balanced vs unbalanced mixer

Hi,

I was wondering - according to doug self a simple unbalanced line input is much less noisier then a balanced, at least without quite some effort for the balanced input.
would it be worth trying unbalanced inputs  in my 12 + 4 channel mixer project?  the outside connections are very controlled, short cable runs from and to audio interface, so the possibility of picking up hum and noise is relatively 'under control'.

adding a balanced output on the other hand might be still a good idea to drive the line input.

any thoughts on this?

cheers,

Michael

audiomixer said:

Hi,

I was wondering - according to doug self a simple unbalanced line input is much less noisier then a balanced, at least without quite some effort for the balanced input.

Click to expand...

Yawn...  While perhaps accurate for the unrealistic example of a shorted or no signal input, in practice a good quality balanced input will be lower noise than most real world signal sources. So the modest noise is insignificant in the context of the typical real benefit from balanced (differential) inputs. 

would it be worth trying unbalanced inputs  in my 12 + 4 channel mixer project?  the outside connections are very controlled, short cable runs from and to audio interface, so the possibility of picking up hum and noise is relatively 'under control'.

adding a balanced output on the other hand might be still a good idea to drive the line input.

any thoughts on this?

cheers,

Michael

Click to expand...

yes, the noise benefit of operating unbalanced is imaginary in practice, with real signal sources, while the potential for signal degradation from unbalanced (not differential) interfaces is real. 

JR

also, if you have a device plugged into a wall socket that is on another circuit, it is nice to have the isolation just in case something weird happens, and something weird will happen if you get enough unmarked patch cables dangling all over the place, :'(

hey they have some Fernwood Tonight reruns on youtube, beats mary hartman mary hartman, eh?
dont forget about America Tonight if you run out of Fernwood Tonight stuff, Tony Rolettti, Virgil Simms, all the big stars,  tedesco and devol as happy kine, 

anybody catch the Wrecking Crew movie that is touring?

youtube has snippits

Cj,

your post made me think further: I was thinking of the inputs from my audio interface only, the desk is for mixdown in a controlled environment, not live / studio tracking etc.... so the additional equipment will be selected / limited. but it was more about the theoretical aspects. I have started to measure around in my first installment of the mixing desk, so many questions come up. I need to do some measurements on a quality desk....

- michael

audiomixer said:

.. would it be worth trying unbalanced inputs  in my 12 + 4 channel mixer project?  ...  adding a balanced output on the other hand might be still a good idea to drive the line input.

Click to expand...

There are NO advantages for a balanced output unless there is a corresponding balanced input.  There are also a alot of REALLY bad balanced outputs, some of which go haywire if feeding unbalanced.

But a balanced input solves many problems and is useful even (??) if all your sources are unbalanced.  Ever had an earth problem?  This is FAR more important than theoretical noise performance.

http://sound.westhost.com/articles/balanced-2.htm An excellent summary of Bill Whitlocks encyclopediac articles. Originals are on the Jensen website.
http://www.douglas-self.com/ampins/balanced/balanced.htm
http://www.uneeda-audio.com/zbal.htm

> inputs from my audio interface only,.. in a controlled environment,... equipment will be selected / limited.

Then put 8 resistors off the innerface and mix unbalanced.

Balanced is only about rejecting the Universal Humm/Buzzz, NOT random hiss. If you have your hum/buzz under control, unbalanced may be simpler and thus more transparent.

ricardo said:

There are also a alot of REALLY bad balanced outputs, some of which go haywire if feeding unbalanced.

Click to expand...

Amen! I have experienced this in so called "pro" modern gear, to which my response has always been to immediately sell the gear.

The widely used 192HD from Avid/Digidesign has a "brain dead" "balanced' output which doesn't operate correctly into unbalanced loads.  I would think, then, that thousands of those boxes should be sold! 

Bri

As Ricardo observed, for short feeds there is absolutely no reason to pursue a balanced send. In fact there is little reason to pursue it for 95% of pro applications, but marketers like the extra 6dB of output swing on their spec sheets.

OTOH IMO there is much potential benefit from differential inputs.

