1
\$\begingroup\$

I just discovered an app for mac that uses the Audio Input of the computer and the AD-converter in the sound card as an Oscilloscope.

First, I was thinking this can't work... but theoretically, it's a great idea! I know it can only measure for frequencies starting from ~20 kHz, but if you need to measure MHz, then it's just fine.

my question is: How reliable is this technique, independent form the Software? If you build a nice adaptor with overvoltage protection (max 2V), then I suppose there is no danger for the sound card on the computer, right? Any other experiences?

Thanks

\$\endgroup\$
  • \$\begingroup\$ Assuming the audio ins aren't high-pass filtered, shouldn't be a problem (but I don't know what the max voltage is on these ins). If you're in it to learn, or if you have severe budget issues, not a bad approach. You can get a USB scope, with software, for $150, or a real DSO for $350, so if issues get in your way move up a step. \$\endgroup\$ – Scott Seidman Mar 13 '13 at 15:37
  • 2
    \$\begingroup\$ Are there any audio interfaces that aren't high-pass filtered? Blocking DC is standard practice in audio circuits, and as far as I'm aware even the cheapest of cheap chineese audio interfaces don't cut that corner. \$\endgroup\$ – Phil Frost Mar 13 '13 at 15:59
  • \$\begingroup\$ DC is cut, I think this is true for all audio devices. I'll just give it a try. There software has a trial version, so it's easy to test. Thanks. \$\endgroup\$ – otmezger Mar 13 '13 at 16:13
  • \$\begingroup\$ Just be carefull with input voltage range. I'm not sure, but I'd expect it to be limited to +/- 1Vpp. You don't want to damage your Mac soundcard to save few 100$ on a simple DSO. \$\endgroup\$ – jippie Mar 13 '13 at 16:34
  • \$\begingroup\$ I think you mean "I know it can only measure frequencies UP TO ~ 20 kHz, but if you DON'T need to measure MHz, the it's just fine" Other than that, see David Kessner's answer. If you meant what you said, see TheNoonMoose's answer. \$\endgroup\$ – Bjorn Roche Mar 13 '13 at 18:46
9
\$\begingroup\$

Generally speaking, the audio inputs of a PC are terrible. Lots of noise, terrible frequency response, etc.

For example, the frequency response of a PC audio input is typically 20Hz to 20 KHz, but could be much worse. Some audio inputs are only really designed for voice, which has a frequency band of about 300 Hz to 4 KHz.

But none of the PC audio inputs will be good down to DC (0 Hz). This makes the usefulness of this as an o-scope rather limited.

This is useful if you are looking at analog signals that are within the 20Hz to 20 KHz range. It can be very useful, actually, for this. But when looking at digital signals, or signals that have components outside of this range, then the usefulness quickly vanishes.

Here's the thing: You want your test gear to be trustworthy, and using a PC audio input is anything but trustworthy. Yes, there are some situations that you can probably make it work-- but it takes extra work to know if what you are seeing is valid or an artifact of the "scope". Your mileage may vary. For many things, it just isn't worth the effort to use the PC audio inputs in that way.

\$\endgroup\$
8
\$\begingroup\$

I think you have your understanding of the frequencies backwards. If it's a standard audio input, it's probably being sampled at 44.1KHz, which means you can only measure frequencies up to approximately 20KHz. If you need to measure MHz, this technique should be useless.

Other than that, the technique should work as you expect it to with correct overvoltage protection.

\$\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.