What is balanced? What are balanced amps? Do they sound better? Discussion for noobs and boobs.

Discussion in 'Headphone Amplifiers and Combo (DAC/Amp) Units' started by purr1n, Dec 5, 2015.

  1. OJneg

    OJneg The Most Insufferable

    Pyrate BWC
    Joined:
    Sep 25, 2015
    Likes Received:
    3,923
    Trophy Points:
    113
    Location:
    Grand Rapids, MI
    My take on the balanced question as a relatively inexperienced engineer:

    • Balanced transmission: ALWAYS GOOD. Rejects common mode noise/interference. Haven't heard any good arguments against it
    • Balanced input: Also GOOD, but with tradeoffs. At the end of the day, you need a balanced input and output to take advantage of above.
      • Active differential, "balanced" input: GOOD, but might not be the best choice for all applications. All things being equal, a classic LTP will double the input referred noise of the input stage compared to an equivalent SE input. Other than that, an active balanced (LTP, instrumentation amplifier, etc) is great because it gives you an easy way to setup an accurate balanced high impedance input, with the option of getting both differential or SE outputs for the next stage
      • Passive, transformer balanced input: MAYBE. A passive device like a transformer will have lower inherent noise (limited by winding resistance) than a balanced input, but will simply not be as linear across the whole range. Also care needs to be taken with following stage of course, as an incorrect termination could lead to loading the transformer down, killing bandwidth, or causing oscillation. This option is best implemented in mic, phono, or other low signal level designs when you're going to be getting some gain out of the device at the same time, and the levels will also be low enough that you won't be swamped by LF distortion. Using a line level transformer to go balanced-to-SE goes against most of my sensibilities.
    • Balanced output: DEPENDS
      • Balanced bridge: MEH. I see this more as a convenient way to get more power out of the same amps (or amp channels) you have lying around. I relegate this option to subwoofer duties, as headphone and most speakers don't need the jiggawatts and are more sensitive to other circuit choices
      • "True" balanced output (Circlotron and other permutations): GOOD. Awesome way to take advantage of fully balanced, differential signal path. Design considerations can be PITA, ask Jason.
      • Push-pull output stage (not necessarily Class AB): GOOD. Assuming good circuit symmetry, an easy way to get a lot of swing out of a pair of devices. Blameless in the scope of things. Usually requires transformer though
      • Single-ended output stage with "balanced" lines: OK. Same issues as any other SE amplifier, but a convenient way to keep the load off ground.
    • Balanced (or rather, differential) wiring: GOOD
      • Mic/Instrument/Phono: AWESOME. Refer to bullet #1. Of course most audiophiles are dummies who will use the SE inputs from their turntables because that's what it came with.
      • Line-Level: GOOD.
      • Digital: AWESOME.
      • Headphone/Loudspeaker: QUESTIONABLE. Good obviously if you want to take advantage of a good balanced output. Will also reduce crosstalk significantly (and audibly). 4-Pin XLRs are more robust connectors all around
     
  2. peef

    peef Friend

    Pyrate
    Joined:
    Oct 5, 2015
    Likes Received:
    311
    Trophy Points:
    63
    Location:
    Montreal
    Here's a recent (?) paper by Bruno Putzeys on ground and differential signaling that I think someone else linked in another thread. It's a gem. The gist of it is that in a single ended design "ground" functions as a voltage reference (0V), a signal reference, a signal return path, AND a shield. You can use differential signaling techniques in a single ended circuit to get around that and retain the CMRR of a balanced line while only driving one of the two phases.

    Hope purrin doesn't mind if I play with his drawing a bit. :)

    [​IMG]

    The floating source can be the secondary of a transformerm or anything with a single current path exists between its two terminals. RCM can go to the amp chassis, or to +10000V, or to a unicorn. It doesn't care, because the circuit is floating and RCM will draw no current provided that R+ and R- are accurately matched. If both sides are driven, you'll have a balanced output with a CMRR dependant on R+, R-, and RCM.

    Redrawing again for a single ended source, we have this.

