[Steve Eddy]

The Balancing Act is single ended. It has input and output transformer and XLR connectors, but you'd need twice as many tubes to make it balanced.

No, you don't. Balanced is all about the interface. Inputs and outputs. It's all about common mode noise rejection which takes place at the input. Even in the pro world, virtually all of the circuitry between the inputs and outputs is single ended. Craig's using CineMag input transformers. That gives the Balancing Act proper balanced (and differential) inputs.

se

 

[paradoxper]

Seeing as my understanding of balanced is CMR and differential signaling, I am completely confused by the notion
that the EC BA would need twice as many tubes to make it balanced. Even though Steve disagrees, I'd like to try
to understand the reasoning behind it.

 

[Steve Eddy]

No to be too harsh, but really, the entire headphone industry should just stop using the word "balanced." Next to no one has any real understanding of it. It just makes the industry look like some sort of amateur technological throwback. Seriously, it's embarrassing. It really is.

*sigh*

se

 

[Steve Eddy]

Seeing as my understanding of balanced is CMR and differential signaling, I am completely confused by the notion
that the EC BA would need twice as many tubes to make it balanced. Even though Steve disagrees, I'd like to try
to understand the reasoning behind it.

Balanced and common mode rejection has absolutely nothing to do with the signal. It doesn't matter if the signal is symmetrical on both lines or asymmetrical on just one line. Or if there's any signal at all. Yes, common mode rejection works even in the absence of a signal.

se

 

[money4me247]

No to be too harsh, but really, the entire headphone industry should just stop using the word "balanced." Next to no one has any real understanding of it. It just makes the industry look like some sort of amateur technological throwback. Seriously, it's embarrassing. It really is.

@Steve Eddy. honestly, I don't think I understand everything that is being said in this discussion as my knowledge of the technical terms and their meanings is quite limited.
 
So basically bridging = having one or more amplifier per channel that are working in sync. bridging =/= 'balanced" as the term that professional non-headphone uses. so what are the technical benefits of bridging if it is not offering the common mode noise rejection benefit?
 
'balanced' via the professional definition = a balanced interface at the input. what is a differential input mean exactly? I was under the impression that RCA input connectors are single-ended, so does that mean that XLR inputs would be differential? Or is it relating to something else completely?
 
Also, if being "balanced for common mode noise rejection" only relies on the input, is it possible to have a single-ended output with a differential input that is 'balanced'?
 
Sorry if these questions are silly or obvious or just incorrect. I don't think I understand the topic as well as I would like, and I would appreciate some clarification. Or at least perhaps a link where I can do some personal reading & learning. thanks!

 

[paradoxper]

Balanced and common mode rejection has absolutely nothing to do with the signal. It doesn't matter if the signal is symmetrical on both lines or asymmetrical on just one line. Or if there's any signal at all. Yes, common mode rejection works even in the absence of a signal.

se

Well, that clears it up for me, further. There is just a lot of convolution when it comes to "balanced". Still don't understand the bit Doug said about needing more tubes.
Care to expand if you have an idea where his thinking was?

 

[Steve Eddy]

So basically bridging = having one or more amplifier per channel that are working in sync. bridging =/= 'balanced" as the term that professional non-headphone uses. so what are the technical benefits of bridging if it is not offering the common mode noise rejection benefit?

Ok, with a typical single ended amp, you have just a single amplifier for each left and right channel. At the input, one terminal is the input and the other terminal is ground. At the output, one terminal is the output and the other terminal is ground.

With bridging, you take two identical single amplifiers and bridge them together. At the input, one terminal is the input to one of the single ended amplifiers and the other terminal is the input to the other. At the output, one terminal is the output of of one of the single ended amplifiers and the other terminal is the output of the other.

The load is tied across these two output terminals and effectively "bridges" the two amplifiers together, which is where the term "bridging" comes from.

Typically, the input to one of the two single ended amps is fed a normal signal. The input of the other amplifier is fed an identical signal, but of the opposite polarity. In other words, if the normal signal is at say +2 volts, then the other signal will be at -2 volts. So now the differential voltage across the two output terminals will be (+2) - (-2), or +4 volts. Twice what it would be if just using one single ended amplifier.

The "benefit" is that if your power supply voltage is such that the single ended amp can only swing x volts, with bridging, you can use the same power supply voltages but achieve 2x volts.

Bridging was commonly used for car audio back in the day where you only have 12 volts to work with. So bridging, combined with using lower impedance 4 ohm loudspeakers let you wring as much power as possible from the limited power supply voltage available.

Of course now car audio uses switching power supplies (DC to DC converters) to achieve whatever power supply voltages they want.

But for AC powered home amplifiers, you can choose whatever power supply voltage you want, so bridging is of dubious value. And even battery powered devices can take advantage of DC to DC converters to achieve higher voltages as well.

