help me to understand: bal to unbal, 6 bd loss?

hi guys. i'm wondering:
i heard that going from balanced to unbalanced at unity gain produces a loss of 6db volume.
is it correct?
but mainly, why?
so a -10dbv unbal signal has a 17db difference compared to a +4 dbu bal signal?

thanks guys

here a short, not very in depth, answer.
A balanced connection has two wires carrieng the signal. These are beeing summed which gives you the +6db.
When using only 1 wire you don't get the 6db boost. You can call this a loss of 6dB...

thanks. now it makes sense.
so if the signal gets balanced at unity gain it's going to get a 6db boost, right?

It depends very much of the way the in/outs are balanced. If they are floating (like transformer-balanced), connecting just one leg leads to complete loss of signal. But if you go unbal with xfmr connections, you don't loose anything.
It's only with poorly designed electronically balanced circuits that you loose 6 dB when disconnecting one pin.

If the circuit loses 6 dB when becoming unbalanced it might not be a good idea to tie outphase to gnd - like you'd normally do. It might short the outphase out to gnd and cause distortion. A 'good' circuit would sense that and automatically raise it's gain by 6 dB while the outphase out is still under control. I wonder why such circuits are even still being used while decent ics or opamp implementations are around for 20 or more years by now?

Michael

From -10dBV to +4dBu (assumed re:dBm in 600ohms) is just about 11.78dB. Usually rounded to 12.

Either level may be sent/taken balanced or unbalanced.

It is customary to have cheap unbal -10dBV and expensive bal +4dBm/u, but that's not required by the level spec.

Any other answer is about the flaws in balanced outputs.

While a true balanced active is possible, it isn't that easy, and most meter-pegging users don't know any difference.

Hello guys,
I m using an RME converter that automtically compensates the 6db loss when used unbalanced
http://www.rme-audio.de/download/adi8ds_e.pdf

  I guess a kind of switch is used when the negative pin is grounded,
Do you guys have ideas of such a circuit maybe using CD4053 or other switchs??

saint gillis said:

Hello guys,
I m using an RME converter that automtically compensates the 6db loss when used unbalanced
http://www.rme-audio.de/download/adi8ds_e.pdf

  I guess a kind of switch is used when the negative pin is grounded,
Do you guys have ideas of such a circuit maybe using CD4053 or other switchs??

Click to expand...

No switch.. there are active balanced outputs with cross connected feedback that set level based on the sum of both outputs. When one leg is shorted, the level on the other leg increases to make up the difference.

These outputs can give errors when both outputs are not read by following inputs.

JR

There is no 6dB loss when you go from balanced to unbalanced. There can be a 6dB drop when you go from differential to unbalanced.

Most so called balanced outputs these days are in fact differential.

Cheers

Ian

+1

And on top of that, most off-the-shelf "balanced receivers", iow diffamps for pro-audio boxes, are inherently imbalanced by the nature of their design (impedance/curent mode imbalance by 200%)

I'd write about it myself, but here's a blogpost explaining the situation:
http://www.edn.com/electronics-blogs/analog-ic-startup/4403059/Balanced---audio-line-receivers-and-the-physics-of-fields

My 2c: if you roll your own, you can compensate for said inherent defficiencies, with clever choice of resistors or using some other totp, the most "fool-bulletproof" design being a dual-inverter topology, but not necessarily best from audio POV. Superbal is good but itchy with fast opamps..

Some types of connections while claiming "impendance balance" are in fact, if you look at it this way actually a "remote-sensing" receivers, compensating for dirty connections (gnd currents etc.)

"Balanced receivers" actually don't "sum", but they substract signals, a - -b (twice neg. becomes 2x the value)

The typical way that a 6dB (or 1/2 the signal voltage) loss occurs is when a 2 legged output (signal swinging on both legs) gets connected to an input in such a way that only one leg is received by the following input stage.

There is a common ASSumption that all balanced outputs have active signal on both legs but that is not so for the class of impedance balanced outputs that only have one active signal leg.

There are many different variants on how such interfaces are executed. Some are more tolerant of misconnection than others. In general an interface that loses half the signal may have other subtle issues depending on whether that 6db loss is from an open or short.

JR   

In layman's terms: If the "-" input is left unconnected (floated), your "loss" will be even greater (quieter). So there's even more to this story.

