How Balanced Is Balanced?
Sep 16, 2009 at 3:05 PM Thread Starter Post #1 of 16

jazznap

500+ Head-Fier
Joined
Oct 1, 2004
Posts
803
Likes
12
I'm considering entering the balanced amp arena at some point, but I'm wondering how far a balanced system needs to go. I figure to get the max benefit, I would need to upgrade my source and ICs, but does a balanced system really need to be "fully" balanced end to end?

I ask because I love my source (Arcam fmj cd23 -> kimber kcag) and am not really looking forward to trying to find a balanced source that matches this combo. Do most people here with balanced amps balance their system end to end?
 
Sep 16, 2009 at 3:19 PM Post #2 of 16
I can't really answer your question, but the source issue is the same reason I chose not to go balanced. I have several sources that I love and do not want to replace. After reading a lot and listening to a few balanced rigs, I decided that I'd rather stay single-ended.
 
Sep 16, 2009 at 3:42 PM Post #3 of 16
You probably do not need to, but you will most probably benefit from going there. Meaning balanced source, interconnects, amplifiers and headphones.

If most people with a balanced amplifier balance their system end to end I do not know. I used to, but no longer do since I went fully balanced to unbalanced DAC last year.
 
Sep 16, 2009 at 8:26 PM Post #5 of 16
Yes and no. The biggest benefit to going balanced is in the performance of the amplifier itself, as far as headphones go. Balanced transmission was really developed for driving long cables such that any common mode noise picked up in the cable would be cancelled at the destination. Since headphone cables are so short, this isn't really a problem. For headphones, the main advantage is that a balanced headphone amplifier will have twice the slew rate, twice the voltage gain, twice the available voltage swing, half the output impedance, and 4 times the power of an equivalent unbalanced rig at approximately slightly more than twice the parts cost. Pulling that off with a SE amplifier would be much more expensive. However, the amplifier has to be fed with a balanced input for this to matter. If you feed an unbalanced signal to a balanced amp then it behaves simply like an active ground amplifier (see AMB Laboratories DIY Audio Site for more information - look at the technical highlights of the M3 amplifier). In fact, it might behave worse than a simple active ground system, because the ground in an active ground system has unity gain, while the L- and R- of of a balanced amp has some gain (for headphones typically 3-8) which will amplify any noise in the ground signal of your source. So to get the benefit of a balanced amplifier it must either be fed with a balanced source or it must have an internal phase splitter that converts the unbalanced source to balanced before sending the signals to the amplifier chain itself. Most balanced amps don't have this internally, so to use an unbalanced source with a balanced amplifier you need an external phase splitter (probably a transformer) between the source and the amplifier.

If you want a half-and-half system, buy an active ground amplifier. It has some of the benefits of balanced (the headphone transducers "see" the same thing on both of it's terminals, as opposed to an amplifier on one side and usually the capacitor bank of the power supply on the other - again, see amb's website for more info and I credit this explanation to amb - I basically stole this sentence off his website), it costs less than a balanced system, and you don't have to worry about a new source or an external phase-splitter in addition to a new amplifier - and you don't have to reterminate your headphones!
 
Sep 16, 2009 at 9:06 PM Post #7 of 16
If I understand what I have read, the eddie currant BA and the millet 307a have splitters in them. So does that mean they are fully balanced even if they are fed from a unbalanced dac?
 
Sep 16, 2009 at 9:17 PM Post #8 of 16
Quote:

Originally Posted by KingStyles /img/forum/go_quote.gif
If I understand what I have read, the eddie currant BA and the millet 307a have splitters in them. So does that mean they are fully balanced even if they are fed from a unbalanced dac?


Yes your amps totally balanced, even with an unbalanced source.
 
Sep 16, 2009 at 9:22 PM Post #9 of 16
Thanks UglyJoe. That helps immensely. It also sounds like there is no reason to upgrade to a balanced amp early and run it SE with the intent to upgrade the source later, since performance may suffer in the interim.

Are there some high-end SS amps that are not balanced (since most seem balanced) or do unbalanced SS amps sort of cap out at some point in terms of performance (say M3, GS-1, etc; in general w/o discussing the merits of each).
 