    [​IMG]

    If RCM is large, you have a phase splitter with an insertion loss of 6dB because of the current loop: an equal voltage drop is developed across R+ and R-, and they're out of phase relative to the signal reference. If it's a short, nothing, because RS2 and R- are effectively bypassed and the voltage drop is entirely across R+.

    Tyll, to get back to your point, I think that balanced impedances to ground and CMRR are one and the same. On the flip side, I wouldn't expect a pair of SE amps running out of phase to show the same benefits unless they're coupled at some point.
     
  3. purr1n

    purr1n Desire for betterer is endless.

    Staff Member Pyrate BWC
    Joined:
    Sep 24, 2015
    Likes Received:
    89,777
    Trophy Points:
    113
    Location:
    Padre Island CC TX
    Moral of the story: unless you are in a studio environment, don't worry about it if the kick-ass amp you like doesn't do balanced.

    Another story: EC thought about making BW or CH balanced (bridged) purely for sales reasons because people are idiots and we could take more money from people that way. In the end, we decided we didn't want to do this. There were other reasons not to do so:
    1. Double parts count and cost
    2. Increased heat
    3. BW / CH already had very high gain, balancing would have doubled voltage, and we didn't want to increase feedback one bit to dial the gain back, because that made the amps sound not as good. Sort of a one step forward, one step back thing. Besides some people have single-ended sources.
    Now we might be more than happy to take $2500 or more from people who insist they want balanced versions of the amps. We'll throw in a three position gain switch too since most people probably wouldn't notice how the amps start to sound more closed in and deader and flatter at the lower more usable gains. But last time I spoke to Craig, he said something of the sort that he would rather kill himself than do this.

    Balanced for the most part is stupid in the home environment. And let's face it, most idiots want balanced because it gives them a lame excuse to buy fancy aftermarket cables.

    My headphone setup is fully singled-ended (HD650+ZDS+TT+phono-pre). My speaker setup (Homemade OB+Yggdrasil+Rag) is balanced. Absolutely zero consideration was given to whether gear was to be balanced or not when putting the setups together. The only consideration was performance for cost.
     
    Last edited: Dec 5, 2015
  4. johnjen

    johnjen Doesn’t want to be here but keeps posting anyways

    Pyrate
    Joined:
    Sep 29, 2015
    Likes Received:
    505
    Trophy Points:
    93
    Location:
    Well Grounded
    My 2¢…
    The one aspect that I rarely if ever see mentioned in any SE vs differential/balanced discussion is the influence that ground has upon the active circuit itself.

    In a SE circuit if you modulate the ground reference, you also modulate the output, because that circuit ‘stands’ upon that ground as it’s reference.

    In a differential circuit this ground modulation tends to be self canceling, if it even gets introduced into the active circuit in the first place (think ‘floating’ circuit topology aka The Rok amp).

    So how would the ground get modulated in the first place?
    There are 2 sources that readily come to mind, and there are others but these 2 are sufficient for this discussion.

    Back at the panel where ground and neutral are tied together means that any ‘noise’ that any device on any branch circuit that does manage to modulate the neutral circuit will also ‘influence’ all devices that also ‘stand’ on that very same neutral/ground ‘potential’ (ie our audio systems).
    Think motors, transformers, SCR’s, dirty switched circuits etc. as they are turned on/off, or even while running.
    They can all ‘dump’ noise back into the neutral circuit and thus modulate the ground, sometimes with very audible consequences.

    The other source is the earth itself where our ground rods are located.
    The thing is, without a differential circuit to be able to measure our single point ground, it is nearly impossible to detect these modulations.
    So 2 ground rods are, at a minimum, needed to detect/measure such potentials.
    If you have never researched the effects that ground potentials have had, and do have on active circuits, this may seem to be all to easily dismissed.

    For example, back during the late 1800’s and early 1900’s when the telegraph and early telephone systems were first being deployed, the whole topic of ground was a major factor and in some cases completely prevented some telegraph systems from operating, at all.
    In other situations the power supplied from the ground connection was used to power the telegraph function.
    IOW they didn’t need batteries, at all.

