Author Topic: Question regarding balanced / unbalanced audio circuits  (Read 2118 times)

0 Members and 1 Guest are viewing this topic.

Offline david77

  • Frequent Contributor
  • **
  • Posts: 745
  • Country: de
Question regarding balanced / unbalanced audio circuits
« on: September 19, 2011, 07:14:40 PM »
Hello Eevbloggers.

I'm currently working on developing a very specialised piece of professional audio equipment.
While thinking about the circuits some questions did arise that I have not been able to answer for myself yet.

Basically the device will have several balanced inputs and several balanced outputs. Internally all connections will be
unbalanced - because the device will also have to interface to unbal. equipment - , so I need a BALUN / UNBAL circuit.
For the balanced inputs I chose TI's INA217, because it is easy to use and the datasheet promises excellent properties.



Here's the basic circuit I came up with, basically just what TI say in the datasheet. Pin 1 & 8 are unconnected to give
a gain of 1.

When I apply 1,55V (+6dBu) at the bal. input I suppose I'll also get 1,55V at the unbal. output.
But what would happen if I were to connect the - bal. and the GND line at the input to connect an unbalanced device to it, as is common practice? Would that mean I'd loose 6dB?

I haven't tested any of this yet, it's early days and this just popped into my head.

Also, a very basic question: How do I measure across a balanced line using a mV meter that only has a negative and a positive terminal, e.g. your run-off-the-mill multimeter? I suppose across the hot/cold leg, right?
Same problem, just the other way round: How do I get a signal from my signal generator in? Will I have to make up a circuit to balance the sig. gen's output, or is there an easier way?
Maybe I'm just thinking to complex, that is often my problem  ;).

Thanks in advance

Offline DrGeoff

  • Frequent Contributor
  • **
  • Posts: 408
  • Country: au
    • AXT Systems
Re: Question regarding balanced / unbalanced audio circuits
« Reply #1 on: September 19, 2011, 07:54:39 PM »
Forget about the GND terminal completely for your input signal. Think of it like a transformer, where the signal is applied across two terminals and there is a high impedance between each terminal and 'ground' and a low impedance between the two terminals.
The INA217 is an instrumentation amplifier, replacing the SSM2017 device. The GND connection in your circuit is probably meant for the 'shield' connector, as it looks like you plan to use phantom power over the balanced pair, otherwise it's not really necessary since all of the signal is carried in the balanced pair. You also take your measurements directly across the pair, since the circuit GND has nothing to do with the balanced input signal.
Your signal generator probably has a balanced output (ie not referenced to any chassis eartch etc), so it can be connected across the input. If this is not the case you can use an audio transformer to isolate the signal and make it fully balanced for testing.
If you did connect the cold input to the circuit GND point, then you would reduce the output by 6dB.
Was it really supposed to do that?

Alex

  • Guest
Re: Question regarding balanced / unbalanced audio circuits
« Reply #2 on: September 19, 2011, 09:31:15 PM »
You might find this app note useful.

Offline david77

  • Frequent Contributor
  • **
  • Posts: 745
  • Country: de
Re: Question regarding balanced / unbalanced audio circuits
« Reply #3 on: September 19, 2011, 10:07:35 PM »
@Dr. Geoff: Thanks. That explains it perfectly.

@Alex: Interesting read, I wish I could splash out on some Jensen transformers but alas, will have to live with the INA217.

Alex

  • Guest
Re: Question regarding balanced / unbalanced audio circuits
« Reply #4 on: September 19, 2011, 10:19:22 PM »
splash out on some Jensen transformers

I believe that is the purpose of that app note, but you can filter out the contribution of the marketing department. Some topologies are interesting.

Offline david77

  • Frequent Contributor
  • **
  • Posts: 745
  • Country: de
Re: Question regarding balanced / unbalanced audio circuits
« Reply #5 on: September 21, 2011, 12:31:38 AM »
I just found a much more in depth essay on balanced lines here
http://sound.westhost.com/articles/balanced-2.htm

It's written by the same author as that Jensen app note, Bill Whitlock.


