Differential to single ended output

[krazy]

Hello,

I have audio coming out of a PIC DAC as a differential output (it is giving me both the positive and the negative). I need it in single-ended form (one wire referenced to the circuit's ground).

What op amp circuit will do this for me?

Thank you!

 

[Nigel Goodwin]

You almost certainly have no need for an opamp, but what DAC references in this way?, can you post the circuit you're using?.

 

[krazy]

My dsPIC33FJ does it. Check out the pinout here:

I do not have a circuit to reference because I am not using a circuit on the output, yet.

I am trying to drive headphones. I have a headphone amplifier but that requires a single ended input referenced to circuit ground.

 

[dknguyen]

It's not positive and negative. It's above and below a positive bias voltage (the dsPIC does not have a negative voltage supply). Just use a difference amplifier or differential amplifier with a gain of 1 if you want to re-reference the signal to ground rather than the bias point.

 

[krazy]

It's not positive and negative. It's above and below a positive bias voltage (the dsPIC does not have a negative voltage supply). Just use a difference amplifier or differential amplifier with a gain of 1 if you want to re-reference the signal to ground rather than the bias point.

So are you saying I could plug the dsPIC DAC output right into, say, an LM4808M [link] to drive headphones?

I guess the main thing I'm trying to do with this DAC output is just drive headphones. Right now, if you plug the headphones right into the DAC1RN/P, the sound is kind of low (even when audio is cranked up).

Suggestions?

 

[dknguyen]

The 33F DAC was designed to drive an audio amplifier. It does not have enough power to drive headphones directly since it is a data processing device, not a power device. You would use the bias output voltage of the DAC IN+ for the IC.

Just curious, were you using DC block capacitors when directly connecting the DAC to the headphones?

 

[Speakerguy]

Is this a differential PWM output DAC? If so, I have several ways of helping you. Do you want a PWM amplifier (high efficiency but complex) or a simple linear amplifier (inefficient, lower power, but easy way to drive headphones)? Remember you should be lowpassing the signal if it is a PWM output.

 

[krazy]

The 33F DAC was designed to drive an audio amplifier. It does not have enough power to drive headphones directly since it is a data processing device, not a power device. You would use the bias output voltage of the DAC IN+ for the IC.

Just curious, were you using DC block capacitors when directly connecting the DAC to the headphones?

No, no block capacitors.

The DAC1RP is coming out biased at 1.7 volts (Vdd/2). If I can remove that bias (bring it down to 0) and give it some current to drive headphones, I'd be golden....

 

[krazy]

Is this a differential PWM output DAC? If so, I have several ways of helping you. Do you want a PWM amplifier (high efficiency but complex) or a simple linear amplifier (inefficient, lower power, but easy way to drive headphones)? Remember you should be lowpassing the signal if it is a PWM output.

I dont think so? Sorry, I don't know...

 

[dknguyen]

It's not. It's a regular analog output DAC, except the reference voltage is some bias voltage rather than ground. (It's in the datasheet).

 

[krazy]

Any ideas on how to remove this bias? And give some current?

 

[Leftyretro]

Any ideas on how to remove this bias? And give some current?

A series capacitor will remove DC bias. An amplifier stage will increase current.


Lefty

 

[dknguyen]

Those DC blocking capacitorson the output of the audio amplifier datasheet you attached ARE what remove the bias. For more current you just use an op-amp buffer or amplifier. If you remove the bias BEFORE you amplify it then you need a +/- power supply and op-amp. THat's why the bias and DC block capacitors are there...makes it easier to use.

 

[krazy]

I tried putting a capacitor between the output (DAC1RP) and an oscilloscope and I still get a mean voltage of 1.6V.

 

[krazy]

Ok. I set up a difference amplifier like this:

...with all resistors 10k.

Now, on the oscilloscope, I get the top half of the waveform but the voltages never go down into the negative. Suggestions?

 

[dknguyen]

Negative...is that with the DC Block capacitor? and if so, then where is the DC block capacitor?

 

[krazy]

No it is connected like this...

I get the positive half of the waveform but not the negative half.

DAC1RN and DAC1RP are both positive and biased around 1.65V (Vdd2/2) and I hooked them up to a differential amplifier with unity gain.

 

[krazy]

Nevermind! Got it!! The outputs had to be switched. I knew I saw somewhere on the Microchip forums that the N/P for the DACs had to be swapped. Thanks for the help everyone

 

[audioguru]

You used a differential amplifier circuit that uses a positive and negative supply because the inverting input to the input resistor was biased at the PIC's reference voltage but the non-inverting input was biased at half the reference voltage. Then the output of the opamp idled at 0V. It cannot swing to a negative voltage without a negative supply.

My circuit uses a capacitor so that the output DC voltage of the opamp is at the reference voltage and can swing positive and negative from there.