Splitting a balanced line output

[conleycd]

Hi friends,

I want to check my thinking out about splitting (or duplicating) a balanced line signal.  What I am confused about is the need for the input impedance to be 10x greater than the output impedance (so as not to load) and the output's ability to "drive" a specific load (i.e. 600 ohm).

So... if I'm coming out of a microphone preamp (now at balanced line level) and I want to split the signal -> can I just strap on a resistor (? another 51R in parallel with the existing ones on the output stage) on both + and - and run to my second output as long as the total load can be driven by whatever the output stage is (NE5532, TL072 etc)?  And/or... do I need to be able to drive the total impedance of the resistor breakout plus have maintain the 1:10 impedance ratio rule?

In terms of conceptualizing this... I think if I need to respect the 1:10 rule then I need to attach opamps (running at unity gain) ontp however many additional outputs I want.  The inputs of the opamps are near 1meg so I can strap a bunch on and still be well within the 1:10 rule while maintaining low impedance outputs (coming out of the opamps).

Is the capacity drive a 600R load the same as respecting the rule that the input should be 10x the impedance of the output.

Help?  What might I be missing here?

CC

 

[Andy Peters]

Hi friends,

I want to check my thinking out about splitting (or duplicating) a balanced line signal.  What I am confused about is the need for the input impedance to be 10x greater than the output impedance (so as not to load) and the output's ability to "drive" a specific load (i.e. 600 ohm).

So... if I'm coming out of a microphone preamp (now at balanced line level) and I want to split the signal -> can I just strap on a resistor (? another 51R in parallel with the existing ones on the output stage) on both + and - and run to my second output as long as the total load can be driven by whatever the output stage is (NE5532, TL072 etc)?  And/or... do I need to be able to drive the total impedance of the resistor breakout plus have maintain the 1:10 impedance ratio rule?

In terms of conceptualizing this... I think if I need to respect the 1:10 rule then I need to attach opamps (running at unity gain) ontp however many additional outputs I want.  The inputs of the opamps are near 1meg so I can strap a bunch on and still be well within the 1:10 rule while maintaining low impedance outputs (coming out of the opamps).

Is the capacity drive a 600R load the same as respecting the rule that the input should be 10x the impedance of the output.

Help?  What might I be missing here?

CC

It is VERY common to use a simple passive Y cable to split a balanced output into two or more inputs.

Don't overthink it. Most active balanced outputs have low-enough source impedances (generally less than 100Ω) so ten times that impedance is 1kΩ. Given the common 10kΩ or 20kΩ input impedances, you don't really need to worry about voltage-divider effects going into two or three loads.

Hopefully the designer of the preamp chose an output amplifier that's capable of driving your intended load.

-a

 

[Flundran]

One could use a 1:0.7+0.7 splitting transformer. If you then for example connect two 10k loads to the secondaries, the primary will reflect 10k load to the source.

/F

 

[conleycd]

Thanks for the insights...

I recognize that if the impedance of the inputs on two mixers are in the 10k range then the total is about 5k so a Y split is fine.  My problem is that the total load is going to be lower.  The one preamp which I need to use for multiple outputs is going to go into two mixers (into the microphone input) which are about 1.2k (not my choice to use the microphone input as opposed to the padded line input that is about 10k - no control on this one for me).

So here I am now below the 1:10 rule with a 600 ohm input.  Now.. if the opamps are able to "drive a 600 ohm load" that are in the output stage of this preamp am I okay or is my output going to be loaded down?

I really want to avoid using a transformer for a variety of reasons - mostly cost.  I'm more inclined to strap on a unity gain dual opamp and buffer it from the main feed.

Thoughts?

CC

 

[strangeandbouncy]

Why not just hook it up and see?  bet you won't notice the difference, unless you are looking for subsonic shizzle goin' on . . . What's the worst that can happen? slightly less headroom/ slightly less bass, how often do you NEED to be ruler flat down to 20Hz or lower, anyway? I haven't got a single piece of gear in our studio that is 600Ohm input! they are all 10k.

  AndyP

 

[abbey road d enfer]

One could use a 1:0.7+0.7 splitting transformer. If you then for example connect two 10k loads to the secondaries, the primary will reflect 10k load to the source.

/F

Then the levels are down 3dB. May or may not be a problem.
A complicated way of doing a simple thing IMO.

 

[conleycd]

Why not just hook it up and see?  bet you won't notice the difference, unless you are looking for subsonic shizzle goin' on . . . What's the worst that can happen? slightly less headroom/ slightly less bass, how often do you NEED to be ruler flat down to 20Hz or lower, anyway? I haven't got a single piece of gear in our studio that is 600Ohm input! they are all 10k.

  AndyP

Yes... I could just "do it" but I'm a little bit about the "right way" to do (perhaps to a fault). 