JR

ricardo said:

There are NO advantages for a balanced output unless there is a corresponding balanced input. 

Click to expand...

Huh? I believe there are advantages, just the same as when you say: "But a balanced input solves many problems and is useful even (??) if all your sources are unbalanced.". The reverse is also true.  In the process of eliminating noise generated by differential ground voltage, having one side "ground-conscious" is enough. So, I would agree that a non-floating balanced output needs a balanced input, but that's a rarer and rarer situation.
Now, in practical terms, the inclusion of EBOS in a piece of gear has such a limited monetary penalty that is would be ridiculous not to do it. If not confident about their ability for designing a good EBOS, one can always use a THAT 1646.
Having balanced floating or ground-sensing outputs is the only way to prevent hum loops with gear that has unbal inputs.

abbey road d enfer said:

ricardo said:

There are NO advantages for a balanced output unless there is a corresponding balanced input. 

Click to expand...

Huh? I believe there are advantages  .. one can always use a THAT 1646. ..
Having balanced floating or ground-sensing outputs is the only way to prevent hum loops with gear that has unbal inputs.

Click to expand...

Abbey, the 3 links in my post discuss this in great detail.  There is also the huge discussion where I was crucified in http://www.proaudiodesignforum.com/forum/php/viewtopic.php?f=6&t=464

But in fact that discussion shows that of the EBOSs, only the THAT devices provide benefit with unbalanced inputs.  The other EBOSs are seriously flaky into unbalanced inputs; especially with some cable in between.

And I'm not sure that 'non-floating balanced output's (ie c**p ones) are rarer ..  judging from the comments here. 8)

But I thank you for the opportunity for me to point out one can do an excellent balanced output with only 2 resistors  ;D

ricardo said:

But in fact that discussion shows that of the EBOSs, only the THAT devices provide benefit with unbalanced inputs.  The other EBOSs are seriously flaky into unbalanced inputs; especially with some cable in between.

Click to expand...

I need to read in details, because apart from the uncircumstanced assertion:  "The cross-coupled output... Unlike the transformer, this circuit demands that it be unbalanced at the source. Heaven help you if there is a pin 1 problem designed into the unit’s output connector. Unbalancing at the load causes instability." I don't see any proof of it. It is certainly not in accordance with the numerous field cases I witnessed. I agree that some of these EBOS can produce distortion when unbalanced, but it's generally less bothering than constant hum.

And I'm not sure that 'non-floating balanced output's (ie c**p ones) are rarer ..  judging from the comments here. 8)

Click to expand...

Yes, it's amazing how the higher the market share, the more penny-pinching.

I will get some more opamps and report back..... ;-) I'm going to set up the equivalent four mixing channels in unbalanced configuration and see how it performs. I'm not sure what type of output my focusrite audiointerface has, will check that too. of course a later update of the audio interface might change the game again.

it seems we all have a (marketing guided?) view on this....  balanced is better, always and for ever, but please be aware that I am referring to a specific case with my audiointerface on the same power circuit, short (<3m) cable runs, no other equipment connected  an so forth and not a full blown mixing desk to be released into the wild.

-cheers,

Michael

audiomixer said:

it seems we all have a (marketing guided?) view on this.... 

Click to expand...

I know at least a half-dozen persons on this forum who have  very circumstanced views on the matter. With scientific proof and years of practice.

but please be aware that I am referring to a specific case with my audiointerface on the same power circuit, short (<3m) cable runs, no other equipment connected  an so forth and not a full blown mixing desk to be released into the wild.

Click to expand...

I believe we all understand that, but you don't realise how much mischief some of these boxes actually pack.  ;D
Your approach is good, as long as you're ready to rebuild /modify when problems surface.

abbey road d enfer said:

...  from the uncircumstanced assertion:  "The cross-coupled output... Unlike the transformer, this circuit demands that it be unbalanced at the source. Heaven help you if there is a pin 1 problem designed into the unit’s output connector. Unbalancing at the load causes instability." I don't see any proof of it.

Click to expand...

