1
$\begingroup$

I dug my old Behringer UCA202 out of my shed and plugged it into a computer. I did some tests and it appears to be working fine... apart from some noise signals on the line input channels. I'm not sure where this type of noise could be coming from; weather it's the computer, the powered USB lead, or the device itself. What components could cause this type of interference?

Spectrum Analysis and Statistics

enter image description here

8 Second Spectrogram 750Hz - 5kHz (showing ~3.3Hz Period)

enter image description here

1 Second 0-10Hz Low-Freq Noise Signal(related? i.e above ~3.3Hz cycle)

enter image description here

I have unplugged everything else from the computer (except power lead and VGA monitor).

I have tried plugging the device into a powered hub. The DC input on the hub states 5v 2.5A, but I'm using a 5v 2.0A power source. I don't know if this matters.

This is the recording:


Only the monitor, mains lead and UCA202 are plugged into computer, and I unplugged the monitor from the mains while recording. There is nothing plugged into the UCA202.

$\endgroup$
  • $\begingroup$ Can I please ask you to upload a WAV of this segment somewhere? Such high frequency noise is usually coming from some sort of switching power supply. Is this recorded with something on the input? (e.g. an instrument). Can you please record one with nothing on the inputs? Also, you might want to set the input to "dynamic mic" type and then plug in a short circuited jack. This will force the input to zero and it should record silence. If it doesn't then it is noise creeping in from elsewhere. Please make sure that any LED (on common power socket) or fluorescent lighting is off. $\endgroup$ – A_A Aug 16 '16 at 8:15
  • $\begingroup$ @A_A I updated post. Thanks for the reply. There are no options on the UCA202 for dynamic mic. It's a line input. Shoult I still loop the in/outs? There is a light on the extension lead with which the PC is powered. think it's a mains rated light. No flourescent lighting. $\endgroup$ – n00dles Aug 27 '16 at 19:58
1
$\begingroup$

"The UCA202 uses the ubiquitous TI PCM2902 integrated USB DAC chip"

So this a "Full speed" USB device, and the USB data is being sent at 12 MHz, but it doesn't send data continuously; it sends it in short bursts every 1 ms, which means it has a frame rate of 1 kHz. This 1 kHz noise finds its way into everything and it's really annoying when you're designing USB audio products. 😒

Sometimes it becomes much more audible because of a ground loop and you can improve it by lifting a ground somewhere, like unplugging your laptop from the wall, or connecting to a balanced input on your monitor instead of an unbalanced one, or using an audio transformer to isolate one thing from another.

There is nothing plugged into the UCA202.

What if you short out the inputs instead of leaving them floating? They're RCA inputs so they might not have normals that short themselves when not connected to anything.

$\endgroup$
  • $\begingroup$ Thanks for the ans. I take it you meant 12 Mbps (in 1kHz packets). What I don't understand is how this interference could find it's way into a digital 1/0 signal. I would understand an analog signal being affected, but not digital. Where would this creep in? P.S. the ground loop doesn't apply in this situation. $\endgroup$ – n00dles Aug 28 '16 at 18:43
  • $\begingroup$ @n00dles I mean it's getting into the analog signal before the ADC $\endgroup$ – endolith Aug 28 '16 at 18:56
  • $\begingroup$ yeah,.. ok. So it is added in the device, before conversion and transmission. Ok. So there is pretty much FA I can do about it? I have my solder iron ready. I was going to put a shorter lead on it. $\endgroup$ – n00dles Aug 28 '16 at 19:04
  • $\begingroup$ What about powering it through some other means? Any Ideas? $\endgroup$ – n00dles Aug 28 '16 at 19:15
  • $\begingroup$ @n00dles You could try all kinds of things, sure. Powering it from a battery + linear regulator, filtering the power supply, disconnecting the input grounds and connecting them to different points on the ground plane until you find a clean one, putting capacitors on the inputs of things to filter out the 12 MHz before it gets demodulated, etc. It depends how it's getting in. Put a shorter lead on what? $\endgroup$ – endolith Aug 31 '16 at 5:02

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.