'balanced' via the professional definition = a balanced interface at the input. what is a differential input mean exactly? I was under the impression that RCA input connectors are single-ended, so does that mean that XLR inputs would be differential? Or is it relating to something else completely?

Ok, first of all, "balanced" refers to the impedance of each line with respect to ground. In a balanced interface, those impedances are the same. With a single ended interface, one of those lines is ground, so by definition, is inherently unbalanced.

A differential amplifier only amplifies the difference it sees at its two terminals. Even single ended amplifiers typically have differential inputs, but for common mode rejection you need balanced impedances between each line and ground. This is because interference noise typically impinges equally on both lines, but without impedance balance, that noise won't appear equal on both lines. In other words, the voltage produced by the interference won't be common mode. No common mode voltage, no common mode noise rejection.

Also, if being "balanced for common mode noise rejection" only relies on the input, is it possible to have a single-ended output with a differential input that is 'balanced'?

I didn't say common mode rejection only relies on the input. I said where common mode rejection actually occurs is at the input. The outputs need to be balanced as well in order to keep the noise common mode.

Sorry if these questions are silly or obvious or just incorrect. I don't think I understand the topic as well as I would like, and I would appreciate some clarification. Or at least perhaps a link where I can do some personal reading & learning. thanks!

Wikipedia has a pretty good treatment of the subject.

http://en.m.wikipedia.org/wiki/Balanced_line

se

 

[Steve Eddy]

Well, that clears it up for me, further. There is just a lot of convolution when it comes to "balanced". Still don't understand the bit Doug said about needing more tubes.
Care to expand if you have an idea where his thinking was?

I think his thinking was based on the same erroneous thinking that got bridging to be called "balanced" instead of bridging. Either that or he thinks that everything from input to output needs to be differential in order to be considered "balanced." Sure, you can run balanced differential from input to output if you'd like, but it's not really necessary. As I said, not even pro audio does that to speak of. The biggest problem is noise picked up on the typically long runs of cables, and that is taken care of at the inputs.

se

 

[dsavitsk]

The Balancing Act is single ended. It has input and output transformer and XLR connectors, but you'd need twice as many tubes to make it balanced.

No, you don't. Balanced is all about the interface. Inputs and outputs. It's all about common mode noise rejection which takes place at the input. Even in the pro world, virtually all of the circuitry between the inputs and outputs is single ended. Craig's using CineMag input transformers. That gives the Balancing Act proper balanced (and differential) inputs.

That's a very narrow definition of what balanced means. Many of us are also concerned with balanced circuits The BA circuit is single ended, and this matters because the circuit could be balanced, and if it was, it would reject common mode noise and cancel distortion.

To answer the other question, rejection of common mode noise largely means rejection of power supply noise and isolation from the power supply. Call it differential instead of balanced if you must, but it is decidedly a different thing from bridged.

 

[Steve Eddy]

That's a very narrow definition of what balanced means.

That's absolute nonsense. It is the broadest, most common definition of balanced in the audio world.

Many of us are also concerned with balanced circuits The BA circuit is single ended, and this matters because the circuit could be balanced, and if it was, it would reject common mode noise and cancel distortion.

Well let's just take Craig out behind the woodshed and beat the crap out of him for being such a sloppy and incompetent designer, shall we?

The Balancing Act does reject common mode noise. It does this as a consequence of the balanced input transformers that it uses. In fact, those transformers will offer excellent common mode rejection even if they are fed from an unbalanced source. So I don't get where on earth your notion that the Balancing Act doesn't reject common mode noise is coming from.

To answer the other question, rejection of common mode noise largely means rejection of power supply noise and isolation from the power supply. Call it differential instead of balanced if you must.

The rejection of common mode noise largely means rejection of noise picked up by cables. If by "largely" we're talking about the audio world at large.

Name any competently designed single ended amp that has any audible power supply noise. I've not encountered one yet.

se

 

[dsavitsk]

What's going on, Steve? Maybe I can't read your tone, but you seem very upset about this. Give me a call and we'll agree to disagree, or I'll agree to use the term differential instead

The BA circuit does not reject PS noise, which means that the PS has to be designed to be very quiet. If the circuit were differential, it would have a high PSRR and the PS requirements would be relaxed. That it doesn't doesn't make it a bad circuit, it makes it a design choice. I have made and sold amps with a similar design choice and for the most part, people have liked them. However, I generally prefer differential circuits -- to each his own.

 

[Steve Eddy]

What's going on, Steve? Maybe I can't read your tone, but you seem very upset about this. Give me a call and we'll agree to disagree, or I'll agree to use the term differential instead

Because the subject is balanced and you're going on about differential and they're not one and the same. Differential may be single ended or balanced. And the differential you seem to be talking about is single ended.

