Having a Breakdown
On Thu, 2 Dec 2010 03:35:12 -0800, Kele wrote
(in article ):
PS: Audio Empire, I re-read your response and don't fully understand
DC offset. Does that mean there is some stray AC on my signal path.
It seems you are saying that my preamp's power supply is overtaxed
because of stray voltage. It's power supply isn't necessarily the
culprit then. Maybe the preamp isn't actually faulty? And I don't
understand DC coupled unless that means the stray AC is flowing to the
amp or maybe from the amp... hmmm. May be I should try swapping out
the APC (not PS-Audio) power strip(s)? I'll put in my Irrational But
Efficious CD to remove the stray electrons from my system (ha-ha).
But really, is this an AC issue? I just checked the AC outlets and
they appear "ok" using a polarity & ground tester (the plug-in type
with the lights). I checked some outlets on the power strips too.
What's left... RC patch cables? Anything can do, I'll try.
It's been over an hour with preamp and amp isolated from one another;
the amp wasn't any cooler. Something elso too... The amps distortion
& thermal protection lights don't illum even with the Dolby on &
speaker cone extension previously described.
I'm saying that there is a DC voltage in the signal path. This shouldn't
happen. There are two ways to get an AC signal (such as audio) though an
amplifying device. One is to AC couple the signal (where there are capacitors
in the signal path which will allow the AC (the audio) through, but none of
the DC which is necessary to operate the active circuitry). The second way
is to DC couple the signal through. In DC coupling there are no capacitors in
the signal path (desirable where practical) and the DC is kept out of the
signal by designing the circuit so that the plus and minus rails on the DC
power supply are the same (example: If you put +15 volts on one side of a
resistor and -15 volts on the other, and tap the center of that resistor, the
voltage at that tap will be zero because (-15) + (+15) = 0. The same is true
if you put equal + and - voltages across a transistor in an amp. When the
transistor is conducting (I.E. "on"), all else being equal, the DC voltage at
the output of that transistor should be zero and the only thing there will be
the AC signal (the audio). Now it's a bit more complex than that, but this
will do for the purpose of this discussion. Now. Let's say, that something
has gone wrong with the power supply, and instead of there being + and - 15
volts across that transistor, there is now +20 and -15 volts. That means that
there will be 5 volts of DC along with the AC (audio) signal at the output of
that transistor ((-15) + (+20)= +5). That +5 volts is called the "DC
offset".
If your power amp has a sensitivity of 2 volts AC to give full output and if
the power amp is DC coupled, then that offset will be enough to drive that
power amp into clipping (exceeding it's possible output) because it is trying
to amplify that DC voltage (which shouldn't be there ). That means that
one-half of your amplifier's push-pull output transistors are turned hard-on
continuously and are conducting at their full capacity to amplify that +5
volts (the amp doesn't know the difference between audio and a DC voltage
that shouldn't be there). The problem here is that amplifier output
transistors weren't designed to be full-on all the time. They are supposed to
switch to the negative going output transistors every half cycle. To keep
them on all the time, is "exceeding their duty cycle" IOW, operating them
outside of their worse-case design parameters. Some amp designs can handle
this kind of abuse with no damage and others will simply fail if the fault is
ignored long enough.
In your case, your speaker cone moves out proportionally to your volume
control setting. That indicates that the DC is getting into your audio signal
BEFORE the volume control and that the volume control is altering the
amplitude of that offset DC voltage. You're lucky here because that means the
entire DC offset is not being applied to your power amp (keeping it from
failing) and also keeping it from pumping so much DC current into your
speaker system that it destroys the speaker(s) as well. That kind of DC
signal exceeds the speaker's design duty cycle as well. You could burn-up the
voice coil by pumping DC through the speaker.
|