Audio balanced output to unbalanced input - a brief visual guide

Discussion in 'General Audio Discussion' started by atomicbob, Oct 29, 2018.

  1. atomicbob

    atomicbob dScope Yoda

    Friend BWC MZR
    Joined:
    Sep 27, 2015
    Likes Received:
    13,167
    Dislikes Received:
    4
    Trophy Points:
    113
    Location:
    On planet

    This guide will hopefully demystify interconnection of balanced and unbalanced audio devices, specifically balanced outputs to unbalanced inputs. It is in no way comprehensive or complete. For further information please reference the following links:

    RaneNote 110:
    https://www.ranecommercial.com/kb_article.php?article=2107
    https://www.rane.com/note110.html

    Bill Whitlock’s excellent application note AN-003
    http://www.jhbrandt.net/wp-content/...tion_of_Balanced_and-Unbalanced-Equipment.pdf
    pg 3 balanced to unbalanced interfaces, which is from
    http://www.jensen-transformers.com/


    Consider the following diagram in which it is desired to send a signal from an audio device balanced output on the left to an unbalanced audio device input on the right. Signal levels will be ignored for this example.
    01a bal out to unbal in - 3 open.png
    There are two active outputs on the balanced output presenting opposite polarity of the audio signal. The simplest approach, but not necessarily the best, is a cable employing a female 3-pin XLR to 2-pin TS ¼” plug or RCA plug. Pin 2 from the XLR will connect to center of RCA or Phone Plug. Pin 1 ground on XLR will connect to sleeve on the unbalanced plug. Pin 3 on the female XLR remains unconnected (NC.)

    So what could possibly go wrong?


    Plenty, depending on alternate / additional ground paths between the two devices. Power supply noise in the grounds, RFI / EMI, DC offsets between ground references, etc. Additionally there are some balanced devices that use Pin 3 hot instead of Pin 2 hot, though that would only create an inversion in the signal in this example.
    01b bal - unbal - noise + offset voltage on gnd.png

    Without knowing specific implementation details of analog signal ground to chassis ground to mains earth ground for both devices we can’t predict whether this simple interconnection method will produce satisfactory, noise free results.
    01c bal-2-unbal-connect-gnds.png

    Commercially available adapter cables are fabricated with several different connection schemes. Some are implemented with pin 3 NC as in the example given above and some with female XLR pin 1 and 3 connected as shown below:
    02 bal out to unbal in 3-1 connected.png

    Depending on the balanced audio device output design and the ability to current limit active devices in the case of output short circuit, the following scenario may develop:
    03 bal out to unbal in 3-1 connected.png
    This is clearly not desirable, especially for active device outputs. Permanent damage may occur. If unsure, don’t do it. You have been warned.


    If the balanced audio device output uses transformers, the situation may be reversed. The cable that worked well for active device outputs leaves one side of the transformer output unconnected. This is not going to work well.
    05 bal - unbal - xfmr - 3 open.png

    Connecting XLR pins 2 and 3 is the correct solution for this situation.
    06 bal - unbal - xfmr - 3-1 connected.png

    An external transformer is another option for connecting active device balanced outputs to an unbalanced device input. Many are equipped with a ground switch allowing experimentation to achieve lowest audible residual noise.
    07 bal-2-unbal-connect-ext-xfmr-chassis-gnd.png
    Which solution is the best for a specific pair of devices requires knowledge about the two audio components connection implementations. It also doesn’t hurt to have a DMM and be familiar with resistance and impedance measurements. The issue of level mismatch, eg. +4 dBu vs -10 dBV was not addressed in this brief visual guide. That is a topic for another post.
     
    Last edited: Jun 27, 2020
  2. drgumbybrain

    drgumbybrain Science Nut

    Friend
    Joined:
    Jan 9, 2017
    Likes Received:
    2,111
    Dislikes Received:
    44
    Trophy Points:
    93
    Location:
    Lives in Fortaleza, Heart in Girona
    Thanks this is very helpful. Do you think the transformer like Radio Iso or Jansen can degrade the sound?
     
  3. atomicbob

    atomicbob dScope Yoda

    Friend BWC MZR
    Joined:
    Sep 27, 2015
    Likes Received:
    13,167
    Dislikes Received:
    4
    Trophy Points:
    113
    Location:
    On planet
    This will be potentially unsatisfying.

    Every component alters a signal in some way. Whether it is a degradation or improvement depends on personal preferences and priorities. Hiss, hum, surface noise, wow, flutter, etc. cause me distractions from the sound. For example, if an isolation transformer eliminates hum audibility for a given setup, I may find any other sound alteration, whether real or imagined, to be much lower on my list of priorities and preferences.

    Good transformers are expensive. Jensen transformers are used in mic preamps, stage snake splitters, and many high-end studio analog processing components for good reason. There are other manufacturers of quality transformers, also expensive.
     
