Theory Of Operation - Architectural Acoustics AUTOMIX CONTROL 8 User Manual

ENGLISH
AUTOMIX™ CONTROL 8
8 Channel Automatic Mixer
Thank you for purchasing the Automix Control 8! The Architectural Acoustics Automix Control 8 is a
Introduction
high quality automatic mixer with eight transformer balanced mic/line inputs. Each channel provides
a gain control, 48 volt phantom power (mic inputs), low cut filter, activity/clipping LED, an Aux send
control and a choice between manual or automatic operation. Each channel also provides a defeatable
insert point, 5 volt TTL status output and can be muted individually or multiple channels can be muted
simultaneously via an assignable mutebus. In addition, channels one and two provide an adjustable
priority control.
The master section provides a gain trim control, three 1/9 octave sweepable notch filters, a downward
expander, transformer balanced outputs (main and aux) and remote volume connections. The Automix
Control 8 has been designed to easily link multiple units together to form a single mixer with many
more inputs (16, 24, 32...). The Automix Control 8 is supplied with a see-through plexiglass security
panel to prevent changes to the installer's settings.
Theory of
The Peavey Automix Control 8 is a classic "automatic mixer" combining several time proven techniques
Operation to deliver maximum gain before feedback in an easy to setup and easy to use configuration. The
Automix uses VCAs (voltage controlled amplifiers) and gain computing circuits in each channel to ride
the gain downward as more microphones become active. By dropping the gain 3 dB every time the
number of active microphones double, the overall system gain remains at unity for ideal gain before
feedback performance.
A precision rectifier and logarithmic converter in each channel computes the amplitude of the audio
signal present at each microphone in real time. This channel amplitude, in decibels, is then compared
to the amplitude of the sum of all the channels, also in decibels. A simple computing circuit calculates
the mathematical difference between these two amplitudes and feeds the product to the VCA as a
gain control signal.
To better grasp how this works, let me offer a few examples. If we take the case of one person
speaking into a microphone, the amplitude in that channel will be virtually identical to the "sum of all
channels" amplitude. The difference between these two equal amplitudes is "0". When 0 dB is fed to
the VCA control port, the result is unity gain for that channel. All of the other channels, with no one
speaking, will exhibit significantly lower amplitudes. When compared to the "sum," which contains
a nominal signal, negative numbers result. These negative numbers, presented to the VCAs of non
active channels, command further attenuation of the noise and leakage present at these non active
channels.
A second more interesting example is two people speaking simultaneously. To simplify this example,
let's assume they're both speaking with the same loudness. As these two sources are incoherent,
i.e. not identical, they will sum as the square root of the sum of the squares or 3 dB more than
either individually. When each channel's amplitude is compared to this +3 dB sum, it's VCA will be
commanded to attenuate 3 dB. As before, the non active channels will be further attenuated. The
elegance of this approach is really apparent in this next case. Assume a person is stationed exactly
equidistant from two microphones. The identical signal arriving at both channels, being coherent,
will sum linearly to +6 dB in the sum of all channels. This results in the two channels being turned
down 6 dB. Being coherent, these two identical signals summed at -6 dB will add up to unity. These
gain relationships hold true no matter how many microphones are active and whether or not signals
are coherent. Since all computations are done in the Log domain, it doesn't matter what the actual
amplitudes are, only the relative amplitudes between signals.
7