2
\$\begingroup\$

I have been experimenting with a Soundcard Oscilloscope and I am having some problems understanding what's going on! I have attached a screenshot of my digital scope setup. My hardware setup is simple. I have a male to male TRS jack running from my line out into my mic input. I have been experimenting with a simple sine wave.

Screenshot of Soundcard Oscilloscope Setup

When I measure the output voltage of my line out (Stereo plug - TRS) with a voltmeter I get a reading of 26mV for the right channel and 12mV for the left channel (Left and Right speakers not the left and right scope channels). However, the scope is reporting an effective voltage of 508mV rms for the left channel and the right channel (not in the screen shot) is saturated at 783mV rms.

Why do the 26mV and 12mV readings remain constant on the line out when I adjust the signal from my sine wave amplitude or change the volume and balance control of my PC soundcard? The voltmeter reading never changes. The voltage changes on the scope, but not on the voltmeter. What am I missing?

My second question has to do with the soundcard max voltage readings I am getting. When I boost the frequency or amplitude of the wave it saturates at 2 volts. I am guessing +2 and -2 is the max input/output of my soundcard? The research I have done would agree with this conclusion but I would appreciate some input! If this is the case why can't I get a 2 volt reading on the voltmeter instead of this micro volt jitter?

Thanks for your time!

\$\endgroup\$
  • 2
    \$\begingroup\$ Do you have your multimeter set to DC mode? A cheap multimeter will typically average the signal in DC mode, leaving you with the DC offset (your 12 and 26 mV readings). Try the AC setting. \$\endgroup\$ – mng Dec 18 '11 at 6:58
  • \$\begingroup\$ @mng Good point. Yes it was in DC Mode. Its a cheapo radioshack meter that only has a 200V and 300V AC setting. I get a reading of .1 on the 200 setting. Seems like it could be the problem. Maybe its time to invest in a grown up meter! \$\endgroup\$ – atomSmasher Dec 18 '11 at 7:09
  • 1
    \$\begingroup\$ On my computer I can output voltage from 25 mV to 1.99 V depending on the volume setting. I tried this with two different sound cards. Somewhere I also read that the maximum effective voltage for soundcard output is set to 2 V, but I can't find an authoritative source for that information. As for the voltage measured by the scope, I'm pretty sure that it needs to be calibrated. I can't find any mention of that in the program right now, but it has been there in previous versions and is mentioned in other similar programs. \$\endgroup\$ – AndrejaKo Dec 18 '11 at 10:03
  • 1
    \$\begingroup\$ A good source of information on powering devices from sound card - slightly tangential but a lot of good read :) eecs.umich.edu/~prabal/projects/hijack \$\endgroup\$ – qdot Dec 18 '11 at 10:47
1
\$\begingroup\$

Firstly, as mentioned the multimeter in DC mode will only give you the DC level of the signal, which if it's swinging around 0V will be 0V (or close enough) You would need a half decent meter with a low range AC mode to get a reasonable reading.

Secondly the sound card oscilloscope software will almost certainly need to be calibrated. All it receives from the card is a value ranging between 0 and its full scale, e.g. in a 16-bit card this would be from 0 to 65,536.
It does not know how these values translate into voltage without calibration (the software may have a default setting based on usual sound card ranges, e.g. +/- 2V or whatever, which may be what you are seeing now, and may or may not be so accurate)
For example if your sound card range is +/- 1V then 65,536 would equate to +1V. If the software is set to a default of +/- 2V for full range then it will see a value of 65,536 as 2V, when in face the actual signal level would be 1V.
The fact you have your line out feeding mic in may cause the default calibration and reported levels to be a fair bit off, as line level out is a fair bit higher than what a mic input will expect.

If the software is any good it should have a calibration setting which you can use to set things up correctly. This will probably involve feeding a signal of a known level into the card and telling the software what the level is, then it can work the rest out. Since most soundcards have a DC blocking capacitor you will need an AC signal for this.

| improve this answer | |
\$\endgroup\$
  • \$\begingroup\$ Excellent answer. Thank you. I noticed in the settings there is a "calibration of amplitude" text box that is set to 1 V/DIV by default. I will experiment with it some more. Thanks again for your help. \$\endgroup\$ – atomSmasher Dec 18 '11 at 16:40
  • \$\begingroup\$ To follow-up I have come to a final conclusion. After purchasing a multimeter that was able to read small amounts of voltage I was able to determine the voltage at the end of my line in wire. I also adjusted the frequency to 60hz. I discovered that the calibration was indeed slightly off by a factor of .23. When I changed the value to 1.23 instead of 1.00 my voltmeter read within a thousand of the reported Vrms on screen. I think its safe to say my input output voltage for the sound chip is +1/-1 volts. Now I can create my voltage divider and measure very very small signals! \$\endgroup\$ – atomSmasher Dec 18 '11 at 23:23
2
\$\begingroup\$

You should also consider that sound cards usually have a capacitor in the microphone input to cut off the DC.

| improve this answer | |
\$\endgroup\$
1
\$\begingroup\$

Even when you get a decent multimeter with an AC range it will probably only measure correctly sine waves with frequency below 1kHz or so (they were designed to measure 50/60Hz). With different frequencies and waveforms you will probably get a reading that is proportional to the amplitude of the signal but for absolute voltage readings you really should use an oscilloscope.

| improve this answer | |
\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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