Damping Factors
Gerral Hubbard
I know that it is effected by somethings like speaker cable resistance, etc.
MKellom
Damping factor is a measure of the output impedance of an amplifier. It's
calculated by dividing the load impedance you're driving (4 ohms, 2 ohms,
etc.) by the output impedance of an amp.
The reason it's an important number is that speakers, and especially
woofers, don't stop moving as soon as the amp stops sending signal. The
inertia of the driver cone keeps the cone moving, and now all of a sudden
the speaker turns into something of a "microphone" in the sense that the
motion of the cone will induce some current flow in the voice coil of the
speaker. This current is going to get back to the amp, and an amp with a
lower output impedance is going to do a much better job of "resisting"
this current (and therefore resisting the continued motion of the cone)
than a model with higher output impedance. The net result is better
control of LF drivers, i.e. "tighter" bass.
The speaker wire gets involved in the picture because it has some
resistance of its' own. Any resistance the wire possesses will be added in
series to the output impedance of the amp. So if an amp has an output
impedance of .01 ohms (at 2 ohms, 2/.01=damping factor of 200), and you
add 1 ohm of resistance due to speaker wire, your damping factor is now
about 2/1.01=2.
Hope this helps!
Marc Kellom
Design Engr., Crown International
MKe...@aol.com
Shankar Ramakrishnan
>>Could someone (who speaks english) please explain exactly what Damping
>>Factors are?
>
>>I know that it is effected by somethings like speaker cable resistance,
>etc.
>
>Damping factor is a measure of the output impedance of an amplifier. It's
>calculated by dividing the load impedance you're driving (4 ohms, 2 ohms,
>etc.) by the output impedance of an amp.
>
>The reason it's an important number is that speakers, and especially
>woofers, don't stop moving as soon as the amp stops sending signal. The
>inertia of the driver cone keeps the cone moving, and now all of a sudden
>the speaker turns into something of a "microphone" in the sense that the
>motion of the cone will induce some current flow in the voice coil of the
>speaker. This current is going to get back to the amp, and an amp with a
>lower output impedance is going to do a much better job of "resisting"
>this current (and therefore resisting the continued motion of the cone)
>than a model with higher output impedance. The net result is better
>control of LF drivers, i.e. "tighter" bass.
The above definiton of damping factor (i.e., the ratio of speaker impedance
to amp. output resistance) is not very significant as the true damping factor
of the overall system (which, by the way, is of the order of unity for most
systems). The first definiton only gives a qualitative estimate as to how
much the non-zero impedance of the amp and the speaker cable would affect
the actual damping factor of the overall system. For sealed woofer systems,
a value of 20 would suufice; for vented systems (which are typically more
sensitive to overall Qts of the driver), a value of about 50 or more
is adequate.
Shankar
Allan Borr
[snip]
>>
>>Damping factor is a measure of the output impedance of an amplifier. It's
>>calculated by dividing the load impedance you're driving (4 ohms, 2 ohms,
>>etc.) by the output impedance of an amp.
>The above definiton of damping factor (i.e., the ratio of speaker impedance
>to amp. output resistance) is not very significant as the true damping factor
>of the overall system (which, by the way, is of the order of unity for most
>systems). The first definiton only gives a qualitative estimate as to how
>much the non-zero impedance of the amp and the speaker cable would affect
>the actual damping factor of the overall system. For sealed woofer systems,
>a value of 20 would suufice; for vented systems (which are typically more
>sensitive to overall Qts of the driver), a value of about 50 or more
>is adequate.
>
>Shankar
This sounds like gibberish to me . Perhaps I'm missing something fundamental
here. All the amps I own are solid state and have typically have
source impedances of a tiny fraction of an ohm. My speakers represent
a load that is tens or even hundreds of times as large over the whole audio
range. Where do you get your notion that damping factors of the order of unity
are typical?
Al Borr
Megatest Corp.
San Jose, California
Gerral Hubbard
I've got two speakers (each is a two way box using an internal passive crossover) and they show an
8 ohm impedence.
1. If I put one speaker on each side of an amp each side would see 8 ohms.
2. If I were to use a cross over and bi-amp the two speakers, using the same amp and running the
high ends on one channel and low ends on the other it would show the amp a 4 ohm load on each
channel.
Advantages of bi-amping aside; I would have a better sound because of the higher damping factor
(using 4 ohms rather than 8 ohms).
Does this sound alright?
David I. Baldwin
wire total) has .08 ohms of resistance, and ten feet of 12 gauge has .032
ohms of resistance. On simple terms, if your amp has .01 ohms output
'resistance', the wire dominates the damping. Connectors also add
resistance. With 8 ohm speaker load and 0 ohm wire, the damping factor
would be 800; with the 16 gauge wire the damping drops to 88. With 4 ohm
load, the damping would be 44. This is a simplified calculation, a
number of things are left out that reduce the effective damping even more.
