894 Chapter 25
25.11.28 A Practical EQ
A three-section parametric EQ with additional versatile
shelving-type high- and low-frequency controls is
detailed in Fig. 25-71. It is designed to be easily short-
ened to high-frequency, low-frequency, plus a single
midband parametric section, for applications that don’t
demand the full complement of facilities. Each indi-
vidual section is switchable in or out to allow preset
controls. Simple in-and-out comparisons with tie-down
resistors maintain the dc conditions of the unused filters
to minimize switch clicks. Even a brief look at the
circuit reveals a major benefit. The signal path through
the EQ is merely via three op-amps, IC2 is an input
differential amplifier, and IC3 does duty as the output
line amp. In the shortened version this path is reduced to
only two op-amps, IC1 and IC3, which serve also as a
swinging-input EQ gain block. IC2 and its associated
circuitry are unused in this simplified version.
The unusual components around the differential
input stage provide unity differential in unbalanced out
levels while providing an identical impedance (with
respect to ground) on each of the two input legs. Natu-
rally, the more precise the component values, the better
the common-mode rejection is likely to be.
25.11.29 The First EQ Stage
IC2 in Fig. 25-71 is the first swinging-input stage. It has
two nonfrequency overlapping filters hanging off it, one
section covering 25 Hz to 500 Hz, the other covering
1 Hz to 20 kHz. Each filter network creates a complex
impedance form against frequency that looks like a
series LC-tuned circuit to ground. This fake-tuned
circuit (formed from two constant-amplitude phase-shift
networks in a loop, named the CAPS-variable filter)
reach parameters ordinary filters cannot reach.
The center frequency is continuously and smoothly
variable over its range using reverse-log potentiometers;
Q remaining consistent over the entire swing. The Q
itself is continuously variable between 0.75 and 5 (very
broad to fairly sharp, representing bandwidths of 1.5 to
0.2 octaves, respectively). Positive feedback inside the
loop, which defines the Q, is balanced against negative
feedback, which controls minimum filter impedance
and, correspondingly, amplitude. Interestingly enough,
this circuit relies on the input impedance of the
swinging-input stage as part of the negative feedback
attenuator. Fortunately, this impedance is reasonably
constant irrespective of boost-and-cut control
positioning.
In the absence of complementary square-law/reverse
square-law dual-gang potentiometers ideally required
for the purpose, readily available log/antilog dual-gang
pots, retarded a bit to a reasonable approximation,
control the positive/negative feedback balance. As a
result of this compromise, the filter crest amplitude
(maximum effect) varies within ± 1 dB as the Q control
is swept; in comparison to the dramatic sonic difference
from such a Q variation, this tends to insignificance.
The result of all this, at the output of IC2, is a pair of
resonant-type curves of continuously variable place,
height, depth, and width.25.11.30 Second EQ/Line AmpA reasonably hefty pair of transistors is hung on the end
of IC3 to provide a respectable line-drive capability, in
addition to the use of the amplifier as a swinging-input
EQ section. There is enough open-loop gain in the
combination of the op-amp and transistors (over a much
greater bandwidth than mere audio) to cope with 15 dB
of EQ boost and output-stage nonlinearities.
Differing from the last EQ stage, this one only has a
single midfrequency bell-curve creator, operating over a
range of 300 Hz to 3 kHz, together with high- and
low-frequency range impedance generators.25.11.31 Low-Frequency ControlGyrating inductance to create a conventional
low-frequency shelving response (variable in turnover
frequency by a 220 k: antilog pot) is achieved around
IC11. A fairly large (2.2μF) series capacitor forming a
resonance is switchable in and out. The value of the
capacitor is carefully calculated to work with the circuit
impedances to provide an extreme low-frequency
response that falls back to unity gain below the resultant
resonant frequency without compromising the higher
frequency edge of the curve. The Q of this arrangement
reduces proportionally to increasing frequency. Typical
resultant response curves, Fig. 25-72, show just what all
this means, demonstrating an extraordinarily useful
bottom-end control.25.11.32 High-Frequency ControlUnusual is one way to describe the high-frequency
impedance generator and its EQ effect. It is essentially a
supercapacitor, or capacitative capacitor. In other words,
it’s a circuit that, when in conjunction with a resistor,
causes a second-order response as would normally be