Converting unbalanced direct outs to balanced

[Michael Tibes]

...I spoke to Allen & Heath directly on the matter and they actually used to offer a balancing kit for the console.  Looks like a really simple and inexpensive circuit:

http://www.allen-heath.com/media/EBOS-BALANCED-OUTPUT.pdf

Am I overlooking it or is this not even cross-coupled, so that it acts 'Transformer-like'? I believe if one of the outputs gets grounded (connected to an unbalanced input), it will be shorted and therefore distort. The solution with the That (and other manufacturers) chips doesn't do that, there will certainly be papers at thatcorp explaining the matter. If you absolutely want to balance it, I'd use one of the chips dedicated to the task, they are pretty good. Though in my opinion the only advantage of balancing the outputs vs impedance balancing is a roughly 6 dB higher clipping point, given the same powersupply voltages. I'd always go with impedance balancing unless the headroom is needed. GML gear uses it and as far as I remember the SSL 4/6k worked the same way on the inserts.

Michael

 

[psb_87]

Am I overlooking it or is this not even cross-coupled, so that it acts 'Transformer-like'? I believe if one of the outputs gets grounded (connected to an unbalanced input), it will be shorted and therefore distort. The solution with the That (and other manufacturers) chips doesn't do that, there will certainly be papers at thatcorp explaining the matter. If you absolutely want to balance it, I'd use one of the chips dedicated to the task, they are pretty good. Though in my opinion the only advantage of balancing the outputs vs impedance balancing is a roughly 6 dB higher clipping point, given the same powersupply voltages. I'd always go with impedance balancing unless the headroom is needed. GML gear uses it and as far as I remember the SSL 4/6k worked the same way on the inserts.

Michael

I expect you're right, as this design from A & H is very old now.  Would it simply be a case of just replacing an IC from their design with one from That?

 

[Newmarket]

Hey everyone, I greatly appreciate all of your input and informative replies.  I spoke to Allen & Heath directly on the matter and they actually used to offer a balancing kit for the console.  Looks like a really simple and inexpensive circuit:

http://www.allen-heath.com/media/EBOS-BALANCED-OUTPUT.pdf

I guess I will just hook everything up when I eventually get my hands on it and see what happens.  If I need to balance the direct outs then at least I can use this schematic to do so.

Good support from A&H. But I'd say it's actually a bit more complicated than I'd  do it myself. More op amp stages.
I'm not going to analyse it all but I guess they are trying to minimise the dc offset out of the 5532 output stages since there's no output caps or 'normal' servo.
22R build out resistors look a bit marginal to me although will probably be fine with cable lengths likely.

 

[Newmarket]

Am I overlooking it or is this not even cross-coupled, so that it acts 'Transformer-like'? I believe if one of the outputs gets grounded (connected to an unbalanced input), it will be shorted and therefore distort. The solution with the That (and other manufacturers) chips doesn't do that, there will certainly be papers at thatcorp explaining the matter. If you absolutely want to balance it, I'd use one of the chips dedicated to the task, they are pretty good. Though in my opinion the only advantage of balancing the outputs vs impedance balancing is a roughly 6 dB higher clipping point, given the same powersupply voltages. I'd always go with impedance balancing unless the headroom is needed. GML gear uses it and as far as I remember the SSL 4/6k worked the same way on the inserts.

Michael

You are correct as I read the schematic. If feeding an unbalanced input you need to leave one of the outputs unconnected otherwise...one hot opamp  :-\
Another alternative is a 'Ground Cancelling' output. Used on (among others) several Dave Dearden DDA designs for Aux Sends etc. Quite economical componentwise.

 

[Newmarket]

I expect you're right, as this design from A & H is very old now.  Would it simply be a case of just replacing an IC from their design with one from That?

No. But it would be simpler - only one IC and fewer external components. Take a look at the datasheet / application info. Thatcorp.

 

[Michael Tibes]

You are correct as I read the schematic. If feeding an unbalanced input you need to leave one of the outputs unconnected otherwise...one hot opamp  :-\
Another alternative is a 'Ground Cancelling' output. Used on (among others) several Dave Dearden DDA designs for Aux Sends etc. Quite economical componentwise.

