What happens when there's a differential in a balanced line?

This question popped into my noggin' the other day (I don't recall why)... Would anything disasterous happen if the "positive" and "negative", or pin 2 and 3, in a balanced line were to have several dBV of difference?

For instance if a mic pre or line amp output is balanced, and for some strange reason there's a rather significant difference between the 2 and 3 pin signal levels (the 3 which will usually be grounded at the input of the next stage), will this provoke some phenomenological audio beast to be awakened?

No, you still have a differential signal, just not fully balanced.

F'rexample: a 4mV signal in a truly balanced signal would have 2mV signal on one leg, 2mV on the other. The noise cancelation should be excellent as long as the noise pickup is equal and identical in both legs (which is in turn usually due to equal impedances and other characteristics in each leg).

If you had a 4mV signal with 3mV in one lg, and 1mV in the other, the recovered signal would still be 4mV (assuming the polarities were still correct!) and so long as the noise pickup were still equal in both legs then the noise would still be rejected, since there would be no differential between the two legs' noise signals.

This is how "impedance balanging" or pseudo-balancing works. There is no signal on one leg, but the impedance should be the same, so noise pickup should be the same. That way you can reject noise diferentially like a balanced line, but without the drive issues.

The most obvious problem is with close-to-maximum signals. In a truly balanced system, the maximum recovered signal before clipping is at its highest when both legs are equal. If one leg is higher than the other, it will clip before the other, therefore limiting the potential maximum signal recovered before clipping, but a theoretical worst case is 6dB for one dead leg and one leg doing all the work. Small imbalances are not disatrous though it is always honorable to strive for true balance.

Keith

Thanks Keith,

I always like learning the 'what ifs' to help fill in the gaps of what little I do know.

:thumb:

Ever look at the output of an API (mic pre, e. q., module, etc.)? Not only is the amplitude vastly different from one leg to the other, the waveform is very different, too! That's distortion, baby! Guess what, they do a great job of driving long cable runs, and some people even like the way they sound.....is lab spec boring?

Hey Niels... -you back in town yet? -Don't I still have your Godzilla crimp tool... let me check.

The API's that I've seen have a fully floating output (assuming you're talking the beasties with the block output transformer) and as such the measured signal on each leg reference to ground will depend entirely on how much impedance the other leg is seeing to ground.

-Think about it this way: if ground is galvanically isolated (as in all good transformer outputs :wink: ) then the ground terminal of your scope probe is connected to a point somewhere in between the hot and cold outputs. The potential divider is comprised of the input impedance of your scope (usually 1MegΩ) in parallel with the load impedance to ground forming the shunt leg, and the impedance of the other leg to ground forming the series leg.

In the case of the API consoles, if they feed one leg to ground via a lower impedance, the signal will become unbalanced, and the noise pickup will also be different in each leg. Within the console, this is probably not a big issue though!

Basically the measured signal on each leg of a floating output is determined by the impedance of the load. Unbalanced impedances will make for unbalanced signals. If you take both legs through a 1KΩ resistor to ground with no other load, the signal should be perfectly balanced on a scope, only the load should affect signal on each pin...

I was round at Greg's place last week, -it's nice to see his API board is still kicking pretty nicely! -He has my LA-2a and one o' my 1176's if you ever wanna check them out in person!

Keith

Even when viewed on a dual trace scope at the same time? Wouldn't the load on both sides be equal? Of course a precision wound tranny should make any single ended source evenly ballanced, could it be the 7$ output transformer?