I don't care about superflat eq but I would like things reasonably replicated on the output.

The problem isn't that I have a 600 ohm input that I'm dealing with - I'm creating a 600 ohm by running my output to 2 - 1200 ohm inputs.  Line inputs are often 10k or even 20k but microphone inputs are typically around 1500 ohms.

CC

 

[JohnRoberts]

Yes... I could just "do it" but I'm a little bit about the "right way" to do (perhaps to a fault). 

I don't care about superflat eq but I would like things reasonably replicated on the output.

The problem isn't that I have a 600 ohm input that I'm dealing with - I'm creating a 600 ohm by running my output to 2 - 1200 ohm inputs.  Line inputs are often 10k or even 20k but microphone inputs are typically around 1500 ohms.

CC

If you are sending a line input to a mic input, you will surely have too much signal level.  That and the mic input is likely optimized for a 150-200 ohm source impedance (typical mic). You can kill both birds with simple resistive pads.  The mic preamp has plenty of gain so any reasonable pad will get your termination impedance up...

JR

 

[conleycd]

Cry and sigh...

Okay... My church insists on running everything through their XLR microphone inputs on their Mackie TT24.  There, I said it.  Actually, with that microphone preamp structure it seems to work pretty fine for the most part.  The +4 line signals that seem not to overdrive the input.

Yes... it makes more sense to run these things into the Line input which (from what I can see on the block diagram) the microphone input with a pad - but they refuse...

So... I'm still stick stuck with my question do I need a 1 (output):10(input) ratio when the spec for the output opamp says, "Can drive a 600 ohm load"?

CC

 

[Andy Peters]

Cry and sigh...

Okay... My church insists on running everything through their XLR microphone inputs on their Mackie TT24.  There, I said it.  Actually, with that microphone preamp structure it seems to work pretty fine for the most part.  The +4 line signals that seem not to overdrive the input.

Yes... it makes more sense to run these things into the Line input which (from what I can see on the block diagram) the microphone input with a pad - but they refuse...

So... I'm still stick stuck with my question do I need a 1 (output):10(input) ratio when the spec for the output opamp says, "Can drive a 600 ohm load"?

CC

Well, the people who run audio at churches tend to not know what the hell they're doing, and yet they will insist that their way is the only way.

In your case, the right thing to do is bring an XLR-F to TRS adapter cable, and tell them: "It has to go in via the line input because I am sending it a line output." Do not confuse them even more by talking about impedance or levels or anything, because it will be too much information for them to understand.

I assume that the only reason you're doing this is because you want to split to a house mixer for the audience as well as driving a line-in recording device. But you describe driving TWO mic-level/impedance inputs, so in that case, why bother with the external preamp?

-a

 

[abbey road d enfer]

Cry and sigh...

Okay... My church insists on running everything through their XLR microphone inputs on their Mackie TT24.  There, I said it.  Actually, with that microphone preamp structure it seems to work pretty fine for the most part.  The +4 line signals that seem not to overdrive the input.

Yes... it makes more sense to run these things into the Line input which (from what I can see on the block diagram) the microphone input with a pad - but they refuse...

So... I'm still stick stuck with my question do I need a 1 (output):10(input) ratio when the spec for the output opamp says, "Can drive a 600 ohm load"?

CC

Your balanced output actual source impedance is bound to be 50-100 ohms, and probably purely resistive (unless it's xfmr balanced), so it has no problem being loaded by ca. 600-800 ohms. A transformer in there would be detrimental to the audio quality.
There is nothing intrinsically wrong with running mic inputs at line level on a modern (transformerless) mixer, in particular the TT24 handles +22dBu at its mic inputs, with an actual impedance of 2k.

 

[conleycd]

Cry and sigh...

Okay... My church insists on running everything through their XLR microphone inputs on their Mackie TT24.  There, I said it.  Actually, with that microphone preamp structure it seems to work pretty fine for the most part.  The +4 line signals that seem not to overdrive the input.

Yes... it makes more sense to run these things into the Line input which (from what I can see on the block diagram) the microphone input with a pad - but they refuse...

So... I'm still stick stuck with my question do I need a 1 (output):10(input) ratio when the spec for the output opamp says, "Can drive a 600 ohm load"?

CC

Well, the people who run audio at churches tend to not know what the hell they're doing, and yet they will insist that their way is the only way.

In your case, the right thing to do is bring an XLR-F to TRS adapter cable, and tell them: "It has to go in via the line input because I am sending it a line output." Do not confuse them even more by talking about impedance or levels or anything, because it will be too much information for them to understand.

I assume that the only reason you're doing this is because you want to split to a house mixer for the audience as well as driving a line-in recording device. But you describe driving TWO mic-level/impedance inputs, so in that case, why bother with the external preamp?