Abbey, this is fairly common knowledge among the better mixer designers who started using the SSM & TI devices in the 80s & 90s.  The proaudiodesignforum.com link I posted has comprehensive test results  ..  at least for the overload behaviour.  I have a hand drawn circuit and notes of a cross-coupled output which I dreamt up circa 1990 when I found the instability with SSM/TI.  Dunno if you'd accept that as 'proof'.  ;D

Michael, whatever works for you is good.  I only commented cos you were going to have balanced outs and unbalanced ins.  To me, seems the wrong way round.  But if you are going to have balanced outs, the ones to use are

• my 2 resistors (cheapest) but good if you have a suitable OPA with the right levels, drive capability etc
• THAT16xx
• big Lundahl / Sowter / Jensen with suitable driver

I don't think anything else is worth the PCB space.

ricardo said:

abbey road d enfer said:

...  from the uncircumstanced assertion:  "The cross-coupled output... Unlike the transformer, this circuit demands that it be unbalanced at the source. Heaven help you if there is a pin 1 problem designed into the unit’s output connector. Unbalancing at the load causes instability." I don't see any proof of it.

Click to expand...

Abbey, this is fairly common knowledge among the better mixer designers who started using the SSM & TI devices in the 80s & 90s.

Click to expand...

Although I have been heavily involved in mixer design in the 70's, 80's and 90's, I have never used these parts. They were not up to the level of performance we were looking for. We used only CCOS or Strahm-type. Only the CCOS went into production, the Strahm's needing too much adjustment in production.

The proaudiodesignforum.com link I posted has comprehensive test results

Click to expand...

I've read the 20 pages of it, I don't see anything conclusive. The last series of measurements done by Wayne indicate they're all in the same order of magnitude; the 1646 is the best overall however. To day, I use only them.
And the oscillations appear only under heavy overload, where distortion is surely the most annoying artefact.

...the ones to use are

• my 2 resistors (cheapest) but good if you have a suitable OPA with the right levels, drive capability etc
• THAT16xx
• big Lundahl / Sowter / Jensen with suitable driver

I don't think anything else is worth the PCB space.

Click to expand...

You seem to underestimate the ground-compensating output. It's remarkable that, on the mediatechniology forum, it has been considered as a non-tested curiosity with no real-life applications. I can testify to the contrary, as the Soundcraft engineers who used them (and probably continue to use them) in thousands of Spirit mixers, but also the Soundcraft 500/600 and 6000 series.

OT, I see that you have a real talent for being chastised on the mediatechnology forum, on par with my talent for being chastised on a well-known and reputable professionel recording forum.

audiomixer said:

I will get some more opamps and report back..... ;-) I'm going to set up the equivalent four mixing channels in unbalanced configuration and see how it performs. I'm not sure what type of output my focusrite audiointerface has, will check that too. of course a later update of the audio interface might change the game again.

it seems we all have a (marketing guided?) view on this....  balanced is better, always and for ever, but please be aware that I am referring to a specific case with my audiointerface on the same power circuit, short (<3m) cable runs, no other equipment connected  an so forth and not a full blown mixing desk to be released into the wild.

-cheers,

Michael

Click to expand...

As Abbey noted, many of us here are not posting from "marketing guided" views.

To clarify my position:

A) 3 circuit differential inputs that allow signals to be carried independently of ground deliver the highest signal integrity between grounded, powered chassis.

B) "balanced" 3 wire differential interfaces only delivers additional benefit when both sends and returns and all wiring in between is balanced wrt ground. 

C) Balanced impedance sends are cheap (one added resistor) and deliver almost all of the possible benefit from a "balanced (dual polarity) signal" send.

For your application if you have a very controlled environment, where you know the exact gear you will "always" be connecting together, you may get away with a KISS interface.  But this simple approach may stop working at some future date when you introduce some new different piece of equipment (to input or output side). 

While such discussions often veer into details about what doesn't work, and much bloviation, do not dismiss advice from experience gained by doing similar tasks over several decades. 

We surely agree on far more than we disagree about, despite appearances, because we have all dealt with the reality of making such interfaces work.