  4. roscoeiii

    roscoeiii Acquaintance

    Contributor
    Joined:
    Feb 15, 2016
    Likes Received:
    81
    Dislikes Received:
    9
    Trophy Points:
    28
    Many thanks for this post. My follow-up post on the Convert 2 thread spurred it I believe. From @purr1n 's reply it seems that 1:1 transformers may have advantages to 4:1 unless you need to take that hot pro balanced output.
     
  5. purr1n

    purr1n Finding his inner redneck

    Staff Member Friend BWC
    Joined:
    Sep 24, 2015
    Likes Received:
    61,829
    Dislikes Received:
    108
    Trophy Points:
    113
    Location:
    Antarctica
    Anything unnecessary in the signal path degrades sound quality to some extent. If you can avoid line level transformers, then avoid them.

    In the specific case of Yggdrasil, I felt the degradation in sound quality was less from the iso transformers than the extra active summing stage for the SE outputs. The Yggdrasil seems to take the balanced outputs through a another circuit for SE out. For other DACs, this may not be the case.

    Keep in mind that transformers will increase distortion, particularly in the bass as the signal approaches DC, and especially as the level increases. When this happens, things rapidly go south. However, there is no need for concern. If you look on a spectrum analyzer, there is little content in the lower bass, and if there is any, the signal levels are low. Movies with explosions and 20Hz test signals at +22dbU are the exception. The way around this is a bigger transformer, but it's generally better to go as small as you need it - less resistance and inductance which results in other issues, bandwidth, etc.

    So fair warning, Amirites should definitely avoid transformers and utilize active iso-boxes with opamps. Those measure excellently and are dirt cheap.

    People who use their ears should consider iso transformers for balanced to unbalanced and compare against built-in unbalanced out of their sources.

    As an aside, watch Studio City. That vintage Neve mixing console that the Foo Fighters guy bought and salvaged was transformer coupled with discrete circuits.

    The transformers in studio gear got replaced sometime in the 80s with opamps. Opamps are dirt cheap. Transformers are not. Personally, I think opamps sound like shit compared to transformers when used for isolation. It's important to note that distortion figures for music recordings run through line level transformers are below the threshold of hearing. We are talking about two or there zeros after the decimal point. But this is probably of no consequence to Amirites who insist on the lowest possible THD and seven zeros after the dot.
     
    Last edited: Oct 29, 2018
    Grattle, msommers, OJneg and 3 others like this.
  6. Azimuth

    Azimuth Rando

    Joined:
    Sep 29, 2018
    Likes Received:
    49
    Dislikes Received:
    0
    Trophy Points:
    13
    Location:
    Somewhere down south
    Reverse the audio path of that last picture and you have a passive DI box.

    Let's not forget that the Neve preamps are not sought after for their crystal clear clean sound. They are sought after for their coloration. And holy crap they sound big. But coloration is part of the engineer's and producer's job to manage for what they want...Neve, API, Trident, SSL, Focusrite...there is a reason Beck only records and mixes on a Trident 80B console.
     
    Grattle, Psalmanazar and purr1n like this.
  7. murray

    murray Friend

    Friend
    Joined:
    Sep 29, 2015
    Likes Received:
    370
    Dislikes Received:
    0
    Trophy Points:
    63
    Location:
    N.Z.
    Heh, heh, I got half way into looking up “Amirites” (thinking maybe it was a religious cult) when the penny dropped. Good one @purr1n!
     
  8. atomicbob

    atomicbob dScope Yoda

    Friend BWC MZR
    Joined:
    Sep 27, 2015
    Likes Received:
    13,167
    Dislikes Received:
    4
    Trophy Points:
    113
    Location:
    On planet
    Measurements for the Radial ISO Twin (Jensen transformers) may be found here
     
  9. rtaylor76

    rtaylor76 Can't wipe his tag

    Friend
    Joined:
    Nov 9, 2015
    Likes Received:
    2,283
    Dislikes Received:
    91
    Trophy Points:
    113
    Location:
    Nashville, TN
    Home Page:
    I read somwhere that one of Radial's Jensen transformer DI boxes was wired out of polarity. Might want to check that @atomicbob.
     
    Psalmanazar likes this.
  10. atomicbob

    atomicbob dScope Yoda

    Friend BWC MZR
    Joined:
    Sep 27, 2015
    Likes Received:
    13,167
    Dislikes Received:
    4
    Trophy Points:
    113
    Location:
    On planet
    That was the Radial Pro48, early production units only. You can read about that incident here:
    https://www.gearslutz.com/board/so-...ime/911576-my-di-signal-always-out-phase.html

    Most Radial boxes have polarity switches on them. The Pro48 is the stripped down version of the J48. The ISO Twin has individual channel polarity reverse switches, so even if a wiring error had occurred, it is easily fixed in use. However, my ISO Twin is wired correctly, as are the vast majority of Radial boxes. Again, the Pro48, early production units, were the unusual exception.
     