    The reason I bring this up is what we think of as ground being stable and uniform and consistent simply isn’t the case, at least not everywhere.
    In fact there were ground based radios that worked quite well and had no antenna, as it wasn’t needed.

    IOW ground is, or can be quite a source of modulation all by itself.

    Now granted much of this modulation is probably of a low order of magnitude, but these days as we strive to achieve 120dB dynamic range, where this degree of resolution reaches down into these very same orders of magnitude, SE designs are more susceptible to these forms of modulation than differential designs are.
    Because standing on a modulated ground means the entire circuit is now ‘locked’ onto that modulation for both SE and differential circuits.
    But the differential circuits ’compensate’ for this as an inherent function of their design.

    And, digital circuits are not immune from this ground modulation either.
    With but very few exceptions, every digital chip has a ground connection along with power and signal I/O.
    And most digital circuits rely upon ground as being stable and ‘relatively’ unchanging.
    And these days with digital circuits operating in the 1 volt range of logic on/off, even small amounts of modulation can ‘trigger’ a change of state either before or after its ’supposed’ to.

    Not to mention that the entire power supply uses ground to ‘bleed’ off any non DC voltage ’noise’, which is a ‘local’ modulation, with the intent that the ground will remain ‘stiff’ and will dissipate the noise energy without affecting any of the rest of the circuit.

    I could go on for a bit more on this but my point is SE and Differential designs both rely upon ground as THE Reference from which to then provide ALL signal modulation.
    But a ‘true’ differential design inherently deals with these sources of noise much better than a SE design can.

    And then there is the whole concept of a ‘transmission’ line which can take this to a whole nuther level of mind bend’n complexity. :confused:

    JJ
     
    • Like Like x 1
    • Epic Epic x 1
    • List
    Last edited: Dec 5, 2015
  5. feilb

    feilb Coco the monkey - Friend

    Pyrate
    Joined:
    Oct 5, 2015
    Likes Received:
    226
    Trophy Points:
    33
    What @johnjen says is insightful, but I want to point out and clarify something he has alluded to. Voltage by definition exists between two points. There is no such thing as absolute "voltage" only the relative electrical potential difference between two points.

    From the perspective of the driver in your headphone, the only voltage that matters is the one between one end of its voice coil and the other. You could attach both ends of your voice coil to the "hot" line in your house (110V AC wrt neutral lets say) and absolutely nothing would happen because there is no potential difference between the two ends. There is no voltage across the coil.

    If "ground" in all of your equipment is always at the same potential (even if that potential is 120VAC relative to some other point outside your circuit) it wont matter. If it all bounces together, it wont matter.

    None of this disputes anything mentioned above, but it is important to clearly understand where issues can and cannot crop up.

    Edit: I'll also add one point about ground modulation that is theoretically applicable to SE outputs: common impedance coupling.

    Common impedance coupling is coupling between two different signals that share a common impedance. In this case, the impedance of the current return path (sometimes called ground, but I prefer to call it "return" for clarity sake). Assuming that the return path has some non-zero impedance, current flowing across it will develop a voltage across it. The larger the current, the bigger the voltage developed.

    This is significant, because an increase in current on one channel raises the ground potential wrt the transmitter on both channels, thus reducing the voltage seen at the receiving end.

    If your headphones have only 1 common wire (until the split to each cup) some voltage develops across the non-zero resistance and causes coupling between the L and R signals. The amount of coupling is dependent on the impedance of the shared return path.
     
    Last edited: Dec 5, 2015
  6. rott

    rott Secretly hates other millenials - Friend

    Pyrate
    Joined:
    Oct 23, 2015
    Likes Received:
    1,176
    Trophy Points:
    93
    Location:
    Nation's Capital of failure
    Not sure if this is the right thread for this, but what audible effect, if any, is there in using balanced interconnects between source and amp that are either comparatively higher (Canare/Mogami) or lower (Belden/Gotham GAC) in capacitance? Does it make a difference for very short ( <= 3ft.) runs?
     
