Ever since I purchased the Xenyx302 back in August 2015, I had this really strange issue where the left audio channel would suddenly be overlaid with some kind of continuous distortion noise.
What made the issue so strange was how randomly it presented itself, sometimes I could go weeks without having the issue but sometimes I couldn't even go a day without having to deal with it (switching the card off and back on again).
Here's a short capture of the noise while I was watching EEVblog:
https://soundcloud.com/user-425960372/pcm2902-intermittent-distortion-issue/s-OHjUSFor the longest time I thought it was either a Pulseaudio or ALSA bug, until one of the Pulseaudio developers pointed out to me the errata for the PCM2902:
http://www.ti.com/lit/pdf/slaz036http://www.ti.com/lit/pdf/slel057Based on the USB Product ID, I was able to determine that the chip inside my Xenyx302 was indeed this badly behaving PCM2902E revision.
Initially I had disregarded this errata because I thought a device sold in 2015 (date coded: 1412) wouldn't have a silicon bug that was fixed in 2008 according to the docs...
But I guess Behringer didn't bother to renew their stock or something..?
So I ordered a pin-compatible PCM2902C (
http://www.ti.com/product/pcm2902c/description) from TI Store to replace the old erroneous chip.
The rework job I did with a 2mm chisel tip soldering iron, a piece of ~0.3mm magnet wire and some Chemtronics flux coated solder wick.
I didn't use enough flux and the magnet wire was a tad too thick, so I ended up breaking one of the pads but that wasn't a big deal.
So if you intend to ever try this, remember to use a lot of flux! Low melting point solder would also work wonders I imagine.
It's been a couple of months now and the issue is finally completely gone!
I'll attach some pictures of the operation below.
Perhaps someone will find this useful some day
Now, of course the mandatory: "While I was in there, I also"-part:
- Added a linear taper 50K stereo pot in-series with the 2-Track input for volume control
- Replaced the 100uF headphone output DC blocking capacitors with 330uF ones (I got 16 ohm headphones..)
- Halved the headphone outputs 47-ohm series resistors (Lower values made the background hiss too loud for my headphones)
- Replaced the AD8602D with a supposedly better suited AD8397 (motivated by:
https://nwavguy.blogspot.fi/2011/03/behringer-uca202-frankendac.html)
The audiophoolery mods I did in the hopes of finding a way to reduce the background hiss (The Behringer-hiss
) at higher volume levels.
Reducing the output impedance via lower resistor values makes the output louder but also makes the hiss louder...
Changing the headphone op-amp didn't have any effect that I could perceive.
Changing the output caps did make a noticeable difference in the low end bass. (a slight improvement imho.)