Biamplification vs. Bridging
Power Amplifiers
By Jon M. Risch, Senior Project Engineer, Peavey Electronics
Corp.
When seeking more output from a sound system, it is common
to bridge the power amplifier to increase the power available
and generate more output from the system. In some cases, this
is the best thing to do, especially if it is a subwoofer being
used with a medium to low output power amp.Bridging the amp
will give the usual 6 dB more power output, or 4 times the
single channel power output into the same load impedance. However,
the bridging must be into a load impedance that is twice the
minimum allowable load for the single channel. Subwoofers often
need some headroom to maintain punch and impact, so use of
a bridged amp is not that unusual.
However, if the load is a full range speaker system, the
bridged amp now has enough power to overdrive the tweeter
very readily if an accident should occur, such as microphone
to monitor feedback, or an electronics related feedback loop.
Before you can even react, the tweeter may be damaged, or
a protection circuit may trip or blow. In this situation,
the increased power may affect reliability.
An alternative to power amp bridging is biamplification.
When using full range speakers that have a biamplification
option, biamplification can often provide more actual sound
output capability than bridging an amp can, with better reliability.
What exactly is biamplification? Biamplification is the splitting
of the full range sound signal into high frequency components,
and low frequency components, similar to what the passive
internal crossover is doing inside the speaker, but only
it is done at line level via an electronic crossover circuit,
and then fed to two separate power amp channels.
In order to more clearly explain biamplification, let's
look at some diagrams. We'll examine some diagrams that are
analogous to an oscilliscope "snapshot" of a particular
signal. In Fig. 1, a low frequency sound is being reproduced
by a single channel of a power amp into a speaker load, and
it is not clipping.
In Fig. 2, a high frequency is being reproduced through
that same amp channel, and is also being reproduced cleanly.
But add the two frequencies together, and the power output
required to reproduce the combination exceeds the available
power output of the single channel. This combined signal
is shown in Fig. 3, and the clipping of the high frequency
component is readily noticed.

If the high and low frequencies are separated at line level,
and sent to two different amp channels, then they can be
reproduced cleanly, and the summing will occur acoustically.
This is shown in Fig. 4. The net result is that actual output
SPL has been increased.

The use of a line level crossover to biamplify has several
advantages. First, the passive crossover has internal losses,
losses in the series inductor for the woofer, and deliberate
losses in the padding for the tweeter to match level to that
of the woofer. The losses through the inductor for the woofer
can be from 1 to 2 dB, so you can see that some output level
is lost from using the passive internal crossover. A direct
connection to the power amp without any intervening series
inductors will usually increase woofer damping, and tighten
up the bass character.
Second, most pro sound speaker systems that are biamp capable
use compression driver tweeters mounted on a high frequency
horn. These compression drivers are as much as 10-12 dB more
sensitive than the woofer at the peak of their output, and
an average of 6-8 dB more efficient over the entire range
of the tweeter, requiring the tweeter to be padded down to
match the woofer output. This loss of the inherent sensitivity
of the tweeter has important implications for biamplification.
Third, there is a hearing phenomenon known as masking, where
one frequency can dominate and overshadow another. In the
case of biamplification, the clipping of the woofer will
be masked by clean output from the tweeter, so that even
when the woofer section of the biamplified full range system
is clipping, it is not as easily heard, due to the clean
tweeter output. This effect will work until the woofer amp
is generating enough clipping harmonics to overcome the clean
tweeter output levels. This is in contrast to the single
channel of amplification, which clipped the high frequencies
first, making the clipping much more obvious. Fig. 5 shows
the woofer signal clipping by 3 dB.
See Fig. 3 again, and compare to Fig. 6, which shows biamplified
woofer clipping of approximately 3 dB.
The undistorted combined waveforms are shown in Fig. 4.
The sound of the clipped waveform in Fig. 3 will be much
more audible than the clipped waveform in Fig. 6, both show
equal levels of clipping.
So how much extra output does biamplification give you compared
to bridging the same amp? First let's start with the assumption
that the power amp is not a very small one, but has a nominally
adequate amount of power for the woofer of a full range system,
say around 100 watts per channel into 8 ohms. If you bridge
such a power amp, and it is capable of 4 ohm per channel
operation and 8 ohms bridged, then the power output will
be 400 watts in bridged mode. This is a power increase of
6 dB, and, if the speaker system can handle the full amount
of this increased amount of power, then the SPL should increase
by about 6 dB also. As a benchmark, it helps to realize that
the common criteria to double the apparent loudness of a
sound is to increase the SPL by 10 dB, or ten times the power.
Biamplification is classically said to increase acoustic
output by 6 dB also. However, this assumes equal power for
the highs and lows, and also assumes equal speaker sensitivities.
In addition, it must also involve a signal with a spread
of frequencies across the audio band, in order to take advantage
of the division of labor that occurs. This describes most
modern music as played through a full range speaker system.
So a 6 dB increase in sound output for full range music through
a biamplified system is a minimum amount, not taking into
consideration those three factors we wrote about earlier.
Let's add those in to the equation now.
Page 2
|