JR

PS: Like Abbey, I never used the off-the-shelf interface chips, when an opamp and precision resistors could deliver similar benefit. It is a fairly recent development that the standard I/O chip sets are worth consideration.

abbey road d enfer said:

ricardo said:

Abbey, this is fairly common knowledge among the better mixer designers who started using the SSM & TI devices in the 80s & 90s.

Click to expand...

Although I have been heavily involved in mixer design in the 70's, 80's and 90's, I have never used these parts. They were not up to the level of performance we were looking for.

Click to expand...

+1 to that.  IMHO, THAT16xx are the first to be acceptable.

You seem to underestimate the ground-compensating output. It's remarkable that, on the mediatechniology forum, it has been considered as a non-tested curiosity with no real-life applications. I can testify to the contrary, as the Soundcraft engineers who used them (and probably continue to use them) in thousands of Spirit mixers, but also the Soundcraft 500/600 and 6000 series.

Click to expand...

If you look carefully between the 20 pages of "kreuzigen!", you'll see I attempt to ask if anyone has tested this.  From what you say, it now appears this is a favourite of Self who did a lot of work for Soundcraft.

I know if Wayne or Self pontificate on distortion or response, their prophecies are based on extensive tests and can be taken on trust.  My prejudices lie more towards bombproof RFI immunity (far beyond EU or other requirements).  I have to admit some disappointment with Self from looking at his Soundcraft mike preamp stage.  I don't know how extensively he tests for EMI.  I thought Eric Benjamin of Dolby used this output circuit but I was mistaken.  He is another paranoid about RFI immunity.

I'll now put 'ground-compensating' in the 'possibly worth trying' category as it uses less real estate than THAT16xx and has some advantages over '2 resistors'.

But to put things into (my) perspective, my current anguish on Balanced O/Ps is whether to go for

• low Zo (as an old BBC spec: 40R +/- 10R < +/- j35R 10Hz - 15kHz) which has known good RFI immunity OR
• 90R to match Belden twisted pairs, better THD & reflections on long lines from Wayne's tests

To answer this to my satisfaction, I need to build another Calrec CB Radio (details in the proaudiodesignforum thread) and confirm my reputation as a crazy beach bum in Cooktown.    As this is unlikely, I'll probably never test ground-compensating properly either.

In the meantime, I'll wuss out & specify 2x36R for my 2 favoured 'electronic' solutions. 

ricardo said:

... ground-compensating output...

Click to expand...

If you look carefully between the 20 pages of "kreuzigen!", you'll see I attempt to ask if anyone has tested this.  From what you say, it now appears this is a favourite of Self who did a lot of work for Soundcraft.
I have to admit some disappointment with Self from looking at his Soundcraft mike preamp stage. [/quote] I'm not sure what mic pre you are referring to... to my knowledge, the Soundcraft elements that bear the Self signature are his patented active pan-pot, several power supplies and a noise-gate. He has certainly had input on several other items, but I think he was more consulting for other designers in a team effort. I don't think the ground-compensated thingy was his baby. And anyway dbx came earler with the idea. There were some brilliant engineers!

I don't know how extensively he tests for EMI.

Click to expand...

Neither do I; at the time he was at Soundcraft, they didn't have an EMI/RFI facility. 

...I'll probably never test ground-compensating properly either.

Click to expand...

Don't regret it; it works as described with all the limitations due to cables and contacts not being pure resistors. It's good enough for most installations with only moderate RFI and "humane" ground loops.

abbey road d enfer said:

" I have to admit some disappointment with Self from looking at his Soundcraft mike preamp stage."
I'm not sure what mic pre you are referring to... to my knowledge, the Soundcraft elements that bear the Self signature are his patented active pan-pot, several power supplies and a noise-gate.

Click to expand...

In his web pages on LN design, (before he took them off to sell more books) he claims credit for the Soundcraft preamp stage which is also used for their virtual earth mixers.

There was a big thread just a while ago on modifying a Soundcraft preamp circuit, removing evil caps, servo etc.  One of the other ex-Soundcraft designers chimed in with the genesis of that design including Self's involvement.

And anyway dbx came earler with the idea. There were some brilliant engineers!

Click to expand...

While I have the very highest regard for David Blackmer, there are some awful balanced INs on early dBX stuff too.