--
Dave Baldwin Internet : dib...@netcom.com
DIBs BBS : (916) 722-5799 Compuserve: 70403,2444
DIBs FAX : (916) 722-3877 Genie : D.BALDWIN3
-=-=-=- @#$%^&* I can't even quote myself! Oh,well. -=-=-=-
John Busenitz
>they show an 8 ohm impedence.
>
>1. If I put one speaker on each side of an amp each side would see 8 ohms.
>2. If I were to use a cross over and bi-amp the two speakers, using the same amp and running the
>high ends on one channel and low ends on the other it would show the amp a 4 ohm load on each
>channel.
So they whole system is a monaural system? Is the crossover line level or is it
speaker level? Firstly, no speaker has a simple resistive 4 ohm or 8 ohm
load, and with a filter (crossover), one channel would have a very large
load impedance at low frequencies, and the other at high. I'm not sure what
the point is here, but I'll play along.
>Advantages of bi-amping aside; I would have a better sound because of the higher damping factor
>(using 4 ohms rather than 8 ohms).
>Does this sound alright?
No, because there is simply no evidence to suggest that damping factor
(as long as it is over 10) is related to better sound, however you
define "better". I bet that the change in system transfer function
would be barely noticable in measurements, and much less audible.
The DCR of the speaker is still so much higher than any other losss
in the system, the changes are negligable. Since one permutation of
the system is mono, and one stereo, you could hear a difference. But
it would not be due to the change in damping factor. You might as
well suggest putting expensive wooden disks on top of the amplifier and
claim that it changed the sound.
--
John Busenitz P.U. E.E. "Any idiot can design a loudspeaker, and,
buse...@ecn.purdue.edu unfortunately, many do." - Dick Pierce
:::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::
Disclaimer: My statements do not represent Purdue University.
Richard D Pierce
David I. Baldwin <dib...@netcom.com> wrote:
>From standard wire tables, ten feet of 16 gauge speaker cable (20 ft. of
>wire total) has .08 ohms of resistance, and ten feet of 12 gauge has .032
>ohms of resistance. On simple terms, if your amp has .01 ohms output
>'resistance', the wire dominates the damping.
Wrong, if you amp as 0.1 ohms output resistance, your cables have .08
ohms resistance, and your speaker's voice coil has 6.5 ohms resistance,
the voice coil dominates the damping. If fact, this is true of pretty
much every situation, save some very rare pathological cases.
>Connectors also add
>resistance. With 8 ohm speaker load and 0 ohm wire, the damping factor
>would be 800; with the 16 gauge wire the damping drops to 88. With 4 ohm
>load, the damping would be 44. This is a simplified calculation, a
>number of things are left out that reduce the effective damping even more.
Yeah, like the DC resistance of the voice coil, which has been shown time
and time again to be the dominant resistance in the entire chain (see
Small, et al).
--
| Dick Pierce |
| Loudspeaker and Software Consulting |
| 17 Sartelle Street Pepperell, MA 01463 |
| (508) 433-9183 (Voice and FAX) |
David I. Baldwin
question. The 'common' definition of damping factor that I know is viewed
from the speaker terminals with the speaker load on one side and the
amplifier and cables on the other. 'Amplifier' damping factor is
something people can play with and make changes to whereas the 'system'
damping factor as I think Dick is refering to is beyond most of us to do
anything about.
Richard D Pierce
>question. The 'common' definition of damping factor that I know is viewed
>from the speaker terminals with the speaker load on one side and the
>amplifier and cables on the other. 'Amplifier' damping factor is
>something people can play with and make changes to whereas the 'system'
>damping factor as I think Dick is refering to is beyond most of us to do
>anything about.
Well, given that adding a few milliohms here, and a centiohm or two
there, and STILL you are faced with a single parasitic series resistance
that is a couple of orders of magnitude larger, what possible differences
does it make.
And what is "amplifier damping factor?" What difference is there between
that and the system damping factor, which, in and of itself, is a pretty
meaningless term invented SOLELY for specsmanship. As such, it is
officially defined as the ration of the nominal load impedance (say, 8
ohms) to the amplifier source resistance). It's a meaningless
specification, in that, again, ignores the fact that a factor of 100
change in the amplifier source resistance (say, from 0.001 ohms to 0.1
ohms), equivalent to a difference of damping factors of 8000 vs 80, is
not going to make one damned bit of difference how the system performs.
Here's an excerpt from a post I made a while ago on an analysis of the
utter myth of damping factor:
> There are a lot of things that decrease the damping factor of the whole
> system:
>
> 1) speaker cable
> 2) coil in the speaker
> 3) plugs etc.
