834 Chapter 25
sat there with all their inputs and outputs accessible by
way of a jackfield for the prosperous or by small screw-
driver and sore knees for those who weren’t.
Mixing sources was accomplished by directly paral-
leling amplifier outputs (possible because all the old
tube gear was designed with a particular termination
impedance in mind, usually arranged to be a conven-
tional balanced 600:) and either hoping or arranging
that the destination had enough gain to make up accrued
paralleling losses. Crude as that may seem today from
an engineering viewpoint, it has a sheen of pure
elegance. An amplifier was just that, a box that had a
balanced 600: source and termination impedance. It
might also have an alternative bridging (> 10 k:) input
terminal and a selectable amount of gain offering
universal application from microphone amplifiers
through mixing amplifiers to headphone amplifiers. To
do more things, more boxes were added. Equalizers and
limiters, a treasured few if there were any, were simi-
larly universally applicable. Variable-level control was
again attained by true balanced 600: source and termi-
nation, via studded rotary attenuators. The utter beauty
of the systemless studio was that anything could go to
anywhere via anything else and be mixed or distributed
at any point on the way.
Soon enough amplifiers were hardwired to attenua-
tors and designated specifically a microphone ampli-
fier, and a system had been created. Some of these
together with a mixing gain makeup amplifier were
thrown in a box. The mixer was born.
It has been downhill ever since, with ever-increasing
numbers of system elements being tied together in
increasingly knotted manners in order to maintain some
kind of flexibility. Perversely, a system can be defined
as a means of reducing the ultimate versatility of its
constituent parts.
Once a mixer was accepted as a system element
itself, the problem set in further. There was no need to
provide for convenient connection of its internal inter-
connections to the outside world, so the balancing trans-
formers disappeared, and more economic alternatives to
the stud attenuators operating at more convenient
internal impedances evolved. By a more positive token,
the electronics were gradually becoming optimized for
the specific functions to which they were designated,
such as the microphone amplifier and the mixing ampli-
fier. (The question nags us whether a universal ampli-
fier, by now all but obsolete, could be optimized for all
the varying requirements, this is unlikely.) Still, at least
all the inputs and outputs of the mixer were conven-
tional. This held true until the slow demise of vacuum
tubes in professional audio.
25.6.1 TransistorsTransistors were justifiably unpopular for a long time
because of the numerous limitations they placed on
design. The headroom was severely limited because of
the low supply voltages that could be applied to the
early devices. They were noisy. The lower operating
impedances and differing modes to tubes took some
getting used to and, when they clipped, they actually
clipped rather than gracefully bending (characteristic of
tubes that people had known, loved, and frequently
taken advantage of even now). To realize a reasonably
low stage distortion, many transistors in compound
configurations using heavy amounts of negative feed-
back were used—a far cry from a single tube stage
operating virtually wide open with little feedback. This
gave rise to a peculiar phenomenon that sounded as if it
hailed from science fiction—zero impedance.
The heavy negative voltage feedback employed
around transistor circuits could be made to render the
output of an amplifier insensitive to varying load
impedances; they would deliver the same output voltage
level almost regardless of their termination impedance.
This eliminated termination problems with the attendant
worry of compensating in level for differing load
hookups. With the exception of long line feeds, 600:
terminations were as good as dead. High-level balanced
inputs were now almost exclusively bridging; they had a
sufficiently high impedance (usually >10 k:) not to
disturb the level of the source to which they were tacked
on. For better or worse, it has become the conventional
studio interconnection technology. It has taken until
fairly recently for a distinction and separate level speci-
fication for the two technologies to be accepted.25.6.2 Level SpecificationsThe original transmission line level specification
referred to a power level of 1 mW regardless of imped-
ance. This was 0 dBm. It was a universal specification
applicable to any signal of any frequency being trans-
mitted along any length of wire for any purpose at any
rated impedance, and it is used extensively in
radio-frequency work and other things entirely unre-
lated to audio. The dBm definition is sacred and can’t
be changed. Zero dBm in a 600: load works out to
0.775 Vrms; this was adopted de facto as the reference
for use in general audio work. With zero impedance
technology, although the working voltage is specified,
the impedance isn’t. It can be anything, but the power
(as measured in dBm) necessarily varies as a result; for
instance, 0.775 Vrms across a 100: load is +7.78 dBm