It will not only run hot, some amount of the distortion will find it's way into the signal pass. The That or TI stuff is a better version of what  DDA did, if you want an electronically balanced output you probably can't do it any better than that. If you really want to get into it you can find a lot of papers at thatcorp.com, also comparing their design with others. I'd probably still simply impedance balance the stuff and go on making music ;-)

Michael

 

[psb_87]

No. But it would be simpler - only one IC and fewer external components. Take a look at the datasheet / application info. Thatcorp.

Simple is good.  Okay cool, so the first product I found should do the job then? I could replicate that design and incorporate into my desk if needs be.  Is this circuit universal or would I need to adapt the values to the Saber?

Here is the product I'm referring to:
https://www.diyrecordingequipment.com/products/balanced-output

And here is the schematic and build instructions:
https://www.diyrecordingequipment.com/pages/balanced-input-output-assembly-guide

Thanks for your help

 

[Newmarket]

Simple is good.  Okay cool, so the first product I found should do the job then? I could replicate that design and incorporate into my desk if needs be.

Yes you could - but a single resistor from the cold leg to ground to impedance balance is simpler !

 

[Newmarket]

It will not only run hot, some amount of the distortion will find it's way into the signal pass. The That or TI stuff is a better version of what  DDA did, if you want an electronically balanced output you probably can't do it any better than that. If you really want to get into it you can find a lot of papers at thatcorp.com, also comparing their design with others. I'd probably still simply impedance balance the stuff and go on making music ;-)

Michael

Yes indeed - not just hot - all the distortion too !

re the DDA stuff - I guess you are referring to the cross balanced type circuit topology used by DDA and others and similar to the ICs  mentioned. I recall DDA had the circuit on a 'module' that plugged into the circuit - EBO (electronically balanced output) - and the trimpot setting gets all important wrt CMRR as the circuit topology is sensitive to component tolerance.
I was refering to the 'Ground Cancelling' or 'Ground Sensing' - the terminology is variable - where it can also work effectively into an unbalanced input with appropriate wiring. eg DDA DMR12 Direct Out. ( whereas 'MAIN' Outs from the desk are Cross Balanced EBO type)
Guess I should say I worked on automation / interfacing etc for DDA amongst others so some specific knowledge but the circuit types are widespread.

 

[JohnRoberts]

Hey everyone, I greatly appreciate all of your input and informative replies.  I spoke to Allen & Heath directly on the matter and they actually used to offer a balancing kit for the console.  Looks like a really simple and inexpensive circuit:

http://www.allen-heath.com/media/EBOS-BALANCED-OUTPUT.pdf

I guess I will just hook everything up when I eventually get my hands on it and see what happens.  If I need to balance the direct outs then at least I can use this schematic to do so.

Hi John, thanks for the great amount of information and for correcting some misinformation.  I'm interested in the high end consoles you mention here that have unbalanced direct outs?

The list is far shorter for consoles that do use active balanced for DOs and/or inserts. IIRC it was an option on some Crest or APB consoles. 

I repeat use care in your wiring between console and destination. When passing the signal with only 2 wires do not corrupt one of those two wires with noise/ground current.

see  http://www.rane.com/note110.html  #13

JR

 

[JohnRoberts]

One example, that may relate to your scenario, is that I have a compressor that has unbalanced and balanced inputs/outputs.  The compressor, as many pieces of gear are, is really unbalanced internally. Meaning, it takes the balanced input and unbalances it and....takes that unbalanced signal and re-balances it on the way out. This particular compressor lets me choose if I want to use the unbalanced jacks for input and output so, it doesn't have to go through these balancing/unbalancing stages. It just stays where it is.

The extra stages are where the signal transparency can be affected. Of course this is subjective but, when I use the balanced outputs of this compressor, the balancing circuit adds 2 op amps, a number of large and small capacitors and some resistors to make it balanced. The signal bypasses this balancing stage if I use unbalanced and I hear a difference that is not subtle.

=====
It is a rabbit hole so, be aware that there are considerations when sending your balanced gear unbalanced signals and especially vice versa...

Good luck...

This is probably TMI and unlikely to help the OP, but back last century I influenced the feature set of a tube preamp (AMR/Peavey VMP-2) to add a high Z unbalanced output that bypassed the low Z balanced output transformer.  My goal was to simply eliminate the output transformer from the audio path for short distance studio interfaces that didn't need balancing.

