Balanced headphone amp with DAC, Community project perhaps?
May 21, 2010 at 4:47 PM Post #16 of 41
I'm not too great at explaining things, so I'm probably not saying the right things to be understood.  :)
 
The way I see a balanced setup, or at least how I would prefer it to be:
 
Each channel has a +/- connection as opposed to say L/GND or R/GND.  When current runs from neg to pos the voltage is below 0v and if it runs pos to neg then the voltage is above 0v.  Is this the voltage swing?
 
I'm assuming that this would be the same for L/R channels?  So that would be four channels.  The way I see the current is like a belt around two wheels, the belt on one side of the wheels is + and the other -.  So if you pull one side the other side will go in the opposite direction.  So it's going full circuit, from the dac through the amp through the headphone back through the opposite amp and back to the DAC and flowing until it swings in the opposite direction.
 
I don't think it's the balanced i'm not understanding, I think it's the output of the DAC itself.  I know the DAC has L+ and L-, so the current would need to flow from one to the other.  If the ground is used as a reference, it's still shorting L+/L- out.  So it should still be balanced.
 
May 23, 2010 at 7:04 AM Post #17 of 41
I think you are misunderstanding current as being something that 'flows' at all. the negative phase DAC does not accept signal, it outputs a signal of the opposite phase. in some applications or components this is in reference to ground, others it is in reference to the positive phase and in some there is an oscillator that swings both directions  'VOSC' and '0' can be biased at a voltage of your choosing 'VREF'
 
May 23, 2010 at 10:17 AM Post #18 of 41
So that means a DAC has 8 different connections for audio?
 
L/R have both positive and negative points, but the positive and negative points also have grounds?
 
I think it's slowly making sense now.  I was under the impression you could take audio directly out of the DAC itself.
 
May 23, 2010 at 10:46 AM Post #19 of 41
What he is saying is that the DAC designers have to take many different situation into account so they
design a great deal of flexibility into the outputs. If you look at the PCM1794 data sheet you will see that
half of the 28 pins deal with the Analog section. You can even power the left are right separately.
They even split the analog and digital sections so you can split the ground plane down the middle of
the chip. Don't let this scare you though...you only really need the L/R + and - outputs.
 
Quote:
So that means a DAC has 8 different connections for audio?
 
L/R have both positive and negative points, but the positive and negative points also have grounds?
 
I think it's slowly making sense now.  I was under the impression you could take audio directly out of the DAC itself.



 
May 23, 2010 at 11:03 AM Post #20 of 41
Lol, I thought that's what the case was originally.
 
Unless different people have different ideals about what balanced should be?
 
I dunno, I have thumping headache and need to lie down.  lol
 
May 23, 2010 at 11:42 AM Post #21 of 41

The usual reason for using balanced signals is for the CMRR.
If you need to run a cable over a long distance, any noise picked
up will be the same phase on each signal line. By using a differential
amp on the receiving end, all this noise is canceled out.
 
Quote:
Lol, I thought that's what the case was originally.
 
Unless different people have different ideals about what balanced should be?
 
I dunno, I have thumping headache and need to lie down.  lol



 
May 23, 2010 at 2:08 PM Post #23 of 41
I was saying more than that, but i'm not sure we want to go into it all just yet :D. I recommend you stop thinking of current as something that flows from the positive terminal to the negative terminal, that just doesnt happen. circuit digrams are often drawn that way to help people visualize it, but in a differential amp especially, it couldnt be further from the truth,
 
both positive and negative phase signals are 2 halves of the same thing, one is not the end result of the other. we arent talking about a positive charged partical and a negative charged partical, we are talking about the 2 opposite phases of a wave, + swinging one way and - the other, both in reference to a chosen '0' or center point, which may or may not be actually neutral. in fact most dacs will be biased one way or the other, or allow you to set your own bias for VOSC. I hope I was clearer this time, because I obviously didnt last time.
 
the above links in quite nicely with what Avro-Arrow is saying about CMMR, but i'll leave you to do some research
 
May 23, 2010 at 2:40 PM Post #24 of 41
I wasn't suggesting that the current only flows from positive to negative.
 
So for an audio triangle waveform starting at 0v, the voltage will increase proportionately with time from positive to negative to give the rise of the top part of the waveform.  One the voltage has reached it's apex, the voltage will drop, again proportionately with time down to 0v (1st half of phase).  Then the current would need to flow from negative to positive to produce a negative value and increase in voltage proportionately with time to drop to bottom part of the trough, once the voltage has reached it's apex, it will then drop back to 0v (2nd half of phase).
 
This is the best way I can think of to visualise in my head how the alternating currents make up the waveform.
 
May 23, 2010 at 5:46 PM Post #25 of 41
I'll say it a different way...
 
The full signal is output from the + output.
The same full signal is also output from the - output but 180 degrees out of phase.
That is how a differential signal works.
 
May 23, 2010 at 6:21 PM Post #28 of 41

Actually, in a current output DAC like the 1794, the output is increasing and decreasing current, not voltage.
You use an I/V stage to convert the current to voltage. If you look at the schematic in the 1794 data sheet,
the first op amp is the I/V stage.
 
Quote:
Oh, I'm more confused now.
 
So what I was going on about is pure gibberish?  Not being funny, just was under the impression that audio is an AC and not a DC current.



 
May 23, 2010 at 6:53 PM Post #29 of 41
Ahhh.
 
I hate asking all these questions....  But yes, everyone is right, I'm very confused about how the analogue out put from the Dac works, I will do some reading on it next week.
 
As long as the dac can output a pure balanced output, that's all I should be worried about.  :D
 
I'll leave this to the pros.  lol
 
May 23, 2010 at 8:00 PM Post #30 of 41
I said I would draw something...
Here is one channel.
In the diagram, Headphones means on one side,
not both channels.
Your headphones would need to be bi-wired for
this to work.
While this would most likely work, it is meant to
be more of a study drawing.
 

 
Revised to correct error in filter feedback loop
 

Users who are viewing this thread

Back
Top