Improved Sound Card Interface - Ideas and Suggestions?

Status
Not open for further replies.
I am running some open source software to use my soundcard as a signal generator and digital oscilloscope. The programs are siggen and xoscope. Every time I go above 4000Hz I get strange waveforms no matter what amp is being tested. I use voltage dividers to drop the voltage down accordingly and also use the soundcard level controls for adjustments.

Anybody else have this happen to them?

Here are some snapshots of 5k, 8k, and 10k sine wave input.
 

Attachments

  • 5k.jpg
    5k.jpg
    87.1 KB · Views: 635
  • 8k.jpg
    8k.jpg
    85.2 KB · Views: 619
  • 10k.jpg
    10k.jpg
    86.3 KB · Views: 642
Last edited:
Call me old fashioned, but I'm going with the incandescent lamp. I'll bet its quieter than the FET circuit. Cold resistance is around 1k.

Pete

I dug around and found 2 small incandescent bulbs from Radio Shack -- the cold resistance of each was a lot lower -- one 3.7R, the other 40R -- the noise and THD was negligibly different from the dual FET protection circuit. Appears your guess is correct, at least at these low current levels (the AP input impedance is a low-ish 100K)

If the bulb fails, it certainly is a cheap replacement.
 

Attachments

  • BulbLimiter.gif
    BulbLimiter.gif
    50.7 KB · Views: 572
I would like to see autorange, larger smd's, shorter traces, true differential and properly shielded. The shielding should also protect the more delicate parts of it self from the not so delicate. Maybe adjustable input impedance which of course shows in the autorange. Further again a driver so that it can be monitored and tweaked throug our PC's and interact with the settings of our favourite analyzers 😉 preferably through some industry standard interequipment protocol... Keep a few I2C chanels open to be able to steer different ADC/DACs from the driver...

The only items of my list I would be able to do something with myself is the shielding and maybe test the logic.

Regards
 
Last edited:
If we run the adc/dac in max volume the preferred atteunuation could be saved in memory for that adc/dac (as could probably alot of other settings). Switch to another one and save its (what do the Borg call it in Star Trek movies) OMG - lost the word...Not personalities as it is for an individual.

Regards
 
I guess I missed this thread when it was new but I thought I'd weigh in anyway.

I was developing a driver circuit with lots of swing for a Unity Coupled amp and wanted to be able to measure distortion on those big (200Vrms+) swings without loading the circuit so I decided to do a 1M version of Pete's board. It works great but noise is increased, as has been discussed. Not a problem for this type of measurement.

If I wanted low noise, I would just build a second interface with the low input impedance. So my suggestion would just be to have instructions for the user/builder to build whatever version they wanted to.
 
Hi, Any updates? Still really interested in this project...

I'm sorry that I found the relevant threads by Mr Millett too late... although I run all through them, also I found that the Mr Millett sells on eb*y a pcb and two panels. Are there any complete kits left?

Regards.
 
Status
Not open for further replies.