    Last edited: Dec 7, 2015
  7. zerodeefex

    zerodeefex SBAF's Imelda Marcos

    Staff Member Pyrate BWC
    Joined:
    Sep 25, 2015
    Likes Received:
    14,051
    Trophy Points:
    113
    Ring a ding ding! We have a winner.

    I never understood why people chased having a fully balanced system over maximizing performance given their spend.
     
  8. zaisan

    zaisan New

    Joined:
    Dec 9, 2015
    Likes Received:
    0
    Trophy Points:
    3
    For amps that have similar power outputs between its balanced and single-ended outputs (CL Theorem, CL Duet), is there any merit in going balanced besides the fancier cables?
     
  9. rott

    rott Secretly hates other millenials - Friend

    Pyrate
    Joined:
    Oct 23, 2015
    Likes Received:
    1,176
    Trophy Points:
    93
    Location:
    Nation's Capital of failure
    Straight from the Theorem's designer (email): "The balanced output is supposed to be just a little bit cleaner, more open space. It is not a giant difference, but we are just trying to bring out every detail." I haven't tried it, but my guess is very negligible gains.
     
  10. Armaegis

    Armaegis Friend

    Pyrate BWC
    Joined:
    Sep 27, 2015
    Likes Received:
    7,464
    Trophy Points:
    113
    Location:
    Winnipeg
    Just mentioning this for the sake of completeness: in some pro audio gear, the negative half of the balanced line is actually driven at zero volts (but it is not "ground") rather than inverted from the positive.
     
  11. johnjen

    johnjen Doesn’t want to be here but keeps posting anyways

    Pyrate
    Joined:
    Sep 29, 2015
    Likes Received:
    505
    Trophy Points:
    93
    Location:
    Well Grounded
    The only thing I can offer as a thought to ponder is, with our ability to continue to push the envelope ever closer to systems with fully capable 120dB dynamic range, even the little things become more important.

    Some aren't pursuing this as a goal, but some are.

    JJ
     
  12. Thenewerguy009

    Thenewerguy009 Friend

    Pyrate
    Joined:
    Oct 11, 2015
    Likes Received:
    385
    Trophy Points:
    63
    Does anyone know if the connector tips only determine if a headphone will be balanced or single ended?

    Or does the connection physically have to be to the cups/drivers?
    I ask because everyone talks about re-terminating their headphones if they want to go balanced or vice versa, but there are adapters out there being sold & I never see people talk about them

    Like can a cheap adapter like this turn a 6.35 single ended cable into a fully balanced cable & vice versa?
    [​IMG]

    I am having a hard time finding one with 4 pin configuration outside of norneaudio.com

    Or is there more than just the connector part & headphones have to be fully re-terminated to take advantage of being fully balanced?
     
  13. Merrick

    Merrick A lidless ear

    Pyrate
    Joined:
    Jan 6, 2016
    Likes Received:
    12,569
    Trophy Points:
    113
    Location:
    Portland, OR
    The headphones have to be wired as balanced to the drivers in order to get balanced playback. Some headphones are wired this way out of the box, especially the higher end ones. All the Sennheiser headphones from the HD600 and up are balanced out of the box, although you'll need balanced cables for them (the HD800S includes a balanced cable in the box). I think all the Hifiman cans are balanced capable, and the Oppo PM-1/2/3. I assume virtually all the high end Audezes can.

    If a headphone has a hardwired cable and terminates in a 1/4" or 1/8" TRS connector, like Grados or the TH-X00, it wasn't designed for balanced playback and will need to be rewired.

    As far as adapters go, you can have a cable that adapts balanced cables to single-ended playback (note that this does not mean you're getting a balanced signal, you're just using balanced cables to play back a single-ended signal). I have one of these for all of my balanced headphones so I can use them with single-ended sources without swapping cables. Going the other way doesn't really work though, you'd need a piece of hardware to convert the signal from balanced to single-ended.

    I'm just going off memory and what I've learned, others may be able to answer you more authoritatively, but I can definitely say that no, that cable you've linked to will not turn a single-ended headphone into a balanced headphone.
     
