USB Sound Card Hack
USB Sound Card Hack
USB Sound Card Hack
Let’s begin with the above photo of a cheap electronics gizmo that made me curious. This
$1 product that I bought from Amazon is nothing but a USB sound card (integrated two-
channel USB audio controller) with audio connectors. Because the tiny device uses a USB
port for power, an external power supply adapter is not required. Above all, it’s a true plug-
and-play (PnP) device for Windows PCs.
You can obviously connect your hardware to your PC through the microphone input of the
USB sound card. However, a closer look at the electronics of my USB sound card revealed
that the “monophonic” microphone input is AC-coupled to the chip’s input by a 1-µF
capacitor, and the microphone input also has a bias feed (for electret microphone) that is
about 2.8 V (measured in an open state).
Simply bumping off the 1K2-Ω resistor (R7) from the PCB will remove the microphone bias,
though it’s not important in some situations. For DC coupling, you can bridge the 1-µF
capacitor (C3) with a piece of wire (0-Ω jumper). As far as I know, there’s no digital high
pass for the microphone input (Pin 27) inside, so CM108 (and CM119) can accept DC input.
Nevertheless, a few more modifications may need to be carried out aside from the above
alterations to make the USB sound card fully useful as an adapter for a PC sound card
oscilloscope. For more convenience, it’d be better to take off the microphone socket from
the PCB because then you can easily extend a three-core screened cable (5 V-GND-MIC)
from the PCB to the outer world through the space leftover.
Afterwards, you can start the evaluation with the help of Audacity (or similar software).
The evaluation can be pretty easily done when you know what you’re looking for. See the
partial Audacity screenshot captured while I was only babbling — the microphone pays
great attention!
*Why 44.1 KHz that only bats can enjoy? As you’re aware, the range of human hearing is
20 Hz to 20 KHz. In digital recording, the Nyquist theorem states that a recording rate
should be double the highest frequency recorded. Therefore, 44.1 KHz should be perfect
for the human hearing range.
Besides, as we are looking at amplitudes over time, the bit depth simply denotes the
resolution or number of points available to store the amplitude data. For example, 16 bits
results in 65,534 points, and 32 bits worth of data gives us 4,294,967,294 data points. Based
on many experiences, nothing beyond 16 bits and 44.1 KHz is required as the choice always
provides better audio quality that we’re able to experience.
Note that the frequency range of a PC sound card oscilloscope depends heavily on the PC
sound card, but 20-Hz to 20-KHz range is possible with most sound cards. In principle, the
low-frequency end is usually restricted by the AC coupling of the input signal. So the USB
sound card, if “hacked” (for DC coupling), will hand a bandwidth of DC (0 Hz) to 15 KHz.
The below image is a mix of random scope shots taken just before the hardware hack of
the USB sound card (that’s to say with AC-coupled input).
Front-end homework
Notably, the microphone input (MICIN) of the USB audio controller chip of the USB sound
card is surely an analog input (AI) with a recommended input voltage swing in the 0-V to
2.88-Vp-p range. The frequency response is 20 Hz to 17.6 KHz (44.1 KHz). So what we need
is a bit of extra hardware that would scale the oscilloscope’s input, preferably with some
form of DC bias adaptation, and input attenuation. Read this article to learn more about
some front-end circuit ideas.
Finally…
If this article has inspired you without your budget running out of steam, that’s great. This
is just a prefatory guide, so it’s worth experimenting with more powerful and beautiful
projects. However, as just about any engineer will tell you, you should aim to keep it as
meticulous as you can. Happy hacking!
CM108 Datasheet
CM119 Datasheet
Audacity Page
Nyquist Sampling Theorem