Balanced output from AD1937

Hello Mark,

This is a good question. I just checked around for designs on different eval boards and others that were never released and we do not have any designs that keep the audio balanced. Our eval board for the ADAU1966 just has passive filters and then connects that to a dSub as a differential output. No Opamps. Most of the time our customers will take the output of the DAC and process it through level controls, faders,  or send it to amplifier chips so the signal needs to be single ended. 

Thinking about this you would want to filter the outputs so simply treating each N & P output as a single ended output is probably the simplest way to do this. Here is the single ended filter out of the AD1938 datasheet:

So duplicating this for each "N" and "P" output would give you a filtered differential signal with a fair amount of drive. Then add one more amp stage to gain up the signal up to get the +18dB signal you are looking for. You will need to have at least +/-12V power rails but I think you know that. 

I like a 2-stage approach so the filter can be a clean and good design then get the gain out of the next stage. 

You may want to post a question up on the Linear Amplifiers portion of the forum. They are really good at OpAmp designs and know our Op-Amp catalog better than I do. I would probably start with the Amplifiers Homepage: (+) Amplifiers - EngineerZone (analog.com)

Dave T

Thanks, Dave.  I've posted a question over on the Amplifiers Homepage, hopefully someone there can help me.

Out of interest, do you know the Vrms from the output of that DAC filter you included in your response?

Kind regards,

Mark

Hello Mark,

The filter in the EVAL-ADAU1452RevBZ board is similar except it uses a voltage follower configuration. I am able to measure this easily because it is right here at my desk and so is the Audio Precision. 

So with a 0dBFS signal in the DSP going out to the DAC. I measured the following in different standards for you.

-1.11dBV

880mV rms

+1.108 dBu

I hope this helps. 

Dave T

Thanks, Dave!  Just so I know, was your 0dBFS signal a pure sine wave?  (so I can calculate Vp-p and consider crest factor etc).

Do you know the LPF response in that circuit for the DAC output?  Is it recommended?  Or do you think I should design a different LPF configuration?

Cheers,

Mark

Thanks, Dave!  Just so I know, was your 0dBFS signal a pure sine wave?  (so I can calculate Vp-p and consider crest factor etc).

Do you know the LPF response in that circuit for the DAC output?  Is it recommended?  Or do you think I should design a different LPF configuration?

Cheers,

Mark