[Davey]

This is getting tiring. 

I'm not talking about absolute polarity, I'm talking about relative polarity.......of all four of your created outputs.  How difficult is this to understand for an "Industry Participant?"
Your source, if it actually has signal on both lines of either the left or right "balanced" outputs, will contain two signals of opposite polarity.  If you're going to separate those out and plumb all four through your preamp, two of them will still be of opposite polarity from the other two downstream.  You'll have to reverse speaker wires to restore the same polarity on all four of those outputs.  Maybe your crossover requires an opposite polarity connection tweeter and woofer, maybe not.  But, you have to be aware of it.

Cheers,

Dave.

[JohnR]

Dave, he's planning to reconstruct the differential signal between the preamp outputs and the input to a power amp (with "balanced" input).

James, it's strange that you would ask a question with "please burst my bubble" and then argue with the answers given. I'm not an expert but from the answers already given (and the linked papers), it seems your solution may work in your case, but it's not something that can be given as a general solution. The reasons being:

1. A "balanced" signal is defined as a differential signal between pins 2 and 3, not as a "positive" signal between pins 2 and 1 and the inverted signal between pins 3 and 1. While it may be the latter, there's no guarantee.

2. Pin 1 is (correctly) wired to chassis ground, not to signal ground. Confusion over this is the "Pin 1 problem." Exactly what issues this causes seem fairly subtle, but again, it seems that your connection scheme would not be correct in this regard.

I'd just like to add that I believe your assumptions about the internal "architecture" of a balanced preamp are incorrect. My understanding is that it's likely to either be a differential circuit at the front end (see page 1 of Jenson note) followed by "single-ended" circuitry up until the output stage, or "fully differential" circuitry throughout as seen in e.g. Morgan Jones' book.

[randytsuch]

1. A "balanced" signal is defined as a differential signal between pins 2 and 3, not as a "positive" signal between pins 2 and 1 and the inverted signal between pins 3 and 1. While it may be the latter, there's no guarantee.

So this I didn't consider, that the differential signals would only be referenced to each other, and not to ground.  I guess they could be generated by a transformer, and if the tranny is not referenced to ground then the signals would not be referenced to ground.

I was thinking about a dac with differential outputs, where it is likely that the outputs would be referenced to ground.

[Davey]

John,

Yes, I understand what he's trying to accomplish and how the balanced signal would/might be "reconstructed" at a power amp input.  It's not clear to me that James understands the possible pitfalls of this scheme.

Anyways, the most important fact here is still missing.  What does the "balanced" output interface scheme look like internally on the "source" he's talking about here.  If there's no signal on pin 3 the scheme might still work, but it seems an awfully silly exercise to implement it in this way.  The pin 1 situation is also unknown.  It's probably at chassis ground, but maybe signal ground is also tied to chassis ground in this equipment.  Or maybe signal ground is disconnected from chassis ground at frequencies outside the audio band via an RC network. 

Many questions here but no answers.  Just a simple query of will it work or won't it? 

Good fun.

Dave.