Sep 17, 2009 at 3:09 AM Post #10 of 16
Quote:

Originally Posted by jazznap /img/forum/go_quote.gif
Thanks UglyJoe. That helps immensely. It also sounds like there is no reason to upgrade to a balanced amp early and run it SE with the intent to upgrade the source later, since performance may suffer in the interim.

Are there some high-end SS amps that are not balanced (since most seem balanced) or do unbalanced SS amps sort of cap out at some point in terms of performance (say M3, GS-1, etc; in general w/o discussing the merits of each).



Beta22, active ground build. Amb's active ground beta22 had Ray Samuels saying, "It doesn't sound solid-state, it sounds like music... takes you so close... doesn't get any better than this." As far as performance suffering in the interim running a balanced amp with an unbalanced source... don't read too much into that from what I said. Yes, the L-/R- amps of a balanced amp would amplify ground noise... but on any decent source, this aught to be minimal anyway. I think the better reason to avoid balanced is cost, especially if you don't see yourself changing sources or getting an external phase splitter at some point. As far as capping out, yeah, high-end SS amps really are about as good as you are going to get without going to extremes, at least as long as you remain single-ended. The beta22 is about as extreme as you can get. Designing something single ended that would have the test-specs of a beta22 balanced would be.... well, really difficult and expensive. There is no reason - from a cost perspective - to attempt this when you can simply balance the system and have such an increase in measurables at basically twice the cost, as opposed to 4 or 5 times the cost or more trying to replicate it with a SE amp. If you intend to upgrade to a balanced source later, then by all means go with a balanced amp now. Your performance won't be noticeably hurt, and since most professionally built balanced amps are based off of extremely nice SE amps, then running a balanced amp as though it were a SE amp won't hurt anything, and will cost less in the long run if you go fully balanced.

Quote:

Originally Posted by KingStyles
If I understand what I have read, the eddie currant BA and the millet 307a have splitters in them. So does that mean they are fully balanced even if they are fed from a unbalanced dac?


A lot of tube amps (any that are push-pull) have a phase splitter in them. In most cases this is not "balanced" in the way a balanced SS amp is. Very often the signals are summed before being sent to the headphone. The headphone still "sees" an amp on one side and power supply ground on the other. I've even seem some designs that are called "balanced" that accept a balanced source but basically bypass their own phase splitter and then do the same thing as above, with the ground of the transducer being directly tied to power supply ground. This IS NOT balanced, at least not in the way that we think of when we say a "fully balanced system." I'm 99% sure that the Millet 307A at TTVJ falls into a category like this. After all - it's a single ended triode amplifier; the "balanced" outputs are almost assuredly for show - you could tie the grounds of the headphone transducers together and the amp would behave exactly the same. The amp is transformer coupled... depending on how you look at it the transformers can be seen as "balancing" a single-ended source or as "unbalancing" a balanced source. It's really a different monster all-together, and nothing like a balanced SS amplifier, which won't be transformer coupled.

I looked at it more closely just now. The 307A has transformer coupled outputs and inputs. Remember, transformers block DC, so ground is whatever you want it to be after going from the primary to the secondary of the transformer. If you don't force a point of the circuit to ground intentionally, then the circuit after the transformer is floating and you can call any point you want "ground". The transformer doesn't care. So if you call one side of a transformer coil "ground" then the signal is all on the other side of the coil - i.e.- single ended. If you had a tapping point at the center of the transformers secondary and measured the ends of the coil you would see equal but opposite signals on either coil - i.e.- balanced.

This is how a transformer could be used as a "phase splitter" (terrible word, because that's not what it does... the old nomenclature has stuck around for years, though). A transformer would have have a secondary that was tapped at the middle. This tap would be tied to the ground of the input of the amplifier. Either side of the secondary would then feed the inverted signals to the +/- inputs of the amplifier, and away we go. Each amplifier would then amplify it's signal, and the signals would be sent to either end of the headphone itself.... the headphone never see the "ground" of the amplifiers or the source... it just sees the differential signal and plays the music.
 
