Neil said:
Oh, please.
ALL amplifiers have a damping factor. It is a fundamental aspect of
electronic amplification and output impedance.
For your >40 years of experience, you don't come across as knowing much
about amplifiers. Damping factor has nothing to do with amplification
per se. It is the ratio of the design load impedance vs the 'looking
backward' impedance, that is, if you treated the output stage as an
input stage and measured the apparent input impedance. For a good
amplifier, the ratio should be about 40 or more.
This very much depends on the design of the amplifier's output stage. for
example, amplifiers with transformer output stages (many of which would
qualify as "good amplifiers") will not have an output impedance of 0.1
Ohms. Further, very few transistor amplifiers lack resisters following the
output transistors (I don't know of any audio amps that lack them).
To clarify, a 'good' amplifier will not only have low distortion at
rated output, but will also, by virtue of a high damping factor, give
good control of speaker performance. There are plenty of amplifiers,
particularly tetrode or pentode valve amplifiers that exhibit good
distortion characteristics when operated into a resistive dummy load,
but do not give good results when feeding speakers. On the other hand,
triode tube amplifiers with 20dB or greater feedback, even though
transformer-coupled, exhibit a good damping factor, though they
generally are rather low-powered.
Hmmm. I disagree. The main issue here is the physical mass of the cone and
voice coil of the speaker, because they can not be accelerated instantly,
making it impossible to accurately reproduce a square wave, for example.
Further, overshoot and ringing are fundamental qualities of electronic
amplification, so well-damped *pre-amplifier* circuits are typically
employed to reduce the problem before it gets to the output stage.
It is precisely the physical mass and suspension of the voice coil/cone
assembly that causes overshoot and resonance. A high damping factor
aids the acceleration of the cone assembly, as the full output waveform
is available to drive the speakers. The damping factor can be regarded
as resistance in series with the output, and a voltage-divider effect
occurs with low factors, just like the internal resistance of a battery
causes a lower voltage output when loaded. When the cone assembly is
moving, back emf is generated, and if that emf sees a short circuit,
which it does when the driving signal stops, the current flowing as a
result of the voltage generated will cause heavy damping or braking of
the cone movement. This is the principle behind a high damping factor,
and indeed is the very reason it is called the damping factor, because
it damps, or supresses, unwanted cone movements.
Yes, speakers find it difficult to reproduce a square wave, but here is
a simple experiment you can try to verify what I am saying. Set up an
amplifier and speaker ( a good speaker) driven with a suitable square
wave, and then set up a double-beam or double-trace oscilloscope with
one channel connected across the voice coil at the voice coil terminals,
and the other channel connected to the amplifier output terminals at the
output terminals. Use a short piece of cable to hook up the speaker.
Observe the waveforms, which will be practically identical. Now insert
a resistor of about five ohms, and capable of dissipating whatever power
you are driving the speaker with, say 1- or 2-watts, in series with the
speaker cable, so the resistor is between the scope leads, and again
observe the waveforms. The channel connected across the speaker will
show reduced amplitude because of the series resistor, so increase the
gain on that channel. You should see ringing on the speaker waveform
that is not present on the output terminal waveform.
Ringing in pre-amplifiers is generally caused by high- and low-pass
filter circuits used for tone control in equalizers, and has nothing to
do with the mechanical ringing in speakers due to cone mass. Of course
filter circuits have to have ringing controlled in order to maintain
signal integrity, but that is a different matter.
Perhaps you should take a look at typical wire resistances. You won't find
any *lamp cord* shorter than a couple hundred feet with a resistance of
one Ohm. The reason is that *lamp cord* is typically >= 18 gauge stranded
copper wire. Look up the resistance of that if you're curious.
Yes, the 1-ohm figure was to illustrate the point, and is a bit astray
with reality. The AWG figure for 18g wire is 6.385 ohms/1000ft. In
real terms, if you have a run of say, 20 ft to your speaker, that's 40
feet of wire total, so the resistance will be 6.385*40/1000, = 0.2554
ohms. In series with a <0.1 ohm damping factor impedance, 0.255 ohms
will reduce the factor from 40 to 4/(0.1+0.255) = 11.25, practically a
four-time reduction.
And a couple hundred feet is 400 feet return, which will have
6.385*400/1000 = 2.554 ohms. A run of 18 awg with total resistance <= 1
ohm can be no longer than 78 feet of cable.
Wrong... The resistance of the speaker's voice coil is far greater than
any speaker cable (or lamp cord) one is likely to buy. The impact of
speaker cables with lengths of less than 20 ft. on the damping factor of
the amplifier will be practically immeasurable, and certainly far less
than the imperfections in even the best speaker systems. In short, you'll
never hear it.
Other imperfections in speakers are not in question here. As far as
damping of cone movement is concerned, any series resistance in the
cable reduces the ability of the amp. to control it.
Yep. This "new" approach has only been done for the last eighty years or
so. ;-)
It's only with the advent of high-power transistor or f.e.t. amplifiers
that it has become practicable to put the amplifier into the speaker
enclosure. Tube amps had too much heat and not enough damping factor to
make it advantageous.
Wrong, once again. 18 guage lamp cord can carry as much as 10 amps *at 120
Volts*, which is *1200 Watts* (though this is not a recommended loading).
Very few audio amplifiers, and even fewer speakers have 120 Volt output
circuits, with many falling in the 10 to 25 Volt range. You do the math to
figure the current @ 100 Watts.
What??? Voltage has nothing whatsoever to do with current-carrying
capacity, and neither has wattage, being a function of voltage. You
appear to be ignorant of Ohm's Law, which inter alia says that I
(current) equals W (watts) divided by R (resistance). So 100 watts/4
ohms = 25 amperes of current. High output-voltage systems, like
100-volt lines, are used to feed multiple speakers as in Public Address
systems, used in part to make calculation of how much power is fed to
each speaker easier (which can vary according to location), and reduced
losses in the cable, since higher voltage means reduced current.
Neil
(with > 40 years of designing pro audio systems)
(and some *really expensive* lamp cord, er, "speaker cables" to sell you
if you still disagree) ;-)
An amateur designer, perhaps? Some of your ideas appear to run counter
to accepted theory.
Colin D.
PS: another experimant you can try is to find an analog meter, and
oscillate the meter body, noting the amount of movement required to set
the needle swinging. Now, connect a piece of wire across the meter
terminals and oscillate the meter again. Note the needle is much harder
to set swinging, a direct effect of short-sircuit damping. Q.E.D.
