double output cheat or grounding hell?

is there anything wrong with taking a ballanced output from a D.I. or a mic pre for example and splitting the positive and negative polarity signals into two out of phase unbalanced signals with a common ground?

could i then send the positive polarity signal and the ground to some analogue outboard (unballanced) before sending to the DA convertor and the negative polarity version and the same ground straight to another channel on the DA convertor as a backup, incase i don't get the compression/EQ/whatever settings right on the way in?

i'm just trying to use my analogue outboard while the signal is still analogue rather than converting to digital then back and forth during mixdown/mastering. i'm also trying to avoid any extra active stages like distribution amps etc.

would it be better to just build dual output stages or would they probably interact undesirably?

thanks for reading this and any help.

> taking a balanced output from a D.I. or a mic pre for example and splitting the positive and negative polarity signals into two out of phase unbalanced signals with a common ground?

For many solid-state transformerless outputs, this will work, though taking a known out-of-phase signal sounds like Asking For Trouble.

Some solid-state transformerless outputs will "force balance" the output, which could give very unbalanced levels.

Some solid-state transformerless "balanced" outputs are balanced impedance, but don't put signal on both legs. You can call this a fake or you can call it brilliant engineering; but you can't get a signal off the "other side".

Floating transformer windings (many old-pro tube and transistor amps) will give very odd results with two unbalanced loads.

It may work fine. It could drive you crazy.

> send ...signal ..to some analogue outboard ...to the DA convertor and the {other} straight to another channel on the DA convertor as a backup....?

Dammit, this is what Y-cables are for. Split your one signal. This is why inputs are hi-impedance and outputs are low-impedance: you can hang several inputs on one output.

Also, running unbalanced gives a big ground-loop around your outboards; if you have balanced everywhere then you can avoid that (though you need custom Y-cables).

> interact undesirably?

Murphy's Law never sleeps. However in small studio work, Y-ing one good output into a few good inputs is usually no problem at all. The main risk is that the Talent or Helper stomps on a cable and shorts it. If you short one side of a simple Y-cable or split-box, all outputs are shorted. You can insert a proper Distribution Amplifier, with input buffer and multiple output buffers, though this implies two more levels of amplification and subtle signal corruption. What pros often do is use one healthy output and put a series resistor to each input.

You have an output rated 600 ohm load, and two inputs of 20K impedance. Build a 1-in/2-out "Y" but with a 680 ohm (or two 330 ohm for balanced) resistors in series with each output. The 20K inputs see an extra 680 ohms in the source, their level drops a negligible 0.3dB. Now short one of the Y-outputs. The 600 ohm source sees a 680||20K= 657 ohm load, it is still happy. Typical modern "600 ohm" outputs are really around 60 ohms, so the level to the surviving recorder will drop almost 1dB, probably not noticable. If it is truly a 600 ohm output, it will drop almost 6dB, which will probably be apparent, but is really quite easy to fix in a DAW (ask the man who has had to recover from mid-concert "accidents").

K.I.S.S.

Ha! PRR; you're giving away one of the secrets of economic design!!.... The 'balanced' output with no signal on the -ve!
It's a trick thought up by those clever German chaps at AKG.... they even use it on some of their microphones! :shock:
But seriously, it is a useful trick; if you take an unbalanced output of exactly known impedance, call it the +ve leg, then take a resistor (or more usually, a resistor and capacitor) making exactly the same impedance, connect one end to ground and call the other end the -ve port, bingo, a balanced output that will behave nicely and will prevent horrid hums on your system. :grin:
AND the 'y' cable trick can be applied to it just the same! :wink:

Isn't that what's called "impedance balancing"?

> giving away one of the secrets of economic design!!

It's been discussed here before, no secret.

> a balanced output that will behave nicely and will prevent horrid hums on your system.

It maintains equal impedance at both outputs, so it does not upset CMRR of lower-Z differential inputs, particularly the one-opamp diff-input that is so common.

It has at least one less amplifier, one less stage of signal corruption.

It does not cancel external field around the cable. You can. in theory, lay a hot balanced cable next to a weak balanced signal and get zero crosstalk. In practice, true balanced output will give another dozen dB less crosstalk. In the shorter runs and all-same-song work of the average music studio, the crosstalk is not usually a problem. In a big operations center, where Howard Sturn is cussing loud on on one line while Bible Study Hour is on another cable running parallel for a mile, lower crosstalk is better.

I don't object to "impedance balanced": it is a terrific compromise between the clean simplicity of unbalanced and the ground-loop rejection of full balanced. In industrial use, full-balanced may be needed. But in small studios "fake" balanced can significantly reduce the number of amps from mike to master.

> AND the 'y' cable trick can be applied to it just the same!

Right. Maybe I'm just low-tech, but I am always amazed at complicated solutions to simple problems. Y the signal and see how bad it sucks. With semi-modern gear, it may not suck at all. Or not until one side gets shorted: probably not a big deal in a tracking studio, but stuff happens in live concert recording. So push the low-Z/hi-Z interface with some "isolation" resistors: won't be perfect isolation, but can be good-enuff to finish the show with no harm done except the shorted feed.

wow!! thanks for the fantastic answers... just what i was looking for!

much appreciated... Mark.