Sep 17, 2009 at 4:19 AM Post #11 of 16
Bill Whitlock of Jenson Transformers has been instrumental in audio standards organization’s definition of "Balanced" interfaces in audio

the distinction is not the commonly understood symmetrical +/- drive Voltages but rather the equal impedance in the +/- signal lines - the drive and the receiver can have different impedances but the source +/- series impedances must be equal as well as the receiver +/- impedances

only with this condition does induced noise common to both signal lines exactly cancel in the differential signal measurement

this definition/usage of balanced operation is most valuable in transmitting low level signals in long cable that may encounter substantial interfering magnetic feilds or differences in transmitter/receiver ground references - often due to ground loops associated with high current equipment like stage lighting dimmers in professional applications

for headphone drive balanced impedance noise rejection concept does not appear to offer large improvements since the source is at a relatively high level and the cable length is short

transformer drive of the headphones can meet the balanced impedance criteria despite SE amplifier driving the xfmr primary - and regardless of where the "gnd" is connected

Bridged/Complementary Symmetry amplification is also called "balanced" operation by most but it can be useful to draw a distinction – a balanced signal does not have to have symmetrical +/- V swings about any particular “gnd” while Bridged/Complementary Symmetry amplifiers are defined by the symmetrical drive V on the +/- outputs

Bridged/Complementary Symmetry amplifiers can have balanced or SE signal inputs, usually they accept both
 
Sep 17, 2009 at 4:57 AM Post #12 of 16
Thanks for the followup Uglyjoe. A lot of good info there.
 
Sep 17, 2009 at 11:04 AM Post #13 of 16
Quote:

Originally Posted by KingStyles /img/forum/go_quote.gif
If I understand what I have read, the eddie currant BA and the millet 307a have splitters in them. So does that mean they are fully balanced even if they are fed from a unbalanced dac?


It depends on what you mean by "fully balanced". They do have balanced inputs and outputs, however the BA and Millett 307A are single ended triode amplifiers. The amplification is performed single ended. They use input transformers to accept balanced input sources. They are not differential input-through-output amplifiers, by which some mean when they say "fully balanced".
 
Sep 17, 2009 at 9:50 PM Post #14 of 16
Quote:

Originally Posted by Uncle Erik /img/forum/go_quote.gif
I can't really answer your question, but the source issue is the same reason I chose not to go balanced. I have several sources that I love and do not want to replace. After reading a lot and listening to a few balanced rigs, I decided that I'd rather stay single-ended.


By that do you mean that you found the performance gains to be minimal when comparing balanced to SE?
 
Sep 18, 2009 at 6:33 PM Post #15 of 16
Quote:

Originally Posted by jcx /img/forum/go_quote.gif
Bill Whitlock of Jenson Transformers has been instrumental in audio standards organization’s definition of "Balanced" interfaces in audio

the distinction is not the commonly understood symmetrical +/- drive Voltages but rather the equal impedance in the +/- signal lines - the drive and the receiver can have different impedances but the source +/- series impedances must be equal as well as the receiver +/- impedances

only with this condition does induced noise common to both signal lines exactly cancel in the differential signal measurement

this definition/usage of balanced operation is most valuable in transmitting low level signals in long cable that may encounter substantial interfering magnetic feilds or differences in transmitter/receiver ground references - often due to ground loops associated with high current equipment like stage lighting dimmers in professional applications

for headphone drive balanced impedance noise rejection concept does not appear to offer large improvements since the source is at a relatively high level and the cable length is short

transformer drive of the headphones can meet the balanced impedance criteria despite SE amplifier driving the xfmr primary - and regardless of where the "gnd" is connected

Bridged/Complementary Symmetry amplification is also called "balanced" operation by most but it can be useful to draw a distinction – a balanced signal does not have to have symmetrical +/- V swings about any particular “gnd” while Bridged/Complementary Symmetry amplifiers are defined by the symmetrical drive V on the +/- outputs

Bridged/Complementary Symmetry amplifiers can have balanced or SE signal inputs, usually they accept both




I agree with everything you say here. The problem is that whole nomenclature thing. For the average head-fier, balanced means differential drive from a symmetrical +/- signal fed by two identical amps - i.e.- bridged-mono. Unfortunately it's easier to talk people through issues using the common nomenclature instead of trying to explain exactly what balanced vs. bridged mono vs. single ended vs. push pull is and how they are related and how they are unrelated. Most people think that SE and balanced are opposite to each other... in fact, they are really completely unrelated!
 

Users who are viewing this thread

Back
Top