No wonder there  are rumours that some mastering studios convert their gear to unbalanced..

tv said:

No wonder there are rumours that some mastering studios convert their gear to unbalanced..

Click to expand...

Not rumours, actual success stories. Just not related to this particular issue. The two real-world examples I know are both entirely about shortening signal path by removing everything related to balancing stages. But these people know what "ground" means and how to deal with it in relation to a signal path.

Yeah I know ... I was just using a bit "journalistic" lingo.. my point was that there is more than one way you can mess up a "balanced" connection..

Of course, tightening the ground is the first thing to do if you plan any semi-permanent setup..

tv said:

I'd write about it myself, but here's a blogpost explaining the situation:
http://www.edn.com/electronics-blogs/analog-ic-startup/4403059/Balanced---audio-line-receivers-and-the-physics-of-fields

Click to expand...

Have you noticed that this thread is not finished and that Bill Whitlock has shown that the evils of which this simple circuit is accused are due to a too quick and incomplete analysis? Indeed there are better diff input stages, but this one should not be considerd faulty; it is just limited in performance, and quite sufficient in many cases.

I don't think it was considered faulty but - the way it is -usually- made, i.e. with mirrored resistor values makes it less-than perfect, no matter if one-chip receivers have laser trimmed resistors, it's still by any standard NOT doing what it should be - being balanced in most aspects.

With discrete-built simple receivers, one can hack much better performance in said criteria, with simple resistor swap in one receiver leg to balance it in current mode...

fwiw it's LESS silicon in the signal path compared to other recievers and can be hacked/modded to taste easier than others, which may be of importance to those who mod their gear (for whatever end goal in mind, be it simple permanent gain adjustments or "mastering" mods).

However, it is STILL a miniature analog computer substracting environment errors from the sum of said errors and your tone.

Kingston said:

tv said:

No wonder there are rumours that some mastering studios convert their gear to unbalanced..

Click to expand...

Not rumours, actual success stories. Just not related to this particular issue. The two real-world examples I know are both entirely about shortening signal path by removing everything related to balancing stages. But these people know what "ground" means and how to deal with it in relation to a signal path.

Click to expand...

No success story can convince me of rewiring the next Bruce Springsteen tour unbalanced, although (or perhaps because) I know what "ground" means, i.e. a theoretical notion.
In order to achieve a noise rejection of 70 dB, which many diff input stages do, you need to have a "ground" connection totalizing less than 1 milliohm for a nominal 10k connection. I'm not ready to wire my studio with copper bars and power distro terminals.

By all means I think we all agree on that, it's ...selbstverstandlich

However, there is also a "different world" with different approacht towards the underlying problem (the one where RCD ground lifts rule supreme imho)

Same stick - different ends


The smaller the rack, easier it is to "compartmentalize" the audio grounds (for the lack of a better word)

tv said:

I don't think it was considered faulty but - the way it is -usually- made, i.e. with mirrored resistor values makes it less-than perfect, no matter if one-chip receivers have laser trimmed resistors, it's still by any standard NOT doing what it should be - being balanced in most aspects.

With discrete-built simple receivers, one can hack much better performance in said criteria, with simple resistor swap in one receiver leg to balance it in current mode...

Click to expand...

Do you mean the infamous 2:1 ratio that gives equal leg impedance or the (2G+1):1 ratio that some advocate? THese have all been proved wrong 

  However, it is STILL a miniature analog computer substracting environment errors from the sum of said errors and your tone. 

Click to expand...

Would you describe a transformer as a"miniature analog computer"?

abbey road d enfer said:

No success story can convince me of rewiring the next Bruce Springsteen tour unbalanced, although (or perhaps because) I know what "ground" means, i.e. a theoretical notion.
In order to achieve a noise rejection of 70 dB, which many diff input stages do, you need to have a "ground" connection totalizing less than 1 milliohm for a nominal 10k connection. I'm not ready to wire my studio with copper bars and power distro terminals.

Click to expand...

Analog processing for mastering is typically line level, where a noise rejection of less than 70 dB seemingly can be tolerated.  Public address systems are another story.  At least 3 mastering rooms in Los Angeles interconnect their analog processing equipment unbalanced, they have been success stories for decades.