Advertisement
[If you are
interested
in
an explanation
of the
mechanism causing
the
distortion increase in PROOF mode, refer to 5.b of Appendix B (Circuit
Description).
The
relevant
paragraph
begins: " PROOF
mode...".]
Stereo
Performance
Verification:
These
measurements
are
usually
made
most
meaningfully by defeating compression and limiting action while leaving all active
circuitry unbypassed. The existing PROOF/OPERATE and LOUDNESS CONTROLLER
OPERATE/DEFEAT
switches
provide
this
function. (See
Mono Proof
immediately
above.)
Special attention should be paid to headroom. The clipping point of the VCA's in
the Dual-Band
Compressor is
approximately 5dB
above 100% modulation,
although
distortion
will begin to
rise substantially
as clipping
is
approached. In
the BTSC
Stereo
setup instructions in Part 4 of this Manual,
instructions were provided for
adjusting
the
8182A's
OUTPUT
ATTENUATORs
such
that
100% L+R
modulation
(i.e., + 25kHz
deviation)
is
achieved
when
both
Left
and
Right
Channels
are
simultaneously driven. This means that only 50% L or R modulation is achievable
when
a single
channel
is
driven
alone
unless
the
OUTPUT
ATTENUATORs
are
temporarily
turned
up by 6dB. If this
is done,
100% modulation of the Left and
Right channels can be achieved.
There
is
no
need
to
readjust
the
8182A's
OUTPUT
ATTENUATORs
to
achieve
100% L+R or L- R: L+R at up to 100% modulation can be generated by driving both
Left
and
Right
channels
in- phase,
while L- R at
up
to
100%
modulation
can
be
generated by driving both Left and Right channels out- of- phase. ( To make the out-
of- phase connection, temporarily reverse the "+" and "-" input connections of one
channel only.)
When trying to create pure L+R or L- R, null undesired L- R or L+R ( respectively)
by
balancing
the Left and Right channel gains as closely
as possible. There
will
always be some residual linear crosstalk which cannot be nulled. This is caused by
relative phase shifts between the Left and Right channels. Because this crosstalk is
linear,
it
introduces
no
distortion
products
into
the
signal
and
has
no
audible
significance
unless
it
is
large
enough ( perhaps - 35dB
or
greater)
to
cause
deterioration of stereo imaging.
Full assessment of crosstalk performance requires a baseband spectrum analyzer to
separately
measure
the linear
crosstalk ( which
is
usually
innocuous)
and
the non-
linear crosstalk ( which can cause audible distortion if large enough). In the BTSC
system,
if
the
frequency
of
the
test
tone
is
F and
f
H
.15.734kHz,
then
linear
crosstalk
is
found
at
F ( subchannel-to-main
channel)
and
at
2fH—
+
F ( main
channel-to-subchannel). All other parts of the spectrum are non-linear crosstalk or
other
spurious
signals
except
for
the
stereo
pilot
and
the
desired
modulation.
6-2
Advertisement
