3
\$\begingroup\$

In the hope of using the standard 3.5mm audio output found in your average PC/smartphone to control (not power) a DC motor, I was measuring the voltage between GND and Tip and noticed the following:

  • When no audio is playing, there's a voltage of about 2 mV.
  • When I output a constant audio signal, the voltage rises initially to 0.5 V, but drops back to 2 mV after 2-3 seconds.

What purpose does this 2 mV bias voltage serve? Does audio output circuitry usually contain DC blocking capacitors, or is DC blocked via some other means?

\$\endgroup\$
1
  • 1
    \$\begingroup\$ PCs can usually detect whether or not anything is plugged in, and in some cases make a guess whether it's headphones, speakers, or a power amp. I suspect the bias voltage is part of that. \$\endgroup\$
    – pjc50
    Commented Apr 9, 2017 at 15:41

2 Answers 2

3
\$\begingroup\$

In regular line level audio, DC voltage does not serve a purpose.

It's conceivable that some devices may use a DC offset in order to detect activity and to go into standby - or turn on - depending on whether there's a DC offset or not, without having to depend on an actual audio signal to be present at all times. But I'm not aware of any standard for this.

Audio output circuitry usually does contain DC blocking capacitors.

(But not always; certain high end "audiophile" power amplifiers do not. With these amplifiers, any DC voltage offset on the input signal can easily blow the speakers. Yes, it's a dumb design.)

A headphone jack does not output line level audio though; it's a low(er) impedance output designed to drive tiny speakers. If the DC voltage on the output of your phone does serve any purpose, I'm guessing that it might have to do with the small remote control for receiving and ending calls, and changing the volume, often found on the cord, although there's often a separate wire for that (mini-jack with four conductors).

\$\endgroup\$
1
\$\begingroup\$

There is actually a good possibility there are no decoupling caps on the output, but rather, the output is ground referenced. See the datasheet for the [WM8994](this is allegedly used the Samsung Galaxy S3, maybe others)1. One of its features is that it has a charge pump to generate the negative voltage needed to output a signal around 0V instead of 1/2 Vcc. The 2mV you are seeing may just be offset error of the output amplifier. Even looking at some RealTek(pc's) chips, they have ground referenced outputs.

It is probably desirable to omit the output cap due to size and additional cost. I have a couple pairs of headphones, one is 16 ohm and the other 32 ohm; an output cap is going to form a high pass filter. If you wanted to set your cutoff at say, 75Hz, you would need about a 130uF decoupling cap. That's going to be physically large, even in ceramic (an 0805 eats up a lot of real estate in a phone, and ceramics are not ideal for audio).

I can't speak for all devices, but I would expect that you're seeing an offset voltage from the output amps.

Hope this helps and good luck!

\$\endgroup\$
1
  • \$\begingroup\$ That's a good point you're bringing up; your heuristic argument seems to pretty much rule out the use of capacitors at least in smartphones. Interestingly, I see this DC bias on my PC line out, my MacBook audio jack, and my iPhone. \$\endgroup\$ Commented Apr 9, 2017 at 22:37

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.