-a

Hi Andy,

You are right but there are two problems.  One is that it is out of the ordinary for them and the other is that there may actually be some connection problems on the TT24 with the TRS inputs.  I'm not sure if they are soldered directly to the PCB board but there have been some intermittent problems which is why they stick with the XLR.

Although there is a longer explanation why the output impedance starts seeing a lower impedance input ultimately it is that there is a lower impedance in the area of probably 600 to 800 ohms depending on the final details of it all.

So... from what Abbey Road is saying it is probably not a problem (especially if the output impedance of my device is under 100 ohms) with driving a 600 to 800 ohm load.  That still fits in the 1:10 rule too.

But.. from what I gather Abbey Road - a 1:10 rule must be maintained if an output transformer is involved?

CC

 

[JohnRoberts]

You seem to be making this more complicated than it is.

If your output can drive 600 ohms, and two mic inputs in parallel are only 1k, and you are not overloading the mic preamps, you are home free.

(As I already told you) The 1:10 criteria for a bridging termination is a guideline to attain mostly voltage transfer. Your 100:1k meets that criteria, but even if it didn't and you lost an extra dB or two of signal level, is that really a problem in a mic input with several tens of dB gain available???

Termination and loading for transformers is a whole different can of worms and probably not worth opening up unless you actually need to use a transformer to provide isolation.

JR

 

[conleycd]

You seem to be making this more complicated than it is.

If your output can drive 600 ohms, and two mic inputs in parallel are only 1k, and you are not overloading the mic preamps, you are home free.

(As I already told you) The 1:10 criteria for a bridging termination is a guideline to attain mostly voltage transfer. Your 100:1k meets that criteria, but even if it didn't and you lost an extra dB or two of signal level, is that really a problem in a mic input with several tens of dB gain available???

Termination and loading for transformers is a whole different can of worms and probably not worth opening up unless you actually need to use a transformer to provide isolation.

JR

John, - thank you for your answer.  This was the confirmation that I was looking for.  I don't think I am actually making things more complicated  - I, for the most part understand the rules, but you have now confirmed, "even with some signal loss" the 1:10 rule can be broken if your output stage is capable of driving the load (with some possible signal loss). 

When you scroll through this thread there were many suggestions and work arounds offered but only you and Abbey Road really answered the question.  Many of the examples still fit the 1:10 rule but you have provided what I was looking for!

Now... onto Transformer Termination - I wish there was a Meta on this because I can't find very much information.  I don't need the info for this current project but I have a bunch of things coming with output transformers to be strapped on...

Does anyone know of a thread (I've searched) containing information about adding output transformers, impedance, and proper termination?

CC

 

[sodderboy]

Just plug it in.  Engineers split anything- on a patchbay is is called a MULT- 3 to 6 points wired in parallel. 
As far as the church politics go, it sure is a losing game.  If it sounds OK going line out to mic in, keep the gain down on the pre and start the service.  Most church sound systems have low hum in them because a computer is incorrectly connected.  Nobody notices, just like the boutique with speakers out of phase.  You mention something and you might as well be speaking mandarin.
Check the transformers meta.  There is much info there.
Mike

 

[lassoharp]

It seems as though you have the answers you need but for the sake of clarity for anyone following this thread I'm still not clear on the exact setup you're trying to solve interface problems for.

Ok. You need to run a multed line level signal into the XLR mic amp inputs on a MackieTT console.  What exactly is your source here?  How does it fit in with the overall goal of what is trying to be achieved?  Mic split for recording, CD for house mix? Those things were unclear.

You did ask about resistive splitting so I will offer an answer on that general question though I can't say it's going to be the best way of handling your specific situation.

This info can be referenced from the audio cyclopedia pgs 239, 422, 419 ( parallel dividing network ). If you don't have that reference handy about the closest I have on hand at the moment is this:

http://s409.photobucket.com/albums/pp179/lassoharp/?action=view&current=Coil_mixer_schematic.jpg&newest=1

This is showing a simple 2 way split from a single source( X4 ). Ignore the transformers at the end and the 900ohm resistor.  The 200ohms are the buildout resistors - one from the source(cut in half and apply to hi and lo side for balanced) and one(or two for balanced) for each branch.

To figure out what resistor you need for your particular source imp., the formula is:

Rb = (N-1/N+1) Z,  Z is your source imp and N is simply the # of branches you wish to connect.

Note that this is a constant impedance network and the load is expected to be the same as the source. In your case the load will not be the same as the source so what do you do?

The answer to this will also roughly be the answer to your question on transformer termination.

You must terminate the load end of each network branch with an appropriate resistor so that when paralleled with your actual load will give the nominal impedance(the value of your source imp) of the network. Same principle with transformer termination.

The loss for a 4 way split is ~10db.  2 way should be well within the makeup gain range of the Mackie amps.