Polycom Vortex EF2241, Application Note

Page: 71 / 77

The default routing of the EFBus will be sufficient for our design. However, for clarity, the labels have been
changed to reflect the signals that are placed onto the bus and all unneeded cross-points have been muted.
Since the EF2201 uses the P Bus to send telephone audio to the EF2280 and the ID of the EF2201 is 00, the
audio signal PB0 needs to be routed to Input PM0 of the EF2280. Remember that the signals that a unit
places onto the EFBus is put into a placeholder where that placeholder is identified by the device ID of the
unit that put the audio signal into this placeholder. Here, the EF2201 placed audio onto the P Bus and
since the unit has a device ID of 00, the placeholder is identified as PB1 where the "B" stands for bus. The
EF2280 takes that signal from the bus and internally assigns that signal to one of two mixes for the P Bus:
PM0 or PM1, where the "M" stands for mix.
M
ATRIX
M
IXER
The matrix mixer will need to be changed from the default settings in order to use Output A for the
amplifier. Only Inputs B, C, D, and PM0 need to be assigned to Output A. By default, Inputs 1-8 are routed
to Output's B and W. Note that since Inputs 1-8 are microphones, the cross-points are colored blue to
indicate that they are gated to Outputs B and W.
Inputs C and D are attenuated by 3 dB to Outputs A, B, and R1 because if the left and right channels are
mono, both channels will increase 6 dB in gain (Mathematically, two signals of 1V peak-to-peak amplitude
are added together, the result is a signal of 2V peak-to-peak. In terms of decibels, that can be expressed as
20*log (2V / 1V) which equals 6 dB). Inputs B and C are reduced by 10 dB to Output Y because if a program
source is mixed at 0 dB with local speech, it may be difficult for the remote person on the telephone to
distinguish between the program audio and the local speech because of the frequency limitations of the
telephone network. By attenuating the program audio, you "bias" the audio sent to the telephone in favor
of the local speech.
Input B and Output B are set to +10 dB and -10 dB respectively. The reason is that the nominal input and
output of the codec is .3 V rms. The nominal input and output of the Vortex device is .775 V rms. This
equates to a difference of 8.24 dB between the two nominal levels. For simplicity, you may round that
value to 10 dB. It is important to set the gain structure correctly so that you don't clip the input audio stage
of the codec and that the Vortex device gets a good signal to perform processing.
The outputs W and Y are signals that will be sent over the EFBus to the EF2201. See the section entitled
Conference Composer Layout EF2201 to find out how to take those signals from the bus and send the to the
telephone line.
The output R1 is our Acoustic Echo Canceller (AEC) reference signal. This signal is what the internal AEC
uses to remove from the local microphones. In our example, we will remove the codec, telephone audio,
and program audio from the local microphones.
71