[odigg]

I like electronics DIY, but it really is a small part of my life and I cannot justify the expense or the physical space requirement for an oscilloscope, signal generator, etc. So I've been thinking about software oscilloscopes like Virtins and Zelscope

Nowadays purchasing a good sound card, or even an external pro-audio box is fairly inexpensive. Many devices everything can do 24bit/192khz without a fuss, so analyzing frequencies well above the audible range is possible without spending much money.

Does anybody have experience with these software oscilloscopes? Does anybody have any comments on them?

 

[tangent]

There are several things dedicated hardware instruments can do that a cheap sound card can't, things which are valuable to audio DIYers:

1. Generate and measure fast signals. You may think that only audio frequency is interesting, so a sound card's AF bandwidth is sufficient, but that ignores the fact that most audio designs are really broadband designs, and need to be tested as such. One common test for AF analog gear is a square wave test. Pure square waves literally require infinite bandwidth, with practical square waves requiring a minimum of about 10x the bandwidth of the fundamental. So, to test an amp's square wave response at 1 MHz, you need at least a 10 MHz signal generator and scope. I've gotten valuable results while testing at up to 40 MHz before. And we haven't even talked about digital design yet, where 1 MHz is "slow"!
2. Create and measure voltages beyond audio line level. Sound cards are only good up to about 2 V. You can start to play games with custom attenuation, amplification, and cabling schemes, but this eats into the cost advantage you're trying to gain.
3. Tolerate much higher voltages than the instrument can actually measure, rather than just blowing up like a sound card will. Most proper test gear is good up to 400 or 1000 V.
4. Long useful life. I've replaced several sound cards over the years, but never had to throw away a decent quality test instrument. (I'll admit, this is a somewhat circular argument; I have thrown or given away crappy test gear. The decent stuff is the stuff I kept. Lesson: don't buy cheap test gear.)
5. Dedicated knobs and buttons are a far superior way of interacting with a test instrument than a mouse and keyboard.

I won't even give you the cost advantage. You can easily end up paying as much for an AF computer-based test rig as for a low-end scope. First you decide the probing setup is weak, so you build custom cables, and maybe a switch box. Then you run into the voltage problems and start looking at attenuator/gain boxes. Then you find all the weaknesses in the sub-$100 sound cards and start lusting after pro audio interfaces. You can easily find yourself with $1000 in computer-based test gear, and still only have AF capabilities.

There is one clear advantage to a sound card, which is indisputable: 16 or 24-bit resolution. This means cleaner generated signals and lower measurement floor than you can achieve with any broadband hardware instrument of reasonable cost. This only helps at AF, but that's still useful.

My advice: if you're going to get deep into DIY electronics, plan to get both systems. You want some kind of computer-based system for AF, and dedicated hardware for everything else.

 

[vixr]

Thank you for the awesome review and insight, Sensei Tangent...I use a picoscope at work and its like watching a cartoon of a real scope being used. I dont care much for it. It works good enough, I suppose.

 

[diditmyself]

Software oscilloscopes are absolutely worthless because of the soundcards limited sample rate. Sure you can analyze a 1 kHz square, but why?

Soundcards are useful for distortion, frequency and crosstalk analysis. The software is free - RMAA.

If you want to study stability, speed and compensation you need a real hardware based oscilloscope. You can get an oscilloscope cheap nowadays. I paid about 75 USD for a Tektronix 100 MHz digital storage oscilloscope in excellent condition.