  14. Colgin

    Colgin Friend

    Pyrate
    Joined:
    Sep 27, 2015
    Likes Received:
    634
    Trophy Points:
    93
    Yes, but assuming you have a headphone that is balanced -- e.g., Audeze, Oppo -- rather than a balanced cable for each, can you simply get a single TRS female to XLR male adapter than can be used with all such headphones using the original SE TRS cable. I think that was what newer guy was asking and in any event it is a question I have. I have a number of cans I want to run balanced and the idea of a single adapter is more appealing than a slew of new balanced cables.
     
  15. johnjen

    johnjen Doesn’t want to be here but keeps posting anyways

    Pyrate
    Joined:
    Sep 29, 2015
    Likes Received:
    505
    Trophy Points:
    93
    Location:
    Well Grounded
    The "cheap adapter' is still a Single Ended termination.
    As you can see there are only 3 pins in the xlr connector.

    Balanced requires 4 wires that are not intrer-connected in any way (other than at the load - drivers).

    Many true balanced amps are 2 separate amps for each channel (4 amps total) that don't use the chassis ground in terms of their outputs, they 'float' independent of ground.

    4pin xlr configurations are 'rare', mostly because hardly anyone, outside of headphones, actually uses them.
    And even headphone configurations haven't 'standardized' on the 4 pin xlr format, since many still support the dual 3 pin xlr configuration.

    And as was mentioned converting from a 4 wire (true balanced) load (the drivers themselves) to 3 wire (single ended) is easy since all you are doing is inter-connecting the 2 minus connections together, while leaving the positive connections separate, thus 3 wires.
    BUT
    Once any 2 wires (the minus connections) are connected together that means that configuration can't be made balanced, until there are 4 independent connections once again.

    IOW an adapter that changes the load the amp sees from 4 wires to 3, will work (balanced to single ended), but trying to use a 3 wire cable and then add a 4th wire to the common minus wires can't work as a balanced connection.

    I hope this is clear enough.

    JJ
     
    Last edited: Jan 25, 2016
  16. Thenewerguy009

    Thenewerguy009 Friend

    Pyrate
    Joined:
    Oct 11, 2015
    Likes Received:
    385
    Trophy Points:
    63
    So a 4 pin XLR cable won't work on an single ended amplifier even with a 6.35mm adapter?
    I never knew that.
     
  17. Merrick

    Merrick A lidless ear

    Pyrate
    Joined:
    Jan 6, 2016
    Likes Received:
    12,569
    Trophy Points:
    113
    Location:
    Portland, OR
    Wait, you're saying that if your headphone is already wired for balanced, you can use a TRS->XLR adapter and you'll get the fully balanced signal even though you've got an SE connection in the chain? That is news to me.
     
  18. purr1n

    purr1n Desire for betterer is endless.

    Staff Member Pyrate BWC
    Joined:
    Sep 24, 2015
    Likes Received:
    89,777
    Trophy Points:
    113
    Location:
    Padre Island CC TX
    SE amp phono out -> SE to balanced adapter -> balanced 4-pin XLR termination on headphone = OK
    Balanced bridged amp 4-pin XLR out -> balanced to SE adapter -> SE phono termination of headphone = BOOM! (most likely).
     
  19. Thenewerguy009

    Thenewerguy009 Friend

    Pyrate
    Joined:
    Oct 11, 2015
    Likes Received:
    385
    Trophy Points:
    63
    Ah
    So say the new Sennheiser HD800S that comes with two cables (SE & balanced), the only one really needed is the balanced one & a cheap SE adapter for it.
    The extra single ended cable that they sell with it is technically unneeded then?
     
  20. dubiousmike

    dubiousmike Friend

    Pyrate
    Joined:
    Sep 27, 2015
    Likes Received:
    269
    Trophy Points:
    63
    Location:
    Los Angeles
    Yup. In fact, many of us have already made such a cable and adapter by snipping the stock se cable near the end, add xlr4 male to the headphone side and xlr4 fem to the cutoff. Done. The extra cable is total superfluous nonsense.
     

Share This Page