Hi,
So I saw this image and it made me realise that I might not understand balanced, I just thought it separated the grounds so you could double the voltage and that was better for longer cables, resistance to interference and making (headphone) amps play louder (I think a power amp sort of "multiplies" sound, so say 2V x 10W is 20W then 4V x 10W is 40W I think this is probably way oversimplified).
Then it occurred to me (I was told that the analogue section of most DACs is what let them down) that a balanced DAC might have twice the (what I'll call) "bandwidth" like 4V means more "signal". I read 3dB more dynamic range. Do you get more "resolution"?
I also read some ladder DACs improve balanced, like maybe you can have more steps in your ladder?
The CamillaDSP setup in the image is using two channels on a DAC changed into one channel balanced, is that going to "measure" better?
I was also chatting about (digital) volume and that you lose headroom, do you get more headroom with balanced, I have got to admit that I don't fully understand (a lot of advanced audio stuff including) headroom.
Can someone please sum things up for me?
Thanks.
So I saw this image and it made me realise that I might not understand balanced, I just thought it separated the grounds so you could double the voltage and that was better for longer cables, resistance to interference and making (headphone) amps play louder (I think a power amp sort of "multiplies" sound, so say 2V x 10W is 20W then 4V x 10W is 40W I think this is probably way oversimplified).
Then it occurred to me (I was told that the analogue section of most DACs is what let them down) that a balanced DAC might have twice the (what I'll call) "bandwidth" like 4V means more "signal". I read 3dB more dynamic range. Do you get more "resolution"?
I also read some ladder DACs improve balanced, like maybe you can have more steps in your ladder?
The CamillaDSP setup in the image is using two channels on a DAC changed into one channel balanced, is that going to "measure" better?
I was also chatting about (digital) volume and that you lose headroom, do you get more headroom with balanced, I have got to admit that I don't fully understand (a lot of advanced audio stuff including) headroom.
Can someone please sum things up for me?
Thanks.
Hi,
Balanced is a short for 'balanced impedance'.
In no way it implies a symmetrical drive with a differential signal, which is what you have interest into ( despite it's often linked with balanced impedance but not always...).
I suggest you to read this link, it will give you most info you are looking for:
https://sound-au.com/articles/balanced-2.htm
Balanced is a short for 'balanced impedance'.
In no way it implies a symmetrical drive with a differential signal, which is what you have interest into ( despite it's often linked with balanced impedance but not always...).
I suggest you to read this link, it will give you most info you are looking for:
https://sound-au.com/articles/balanced-2.htm
Headroom is not an easy to grasp concept at first, but it is in fact simple.
Let's imagine someone with an arbitrary size of say 6" ( 1,8m) in a room where ceiling is 7" (2,1m). If he start to jump he might hit the ceiling and hurt his head.
Now imagine the same person in a room where ceiling is 10" (3m). If he jump, then there is low chance the head hits anything, right?
In that case, the difference ( distance) between head and ceiling is headroom.
In practice in audio, the person is the signal to reproduce, the ceiling height is the headroom, if head hit ceiling then disortion occurs. If signal is 1,23V (rms) and the whole system is able to cope with 3,88V(rms) without distorting then you have marging before distortion occurs, and it is equal to 10db of headroom ( 1,23V is +4dbu and 3,88v is +14dbu).
Headroom is thus marging left in case a peak ( a jump) occurs. In practice 10 to 12 db of headroom is enough to garantee distortion will never occurs ever ( in pro context as a mastering studio this is the value you'll meet with the most conservative engineers).
That said it is true for the analog part of signal transfer, in no way it'll help regain any headroom lost in digital conversion as it happen before the analog stage.
To not loose resolution (digital) you'll have to ouput ( digital) signal as close as possible to 0dbfs which is the max digital output level. This time it's not rms anymore but peak value (which is rms x 1,414).
About differential signal:
https://en.m.wikipedia.org/wiki/Differential_signalling
Last edited:
The image linked might be confusing because it use two output to create a differential drive stage.
If you take a closer look you'll see there is '0db inv' noted next to one of the pair of output ( the other one is noted '0db'). It means one output is polarity reversed wrt to the other one and both have same amplitude - what you wrongly called 'bandwidth' in your first message, bandwith is frequency related term, not related to amplitude-( the meaning of 0db).
So they satisfy what is explained in the wikipedia link about differential.
I can assume both will have the same output impedance too ( as they are the same output circuitry) and so they are balanced impedance too.
I hope it's clearer now.
I can't help about ladder resistor dac but being made from discrete components, matching is easier so balancing could be better ( than what? I can't tell i never studyed in depth converters schematics...).
And maybe read again my previous message as i edited them to improve explanation ( bad habit i have, i'm guilty!) 😉
If you take a closer look you'll see there is '0db inv' noted next to one of the pair of output ( the other one is noted '0db'). It means one output is polarity reversed wrt to the other one and both have same amplitude - what you wrongly called 'bandwidth' in your first message, bandwith is frequency related term, not related to amplitude-( the meaning of 0db).
So they satisfy what is explained in the wikipedia link about differential.
I can assume both will have the same output impedance too ( as they are the same output circuitry) and so they are balanced impedance too.
I hope it's clearer now.
I can't help about ladder resistor dac but being made from discrete components, matching is easier so balancing could be better ( than what? I can't tell i never studyed in depth converters schematics...).
And maybe read again my previous message as i edited them to improve explanation ( bad habit i have, i'm guilty!) 😉