0
\$\begingroup\$

Im trying to record output from an old synth chip (YM3812). The output waveform is centered around 2.5V with peaks at approx. 2.5 +- 1.5V. I can easily connect a pc speaker with a TSSR cable to hear the sound. I connected 2.5V to GND and the audio signal to either the left, the right or both audio channels.

My problem: If I try to record the signal with a PC line input, the signal is clamped at 2.5V. Every part of the waveform that is "negative" relative to GND is removed.

Do you know, why my speaker could handle the negative voltages, while the line input just clipped? Should I connect 0V to GND instead of 2.5V?

Edit: This is the relevant part of my circuit:

schematic

simulate this circuit – Schematic created using CircuitLab

\$\endgroup\$
1
  • 1
    \$\begingroup\$ It is unlikely that your question can be answered until you add a circuit diagram showing how you are connecting this. \$\endgroup\$ Commented Aug 21, 2016 at 14:34

1 Answer 1

2
\$\begingroup\$

Should I connect 0V to GND instead of 2.5V?

Yes, of course. Otherwise you risk a dead short from your 2.5 V source to 0 V if there is a ground return path anywhere in your system.

schematic

simulate this circuit – Schematic created using CircuitLab

Figure 1. The audio signal can be decoupled from the DC with the addition of a series capacitor. The capacitor blocks DC whilst allowing audio to pass through. A 10 µF capacitor should do the trick. Watch polarity.

\$\endgroup\$
4
  • \$\begingroup\$ Thanks, the capacitor is a good idea. I think that there cant be a dead short in my circuit because the audio ground is not connected to any other part. I edited my post to show the circuit. \$\endgroup\$
    – tly
    Commented Aug 21, 2016 at 16:12
  • \$\begingroup\$ I your circuit it's not connected, but that ground goes somewhere ... \$\endgroup\$ Commented Aug 21, 2016 at 16:18
  • \$\begingroup\$ shouldnt the line input of the pc use a differential opamp to subtract these signals? \$\endgroup\$
    – tly
    Commented Aug 21, 2016 at 18:55
  • \$\begingroup\$ No, I don't think so. Audio signals are AC centred about ground / 0 V. The PC line input probably already has the DC blocking capacitor internally. I'm recommending one externally just in case. \$\endgroup\$
    – Transistor
    Commented Aug 21, 2016 at 19:13

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.