|
[DPRG] Audio filtering
Subject: [DPRG] Audio filtering
From: John Swindle
swindle at compuserve.com
Date: Wed Nov 4 18:44:06 CST 2009
Scott and Ed:
I agree that most audio problems are solved by eliminating ground
loops and isolating power. It seems that some consumer devices
have internal leakage pathways that are sufficient to create
loops within the signal pathways, even if the grounds are
isolated, maybe because of an unbalanced signal going over what
looks like a balanced line. For low-power devices, the signal
lines can source or sink power unintentionally.
A few times, I've transfomer-coupled the systems, with pleasing
results. Fortunately, audio transformers are small. I personally
have never had satisfactory success with filters if I want to
actually listen to the audio. Filtering seems OK to reduce the
noise when the audio is being measured by a program, such as for
speech recognition. But if I want to listen to the audio, a
filter is unsatisfactory.
Beware that the loops can be anywhere. For example, the USB spec
allows the signal ground to be connected to the chassis ground at
both ends of the cable. (The original USB spec prohibited that
behavior, but later specs relented to a "best-practices"
statement instead of a specification.)
Older stereo systems connected the signal lines through
high-value resistors to the safety ground (which is one side of
the AC mains). I don't know if that practice is still done, but
those megohm resistors sometimes created problems with low-level
audio signals.
Audio amplifiers can be so sensitive that even a capacitive loop
can create problems. It doesn't have to be a DC pathway. If
you're lucky, you can measure the AC current between the isolated
systems. Kinda spooky. Sometimes, you can feel the current. Even
more spooky.
Good luck.
John Swindle
More information about the DPRG mailing list
|
