically, these do-dads let you adjust the load going into the pre, thereby changing the frequency response coming in from the mic.

As with most fads in audio, some are fer it some are agin it. The fer it camp (which includes the manufacturers, I might add) claims that variable impedence inputs are an effective way to tailor the sound coming in from the mic. The agin it camp counters that, because the vast majority of modern microphones are specifically designed to see 1200 ohm inputs, the variable impedence dial is just a good way to muck up your sound (and pay more for the privilege).

What's your take?

[smopo24]

i thought they would be good for older ribbon mics; that was my understanding of their purpose for the most part.

[dagosto]

I've heard that they are really only effective on dynamics (ribbons included) since the preamp inside of condenser mics electronically isolate it from the preamp.

[puffpastry]

since the preamp inside of condenser mics electronically isolate it from the preamp

That's not true. Varying the impedance of any load will vary the way that the output driver stage of that mic will operate going into the load, and this will alter harmonic content. How much depends upon the load being driven as well as a few other factors which I won't bore you with.

[dagosto]

Originally Posted by puffpastry

That's not true. Varying the impedance of any load will vary the way that the output driver stage of that mic will operate going into the load, and this will alter harmonic content. How much depends upon the load being driven as well as a few other factors which I won't bore you with.

I've meant to say that it affects the sound of dynamic mics more than other mics. I should've been more clear.

[puffpastry]

Your point is well made.
It will have the largest effect on any mic with a transformer coupled output, since varying the load that the transformer sees will affect harmonic content. Transformerless designs will respond as well, but the response is most likely to involve increases in higher order harmonic distortion in the low and high frequencies. Since dynamics, by nature, must use step-up transformers, they will exhibit the most noticeable tonal changes.

[johnS]

Good technical info, but are any of these effects musically useful?

For example, I have a pair of Microtech Gefell m300s. Transformerless design, very neutral sounding condenser mics--almost invisible imho. What would messing with the harmonic content do to the sound? Something interesting, or just something bad?

[puffpastry]

That's hard to say. It depends upon how much current the output stage of the Microtechs are biassed for. You might see something nice, you might get something ugly, or you might not notice much difference at all.

[dagosto]

I didn't know that about the distortion in all mics. So, if I'm reading you clearly, in condenser mics you will see this distortion while leaving the frequency response relatively unaffected; while in transformer coupled dynamics you will see both of these effects.

[puffpastry]

Well, sort of... The output of the microphone is required to drive both the mic cable (which may be capacitive, and is probably reactive) and the input impedance of the next device. If you know a little about Ohm's law, you know that, the smaller the impedance is, the more current it will require from the output of the previous device's stage, and of course, the reverse is true as well.

So, imagine that we have an output stage that is designed to put out a signal at a given voltage level, say , .001 volts (~-60dB) at a given current level, say 60 microamps. Let's say that we can drive a 1500 ohm load with that much current, although truthfully that's a bit on the lean side, since the load won't always look exactly like 1500 ohms depending upon the frequency of the waveform.

But then, lets say that we change the load from 1500 ohms to 1000 ohms. Suddenly, we don't have nearly enough current to drive that load, and we start seeing a rise in load-based distortions at certain frequencies. These load based distortions are mostly 3rd harmonic (some people claim a 7th and 9th harmonic distortion), and they are at such low-levels that they would be perceived more as a "coloration" than anything else. If we were to drop the load to something egregiously low, my instinct tells me that we'd see a lot of the 7th and 9th as the microphone's active output struggles to push that load.

For transformers, something a little different happens. Instead of load based distortions, you get something more akin to filtration. Because a transformer is an inductor, and inductors alter phase response, most transformers will have resonant peaks. The way that we tame these resonant peaks is to use a Zobel network, which is just a simple resistor/capacitor combination. This will, in a sense, fool the transformer into behaving a particular way within that (previously) resonant band, causing its response to flatten out. Altering the load will alter the behavior of the Zobel network, causing a change in the transformer's frequency response.

If we go the opposite direction and increase the input impedance, we get different effect, at least in theory. Because our load is demanding less current from the microphone's output stage, more current is available to the output stage to actually conduct the AC signal, so you get the opposite effect, at least to a degree. With a transformer, I'm not certain. I suspect you would see an increase in resonant behavior, but truthfully I've never seen that addressed, and I've never actually witnessed it, so I can only guess.

Anyway, that's just a tip on an enormous iceberg.