balanced summing question

Simple question: in a passive balanced summing network connected to a balanced virtual ground summing amplifier,
the channel mute switch,  got to short the phases of the unused channel each other, live the channel unconnected or just live the channel unconnected living the phases floating ?

thank you

M.

mik said:

Simple question: in a passive balanced summing network connected to a balanced virtual ground summing amplifier,
the channel mute switch,  got to short the phases of the unused channel each other, live the channel unconnected or just live the channel unconnected living the phases floating ?

thank you

M.

Click to expand...

Perhaps the question is too simple for me.  A "passive" summer does not generally have a virtual ground (earth?) summing amplifier.

Can you post a schematic of what you mean?

JR

thank you J.
here it is , it's a simplified schematic for what I mean,
I hope this is clear enough.

first channel shorts the phases,  the others  just not. what's the best way to go ?

M.

OK, I now understand the circuit.

mik said:

Simple question: in a passive balanced summing network connected to a balanced virtual ground summing amplifier,
the channel mute switch,  got to short the phases of the unused channel each other, live the channel unconnected or just live the channel unconnected living the phases floating ?

thank you

M.

Click to expand...

The answer is it depends.

If you short the two sends together, and float them, there should be full signal attenuation, while the bus amp will see the same noise gain as with signal present (perhaps not a big concern).  If the now floating send sees and signal coupling from nearby signals, that pickup should be common mode so cancel.

If you leave both sends floating the bus noise gain will be lower, but pickup from nearby signals into just one or the other may be worse.

Another consideration from opening/shorting the sends as this may cause small DC level steps (clicks) if the DC is not 0V.  In general we want mutes to be click free, so this may be a larger concern than bus noise gain.

JR

thank you John !  after your precious considerations
to me the best way is to schorting the sends, becouse channels are on PCB leads by Dsub25, they are to close each other, and lives room for cross talk. what do you think ?
P.S.
DC cliks are not an issue, because this is a "set and forget" device there's not a mute on the fly operation during mixdown.

M. 

mik said:

to me the best way is to schorting the sends, becouse channels are on PCB leads by Dsub25, they are to close each other, and lives room for cross talk. what do you think ?

Click to expand...

That would be a concern only if there was one source that you would not want at all in the mix. There are not many situations where you have that. It happens in broadcast when the candidate must not hear the answer to a question, it happened in ancient times when one track was carrying Time-Code.
The way hardware mixers are used today, the signals come from a DAW and offending signals are killed in the DAW.
Regarding possible clicks due to DC offset, a pair of capacitors of suitable value between the resistor ladder and the summing amp inputs should answer that.

I see...
thank you John !

mik said:

thank you John !  after your precious considerations
to me the best way is to schorting the sends, becouse channels are on PCB leads by Dsub25, they are to close each other, and lives room for cross talk. what do you think ?
P.S.
DC cliks are not an issue, because this is a "set and forget" device there's not a mute on the fly operation during mixdown.

M.

Click to expand...

To keep crosstalk to a minimum shorting the two legs together and to the local input ground should keep that small. I notice you are already using 1% resistors that will help with differential cancellation.

If clicks don't matter DC coupling is fine.

This is a personal preference of mine, but I would be tempted to add a differential stage after the two bus amps to cancel out the common mode. If you want differential (2 legged) output you can add another inverting stage, or just use two symmetrical differential amps cross connected. Many modern audio applications don't need +26dBu level. This will keep any common mode noise/errors inside the box, so cancellation does not have to depend on the following stage's differential quality or wiring integrity. Further a differential can reference the signal between the bus amp ground and the output ground, while a minimalist design could just connect the two bus sum amps to output ground.

FWIW from my experience with larger consoles, input ground, channel ground, bus amp ground, and output ground are often different voltages.

JR 

JohnRoberts said:

This is a personal preference of mine, but I would be tempted to add a differential stage after the two bus amps to cancel out the common mode. If you want differential (2 legged) output you can add another inverting stage, or just use two symmetrical differential amps cross connected. Many modern audio applications don't need +26dBu level. This will keep any common mode noise/errors inside the box, so cancellation does not have to depend on the following stage's differential quality or wiring integrity. Further a differential can reference the signal between the bus amp ground and the output ground, while a minimalist design could just connect the two bus sum amps to output ground.

FWIW from my experience with larger consoles, input ground, channel ground, bus amp ground, and output ground are often different voltages.

JR 

Click to expand...

yes two symmetrical differential amps cross connected it's what I have in minde

thank you.