And I resent being told that my definition of balanced is a "very narrow definition" when it is the quintessential definition across the audio spectrum, save for the headphone industry.

The BA circuit does not reject PS noise...

Of course it does! It doesn't offer as much rejection as a differential circuit would, but it's not completely lacking power supply rejection.

...which means that the PS has to be designed to be very quiet. If the circuit were differential, it would have a high PSRR and the PS requirements would be relaxed. That it doesn't doesn't make it a bad circuit, it makes it a design choice.

It's rather six of one, half a dozen of the other, isn't it? You can have a more complicated and quieter supply to accommodate a simpler amplifier topology or you can have a more complicated amplifier topology to accommodate a simpler and less quiet power supply topology.

I don't know that the Balancing Act has a terribly complicated power supply topology. Does it have audible power supply noise? If not, who cares?

I have made and sold amps with a similar design choice and for the most part, people have liked them. However, I generally prefer differential circuits -- to each his own.

Personal preference is fine. But you're trying to make technical arguments as to why the Balancing Act is somehow an inherently inferior design.

Bottom line, it is a proper balanced amplifier in the broadest sense of the word.

se

 

[tomb]

 And I resent being told that my definition of balanced is a "very narrow definition" when it is the quintessential definition across the audio spectrum, save for the headphone industry.

 
If that were true, this wouldn't be such a discussion.  It's all about semantics and who can use which term and who can't.  Those are things usually decided by a standards committee and even then, it doesn't mean the industry will follow. 
 

But you're trying to make technical arguments as to why the Balancing Act is somehow an inherently inferior design. 

 
You just called it inferior.  He didn't.  He called it a design choice and implied nothing of the sort about one being inherently inferior vs. another.
Quote from Dsavitsk:

That doesn't make it a bad circuit, it makes it a design choice. 

 

[paradoxper]

Still seems like a massive confusing mess. Doug, what are your thoughts per the "balanced" nature of the DNA Stratus? Similar to the BA it also uses
output transformers providing a differential output. 
 
FWIW here's what Donald has to say regarding "balanced"
 
There is a lot of confusion and misuse of "balanced". Balanced means equal impedance to ground for the + and - signal. Balanced interfaces provide a high impedance to common mode noise. Balanced interfaces are beneficial for long cable runs, particularly useful in recording environments like microphone and mic preamp feeds. Balanced can also reduce the potential for ground loops between components. There can be and exist single-ended, differential, and push-push headphone and speaker amplifiers (tube and transistor) with balanced inputs. Likewise, all three topologies when used as a preamp can have balanced inputs and outputs. A single ended amplifier (be it solid state or tube) runs in class A and the active devices amplify the entire music waveform, both the + and - portions. Push pull amplifiers divide the task and dedicate one set for the + waveform and another for the - and combine at the output. Differential amplifiers have two sets of active devices, each handling the entire waveform, but running opposite phase to one another. A differential amplifier is balanced the entire time, input through output. For a single ended preamp or amplifier to accept a balanced input, it either needs to have dedicated electronic circuitry to accept the balanced, differential input signal and convert it to unbalanced or use an input transformer for the same function of converting the balanced signal to unbalanced. This is how some pro audio tube components work: They're single ended electronically but have input and output transformers to accommodate balanced cables. An output transformer-loaded single-ended amplifier can produce a balanced, differential output by center tapping the output secondary winding. Headphones have 4 wires: + and - for each channel. With traditional 1/4" TRS plugs, the - wires are tied together for a common ground. Some listeners feel there is a benefit by separating the - minus leads and driving them separately to eliminate potential back EMF interaction between the headphone transducers & amplifier. This is similar to the rationale for bi-wiring loudspeakers, separating the feeds for the tweeter and woofer all the way back to the amplifier. The 4 wire, balanced headphones need to be driven by an amplifier which provides a differential, balanced output. Again, this can be from a single ended, push-pull, or differential amplifier.
 

 

 

[tomb]

That seems a lucid explanation.
 
This thread seems to be arguing that the definition of balanced is only applied to cables.  That's completely failing to recognize the possibility of a new application to that definition.  It's like creating a rule that no one can discover an expanded application principle, because the original definition doesn't allow it.
 
The analogy is understood only if you consider than an amplifier has entered the arena of the cable.  To keep the signal balanced and to use common mode rejection within the amplifier itself, then the amplifier circuit must be duplicated just as the cables are.  Once the cables are separate, the balancing holds true.  However, introduce an amplifier into that signal stream where the signal is no longer separated, then it's no longer balanced.  If the amplifier also imparts some noise you're trying to remove, then the advantage is lost ... unless you have two circuits that continue to keep the signal separated ... as in two sets of tubes, for instance.
 
Doug will probably respond eventually, but it'll have to be late at night.