I won't make claims about audible superiority but removing the transformer was slightly cleaner on the test bench. YMMV and some people add transformers because they want some transformer "sound" (not me).

JR
 

 

[ruairioflaherty]

This is probably TMI and unlikely to help the OP, but back last century I influenced the feature set of a tube preamp (AMR/Peavey VMP-2) to add a high Z unbalanced output that bypassed the low Z balanced output transformer.  My goal was to simply eliminate the output transformer from the audio path for short distance studio interfaces that didn't need balancing.

I won't make claims about audible superiority but removing the transformer was slightly cleaner on the test bench. YMMV and some people add transformers because they want some transformer "sound" (not me).

JR

I used a Manley Dual Mono Tube Mic Pre for many years and always preferred the unbalanced output for similar reasons. 

 

[psb_87]

Thanks everyone for your discussion on this, I really appreciate it.

So I finally got the desk and I've gone through every channel.  There's only a couple of issues which should be an easy fix.  The entire master section however doesn't seem to be functioning properly, which is a bit of a shame.  I don't seem to be getting anything on the main L&R no matter what I route there.  There's also a low oscillating buzz/hum when monitoring through the headphones, which I'm guessing would also be on the main monitors L & R but I am yet to check.  I took the master section out to see if there was anything obvious and I found 1 single burnt out resistor, and some magic smelly smoke appeared at one point... it smelt like horse manure.  I am not sure what it was that fried, I don't think it was the resistor as the smoke appeared at the top of the strip and the resistor is towards the bottom.  Any ideas on what I should try to get this to work? (I know, there's probably a million things it could be)

Theres a few channels which have their peak light on constantly even though they are fully functioning. 

One channel's pan is not functioning as well as the high band EQ, though the channel is working other than that.

Also on the first eight channels the mutes are weird.  Activating the mute on channel 3 also activates the mute on channel 1, but activating the mute on channel 1 does nothing.  The same with 2 &4, 5 & 7, 6 & 8.  Any ideas? The other mutes are all working fine.

I also discovered that the desk is not a Saber 8 which the owner originally told me, it's actually a Saber mk1 or mk2 or plus.  I don't know how to tell.

All the groups are working, all the meters are working and all the channels pass audio when in 'check' mode.  They are also surprisingly quiet in terms of noise, which is nice.

Any info or help that you guys can offer would be fantastic, here's a pic of the desk:

 

[scott2000]

Nice Desk....

You could always get a decent LCR meter and start checking for the obvious faulty components and see if that gets you anywhere? I've been pretty lucky doing this when some of my stuff has acted up....
. It's pretty quick and painless......

Getting a schematic will be very helpful if not necessary at some point....

Good Luck!

 

[psb_87]

Yes you could - but a single resistor from the cold leg to ground to impedance balance is simpler !

Hi Newmarket, sorry for the late reply but I am pretty interested in doing this to my DOs.  Any idea what resistor value I should use? I've read that it's supposed to be the same as the 'build-out resistors' but am not even sure what they are or where to find them lol.  Any help would be much appreciated

 

[psb_87]

Yes you could - but a single resistor from the cold leg to ground to impedance balance is simpler !

Would it be this one marked in green?

 

[Rob Flinn]

Would it be this one marked in green?

No, because it's not "connected to ground" & it is connected to hot (tip or pin 2).    What you need is a resistor from GROUND to the cold (ring or pin 3).

 

[psb_87]

No, because it's not "connected to ground" & it is connected to hot (tip or pin 2).    What you need is a resistor from GROUND to the cold (ring or pin 3).

Hi Rob, I think there's been a misunderstanding.  I'm asking for someone to help me identify the 'build-out resistor' already present in the circuit.

I can post the entire schematic if needed.

Many thanks,
Paul

 

[keefaz]

Signal goes from IC4 out (impedance like 0R), then goes through R87 22R to direct out.
So output impedance is 22 + 0 = 22 ohms.

Remove the output jack socket ring / sleeve connection, solder a 22 R resistor between ring and sleeve, then you have impedance balanced out

 

[Rob Flinn]

Hi Rob, I think there's been a misunderstanding.  I'm asking for someone to help me identify the 'build-out resistor' already present in the circuit.
I can post the entire schematic if needed.

Many thanks,
Paul

Right, gotcha.