What goes here? -- this week's preamp thread

[fifthcircle]

Interesting points... To make it further confusing, perhaps... Why would 2 transformer based pres (ie the API and the Pacifica) have such differing depths of field. I use the Pacifica for main pairs regularly with fantastic results. I don't do as well, though with the API.

Why would that be? I don't know enough about electronics design to answer the question...

--Ben

[David Rick]

Originally Posted by fifthcircle

Interesting points... To make it further confusing, perhaps... Why would 2 transformer based pres (ie the API and the Pacifica) have such differing depths of field. I use the Pacifica for main pairs regularly with fantastic results. I don't do as well, though with the API.

Why would that be? I don't know enough about electronics design to answer the question...

A transformer-based signal path can ring, but it Dean Jensen showed that it didn't have to ring. It follows that some designs may ring more than others, and some transformer-based preamps may have pretty smeary-looking impulse responses while others don't. It might be interesting to compare the the time responses of API and Pacifica preamps.

(Historical aside: Years ago, Jensen wrote and sold a pretty capable AC circuit analysis software application written in PASCAL. He'd written it to help design transformers and then decided to commercialize it. It ran on Hewlett Packard "desktop" computers in the days before IBM PC's. Back then, SPICE packages ran only on mainframes.)

Of course the stated hypothesis attempts to explain things strictly in terms of linear system theory. But we know that the distortion characteristics of preamps are important and sometimes highly-valued. Transformers can exhibit all manner of interesting distortions related to the magnetic hysteresis curve of their core material. That's why you see designs like Nathan and Tim's new unit which offer a choice of transformers.

One consequence of hysteresis is that the way a system (e.g. preamp) behaves depends on what has already happened to it. (Linear systems are usually also assumed to be "time invariant".) So the signal from, say, a nearby wind player could, in theory, cause the preamp to have a different behavior when processing the (later) sound from a player farther away. One hint that something like this was happening would be if you noticed that the "depth of field" depended on how loudly the featured instrument was playing.

It's interesting to note that one of the more successful approaches to ITB preamp models is based on a Voltera kernel representation. This technique explicitly models certain distortion products.

I hasten to add that the above musings are hypotheses, not proven theories. I'd be more than a little chagrined to some day find the above treated as "scientific" gospel in the glossy literature for yet another mic preamp. But knowing the audio industry as I do, I wouldn't be very suprised.

Aw heck, just spell my name right!

David L. Rick
Seventh String Recording

[mixedupsteve]

I doubt it's a ringing situation. Most decent transformers don't have this problem in the audio range. It's probably the 2520 opamp which is just a forward sounding opamp. It tends to make everything sound close. That would be my guess.