Offline vk6zgo

  • Super Contributor
  • ***
  • Posts: 2413
  • Country: au
Re: Question regarding balanced / unbalanced audio circuits
« Reply #6 on: September 21, 2011, 12:40:39 PM »
I always thought that the reason for going away from 600 Ohm input & output impedances was to save money on transformers,so it seems like a backward step to reintroduce them(this time with low o/p Z & high input Z.

The author points out that matched 600 ohm systems give best power transfer,whereas  high Z inputs bridging across low Z outputs give best voltage transfer.
I would question the importance of this distinction---with modern components we have ample gain in both situations,so the only advantage as far as I can see is that it  is possible to bridge a number of inputs across one output,without having special splitter amps.

A point that was not clarified,is that 0.775 volts is the voltage level which appears across a 600 ohm termination for a level of 1 mW
---a level of 0dBm.
This allows us to test in circuits with a combination of 600 Ohm input/output devices & low Z o/p/High Z input devices without changing the settings on our level meter.
600 ohm input/output devices were used successfully for many years in Broadcasting without any major dramas,as have High Z/Low Z devices since their introduction.
Commercially made equipment meeting all required specifications have been available for many years,so the design of these things "isn't rocket science"!

VK6ZGO




Offline ejeffrey

  • Super Contributor
  • ***
  • Posts: 1341
  • Country: us
Re: Question regarding balanced / unbalanced audio circuits
« Reply #7 on: September 21, 2011, 08:53:31 PM »
The primary reason for the existence of impedance matched circuits in audio is for the telephone system.  This is actually for two reasons: the obvious one is that telephone lines are not short compared to audio wavelengths, so it is a genuine transmission line.  The less obvious reason is that the telephone system needs to to carry bidirectional audio over a a single twisted pair.  The impedance matched line allows the transmit and receive signals to coexist without creating positive feedback loops at repeaters.  Ideas about impedance matching audio circuits and 600 ohm impedances are mostly holdovers from the telephone system and are not required or useful in local audio distribution systems.  Microphone still often have high output impedance, often on the order of 600 ohm, but that is simply due to their nature -- voice coils have a lot of turns of very thin wire.  Microphone preamps should still generally have high impedance.

All of this is mostly independent of balanced vs. unbalanced and transformer coupling.  It is perfectly possible and quite common to have a low impedance source transformer coupled and balanced with a high impedance receiver, also transformer coupled.  Likewise, you can have an impedance matched unbalanced connection.  You can even have transformer coupled unbalanced signals, although there are few reasons why you would do this.  With balanced signals you will typically have non-negligible output impedance for the reason that the most important aspect of a balanced circuit is for the + and - phases to have the same impedance.  This is easier to guarantee if you control the impedance with resistors rather than letting it be set by the device limitations of the output amplifier.

Offline vk6zgo

  • Super Contributor
  • ***
  • Posts: 2413
  • Country: au
Re: Question regarding balanced / unbalanced audio circuits
« Reply #8 on: September 23, 2011, 03:30:04 PM »
For most of its many years of existence,Radio Broadcasting was dependent on the public telephone system for transferring programs around the Networks,& ultimately,to the Transmitter sites,using Program lines which were specially equalised for the purpose.

In Australia,these were often 1200 Ohm balanced lines,which terminated in 1200:600 Ohm transformers,& then into the 600Ohm inputs of line amps,which in turn fed 600 Ohm inputs of limiters,then on to the 600 Ohm inputs of Transmitters.

The 1200 ohm lines were usually physical lines from the local telephone exchange.The long distance transmission between exchanges was very rarely at baseband audio frequencies,being mainly carrier telephony on lines,microwave or UHF radio comm systems.

Around about the time that Transmitter feeds began to be direct STLs or Satellite links,the use of low Z output/High Z input became more prevalent,but many Transmitter program input racks retained the 600 Ohm input/output system,as did many Studios.
Broadcasting for many years was the benchmark in Audio line equipment,as well as its main market.

VK6ZGO   
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf