note # 18
Phase check in a Louspeaker system
In Sample Champion version 2.8 and above, a "Correlation
Meter" and a "Stereo Check" functions are
available for measuring very easily the phase between
the 2 channels.
When an audio system has a large number of loudspeakers, the
possibility of wrong connections is always high. A friend of
mine got a new car, but the loudspeaker system provided by the
manufacturer sounded a little strange and he asked me to investigate.
Here are the results...
the test I used a 250 Hz sine signal, played from the car CD
player. Specifically I used the CD Audio Test & Demonstration
CD by Stephen Court & Alan Parson (track 48, infinite
repeat mode), but any other is suitable, also a self-made recorded
CD with some sine signals.
IMPORTANT: For phase tests I recommend to use a signal with
a low frequency (long wavelength) so that small displacement
errors in the microphone positions will not be interpreted as
250-500 Hz could be a good value. Higher frequencies require
more attention in the microphone positioning.
system in the car analyzed here has 6 loudspeakers: 4 in front
(1 woofer and 1 large-band tweeter in every channel) and 2 in
rear (1 woofer in every channel).
Champion settings for this measurement:
2 channels input
- 4K samples buffer
- Repeat mode (4096 cycles)
- Scope window opened, both channels shown (Red and Green)
I connected a microhone to each of the 2 input channels of the
soundcard, selected one of the 6 loudspeakers as reference (the
left woofer in front) and placed both microphones very near
to this loudspeaker.
the measurement, following waves are shown:
Scope plot is:
measured waves (left channel in green, right channel in red)
are not pefectly overlayed because the microphones are in different
positions (although very near). Using a low frequency test tone
this difference should be small.
I moved one of the 2 microphones near to another loudspeaker
(left front tweeter), leaving the other near the reference one.
measurement shown the following:
a small wave displacement can be observed, but this can be due
to physical factors such as a crossover or a different distance
of the loudspeaker from the microphones (behind the grid). So
the phase between these 2 loudspeakers is correct.
in the right channel front loudspeakers; right front tweeter:
Right front woofer:
all 4 front loudspeakers are correctly wired...
Let's go to rear loudspeakers.
BAD BAD!!! Phase inverted! Here
is the cause of the bad sounding audio system! The waves
(red and green in the plot) are clearly displaced.
the mesurements, we go now to the right rear loudspeaker:
the solution is correcting the phase of the left rear loudspeaker.
Trick: when playing sine tones, it could be useful to
change the fader and balance settings, when possible, in order
to minimize the effect of the sound from all other loudspeakers.
trick: when the second microphone is placed in front of
another loudspeaker, it could be necessary to adjust the input
gain on the audio card mixer or on an external microphone pre-amplifier
(if present), and turn ON and OFF the Y auto-scale button of
the Scope window, in order to have on the plot the 2 waves with
the same amplitude.
We would like to thank also a little guy that helped us and
gave some useful suggestions during the measurements:
Back to Application Notes