"DGD" said:
I have an ASUS A7N8X-E Deluxe mobo mounted in an ANTEC Sonata case.
The front panel audio connections are connected to the mobo front panel
audio connector usng a 10 wire connector (although not all connections
are utilized). For some reason, if I plug a microphone into the rear
microphone jack which is part of the mobo, it is far more sensitive to
sound than if it plugged into the front panel microphone jack, which is
connected to the front panel audio connection on the mobo. I would
like to simply connect a set of headphones with a fitted microphone
into the front audio jacks, but one can barely hear the voice through
that microphone connection. Any ideas as to what is happening and how
does one correct it? Thanks
DGD
I tried looking at the Realtek datasheet for the ALC650, but I
see some errors in the diagrams there. In any case, I can think
of two possible reasons:
1) There is a 20dB microphone boost button. That selector button,
somewhere in the mixer control panel, is useful for dealing with
mics that have low output (dynamic microphones). Perhaps the boost
button setting changes, when you change the microphone selection
between front and back mic ?
2) Electret microphones are a common form of microphone. They
receive a DC bias from the microphone jack, as a source of
power to run them. Now, if the bias was loaded down for some
reason, that might reduce the gain of the mic. That is why
the FP_PANEL header calls the signal MICPWR, even though the
amount of power involved is miniscule. The bias resistor on
the motherboard prevents this power from being dangerous -
the resistor limits the maximum current flow. Typically
a +5V source is used, with a 2.2K ohm series resistor to
limit the current flow.
I'm betting the problem is (1), as I don't see an easy way for
the problem to be (2).
There is some info here on electret microphones. (Note: This
site uses popup adverts...) The GIF file shows how a typical
monophonic electret is connected - the electret is a two
terminal device, while the cable uses a stereo 1/8" plug.
The electret in fact, shorts MIC to MICPWR, so the signals
become one in the same (at least for a mono mic - a stereo
mic like the Andrea Electronics one, actually needs separate
bias on both tip and ring contacts). The output of the
mic is capacitively coupled to the AC97 codec.
http://www.epanorama.net/circuits/microphone_powering.html
http://www.epanorama.net/circuits/sb_micinput.gif
Note the comment on the Epanorama page:
"The microphone capsules are not usually too picky of the actual
votlage, so generally anything in the range of 3V to 9V will work
(although the voltage can sometime affect the microphone output
voltage level somewhat)."
If you own a multimeter, you can do your own electrical analysis.
You won't need the micropnone to do this, but make sure the
volume is turned down on the computer speakers when you do the
test. (You do this test with the computer running, so that it
will be putting bias on the mic jacks. When the multimeter is
touched to the mic signals, there will be loud popping sounds,
so the computer speakers must be turned down.)
Do this test for both the rear mic_in jack, and your front
panel jack. Find an audio cable, with male 1/8" stereo plugs
on each end - these are typically used for TV tuner cards, to
connect from the TV tuner, to a sound card. Plugging one of
those cables into a computer jack, gives easy access to the
three signals "TIP RING SLEEVE".
1) With the multimeter set to the 10 volt scale, measure the
voltage from TIP to SLEEVE. Measure again from RING to
SLEEVE. RING to SLEEVE should measure 5V. If it measures
below about 3V, an electret plugged in there won't
give full volume.
2) Move the input leads on your multimeter, to prepare the
meter to measure current. Set the meter to the highest
range, then turn it down gradually until you start to
see a measurable current. I would expect, when the meter
is down to the 10 milliamps full scale range, you'll see
about 2.27mA of current. Dividing the voltage measured
in (1), by the current measured in (2), tells you the
size of the bias resistor Asus is using. If you got
5V and 2.27mA, dividing 5/0.00227 = 2.2K ohms. If a much
higher resistance value is the conclusion of your
own measurements, that too would be a reason for the
circuit to not work well.
You could also attempt to measure the DC voltage on RING,
while the microphone is connected to either of the
computer jacks. If the voltage is quite low (much less
than 2.5V), again, that implies the circuit is not
being biased properly - it could be the electret is
being "too greedy" for use with a typical computer sound
input.
When you consider the variation in electret mics (not all
use the same voltage), and the variation in motherboard
designs (not all use the same bias circuit), it is really
a miracle that electrets work at all.
Paul