    A1Gear and rtaylor76 like this.
  11. rlow

    rlow A happy woofer

    Friend
    Joined:
    Jul 18, 2017
    Likes Received:
    4,417
    Dislikes Received:
    4
    Trophy Points:
    113
    Location:
    Canada
    Was considering picking up a Radial J-Iso to put between Yggdrasil XLR outputs and Saga SE inputs, but unsure if this is the right device for this purpose, as I believe it is a 4:1 transformer?

    http://www.radialeng.com/product/j-iso/
     
    A1Gear likes this.
  12. atomicbob

    atomicbob dScope Yoda

    Friend BWC MZR
    Joined:
    Sep 27, 2015
    Likes Received:
    13,167
    Dislikes Received:
    4
    Trophy Points:
    113
    Location:
    On planet
    A1Gear, rlow and rtaylor76 like this.
  13. rtaylor76

    rtaylor76 Can't wipe his tag

    Friend
    Joined:
    Nov 9, 2015
    Likes Received:
    2,283
    Dislikes Received:
    91
    Trophy Points:
    113
    Location:
    Nashville, TN
    Home Page:
  14. rlow

    rlow A happy woofer

    Friend
    Joined:
    Jul 18, 2017
    Likes Received:
    4,417
    Dislikes Received:
    4
    Trophy Points:
    113
    Location:
    Canada
  15. Armaegis

    Armaegis Friend

    Friend BWC
    Joined:
    Sep 27, 2015
    Likes Received:
    4,111
    Dislikes Received:
    2
    Trophy Points:
    113
    Location:
    Winnipeg
    An autoformer trades the isolation for potentially better measurements (one less "conversion" step... sorta-ish).
     
  16. atomicbob

    atomicbob dScope Yoda

    Friend BWC MZR
    Joined:
    Sep 27, 2015
    Likes Received:
    13,167
    Dislikes Received:
    4
    Trophy Points:
    113
    Location:
    On planet
    Yggdrasil has an output impedance of 75 ohms with 0 dBFS producing +14 dBu maximum.
    Saga has input impedance of 10K ohms and can handle +22 dBu maximum through the system.
    So it depends on what is downstream of the Saga whether to look at J-ISO or ISO Twin. If you want to use attenuation via Saga then pick the ISO Twin. If you want little attenuation via Saga then pick J-ISO.
     
    A1Gear and rlow like this.
  17. purr1n

    purr1n Finding his inner redneck

    Staff Member Friend BWC
    Joined:
    Sep 24, 2015
    Likes Received:
    61,829
    Dislikes Received:
    108
    Trophy Points:
    113
    Location:
    Antarctica
    You can ask around here about the 4:1 Jensens (Jensen Isobox and Radial). I saw them mostly used with the OG Gungnir MBs and maybe Yggdrasils to warm up the sound because those DACs sounded drier than the latest A2 revisions. They may have some cost to sound quality compared to the direct unbalanced out. Maybe the best way to put it was a different kind of veil? Less gray but a slight sheen of veil (4:1 Jenson converted to unbalanced) compared to grayer (direct from DAC unbalanced).

    I think the Cinemag 1:1 (and by extension, the Jensen - since same wire colors and ancestry) is more transparent and more noticeably better than the SE outputs, at least with the A2 Schiit DACs, compared to the 4:1. However, the differences are not huge between direct unbalanced and balanced to the 1:1 iso (I did volume match). The downside is there are variants to the transformers (you need to pick the right one for your application) and you'd have to DIY if you really wanted to use the 1:1s. Another possibility is that I might be getting better sound quality because I physically twisted together and soldered all the connections, despite using the shittiest cables I have in the house. Some members, like @OJneg feel that connectors (or lack of connectors) are more important than the wires when it comes to "audiophile" cable-lery.
     
    Last edited: Nov 1, 2018
    A1Gear, jexby and rlow like this.
  18. OJneg

    OJneg The Most Insufferable

    Friend BWC Banned
    Joined:
    Sep 25, 2015
    Likes Received:
    3,752
    Dislikes Received:
    37
    Trophy Points:
    113
    Location:
    Grand Rapids
    Sounds nuts, but I have heard differences after cleaning old connectors on my audio cables. It's not totally unfounded in electrical theory; you can look up passive intermodulation in nickel and ferrous metals. Makes you wonder about how many other matsci factors (gold grids, metal oxide junctions, etc) in our signal path might be audible but not talked about.
     
    yotacowboy likes this.
  19. Clemmaster

    Clemmaster Friend

    Friend
    Joined:
    Sep 28, 2015
    Likes Received:
    1,782
    Dislikes Received:
    4
    Trophy Points:
    113
    I piss on my connectors every morning, to clean up the oxide.
     
  20. gixxerwimp

    gixxerwimp Professional tricycle rider

    Friend
    Joined:
    Sep 30, 2015
    Likes Received:
    3,033
    Dislikes Received:
    13
    Trophy Points:
    113
    Location:
    small island claimed by China
    Sorry to necro this thread, but was led back here by a recent All Purpose Advice post.

    Which connection scheme do the SC1.5XJR cables in your first Synergy post have? I searched but couldn't find.

    https://superbestaudiofriends.org/index.php?threads/system-synergy-special-sound.54/
     

Share This Page