I've done this analysis SO many times, I can do it in my sleep, but here
goes again.
The damping factor proponents start throwing things like cables and plugs
and even the inductors (what I presume you mean by "coils") into the
equation. Unfortunately, they ALL ignore the largest single resistance in
the whole chain: the one the utterly swamps ALL other resistances
combined: the DC resistance of the voice coil.
Assume (which is reasonable) an amplifier with a rated 8 ohm damping
factor of, say, 1000. Add .25 ohms for speaker cables, .25 ohms for
inductors, .25 ohms for plugs (pretty awful, but what the hey), for a
grand total series resistance of 0.008 (amp) + 0.25 (cables) + 0.25 (coil)
= 0.758. Now, it's argued, the damping factor has been spoiled to a
whopping figure of 10.6.
That's awful, right?
Well, it's a factor of 100 poorer than the spec of 1000, for sure, but
neither the damping factor NUMBER of 1000 nor of 10 HAS ANY SIGNIFICANCE IN
DETERMINING THE BEHAVIOUR OF THE LOUDSPEAKER ITSELF.
The amount of ACTUAL damping of the driver is determined by a DRIVER
figure called the total system Q, or Qtc. This is the ratio of the amount
of energy stroed in the resonance to the amount of energy dissipated, and
is a DIRECT determination of how well the speaker is damped. It is made
up of two major parts, the system Q due to mechanical losses Qmc and the
system Q due to all electrical losses Qec'. The total Q is related to
these by the following relation:
Qmc * Qec'
Qtc = -----------
Qmc + Qec'
Now, the part the "damping factor" or ALL series resistance play is in
changing the electrical Q, Qec'. There is a figure, Qec, for the
electrical Q with 0 source resistance. The two are related:
Re + Rg
Qec' = Qec ---------
Rg
where Re is the voice coil DC resistance and Rg is the TOTAL source
impedance.
Let's look at the change in Qec' and thus Qtc win the difference between
a damping factor of 1000 and 10, due to a change in source impedance of
.75 ohms. Let's assume a PERFECT 2nd order maximally flat butterworth
sealed box system, with a Qtc of 0.7071. Assume (which is reasonable) the
mechanical losses in the system (including absorbtion andfrictional
losses) result in a Qmc of 4.0. We can derive the electrical losses as
Qmc * Qtc
Qec = -----------
Qmc - Qtc
which means a Qec of 0.859. Now, the typical resistance of a voice coil
on an 8 ohm driver is around 7 ohms (go measure it). Now, we can
recaculate the new Qec' with the added 0.75 ohms of series resistance:
7 + 0.75
Qec' = 0.859 ----------
0.75
or a new Qec' of 0.951. Put that back into the total Q:
4.0 * 0.951
Qtc = --------------
4.0 + 0.951
With a new system Q of 0.768. That's an increase of about 9% in the
system Q. It's VERY important to note at this point that this is less
variation then you find in manufacturing tolerance of high quality
woofers, or variation in performance due to atmospheric changes, so from
this viewpoint alone, the change is swamped by other effects.
Now, how much is the response of the speaker changed at resonance due to
this. In otherwords, how has the new series resistance, which lowered the
damping by a factor of 100 (oh gosh) decreased control of the speaker at
resonance? How much is the speaker underdamped?
Well, the original system is maximally flat, it's response show no rise
to cutoff and then rolls of smoothly. A higher Qtc results in a bump in
the response at resonance. The magnitude of that bump is both measurable
and calculatable:
4
Qtc
Gmax = sqrt [-------------]
2
Qtc - 0.25
Thus, this new series resistance generates a peak of 0.1 dB, FAR less
than perturbations caused in manufacturing tolerances, environmental
tolerances, room response and so on.
Now, this is not to deny that there may be an audible difference between
amolifers with different damping factors, but until you rigorously
eliminate all other possible causes, then you cannot point at damping
factor as a cause of such differences because damping factor, over the
ranges we have explored has an influence that is FAR lower than any one
of a vast number of other influences.
Shankar Ramakrishnan
You forget that the overall damping factor of the amp is largely determined
by the sum of the Dc resistance of the woofer's voice coil, crossover
inductor(s), the speaker cable, and the amplifier's output impedance.
generally, the latter is an insignificant part of the overall contribution.
If the damping factor of the amp is low, this leads to a higher resistance
which, paradoxically enough, increases the damping factor fo the overall
system.
Shankar
Shankar Ramakrishnan
Oops. The last statement was a mistake. Increasing total resistance increase
Qts leading to a lower damping factor.
Shankar
Gerral Hubbard
>>I've got two speakers (each is a two way box using an internal passive crossover) and
>>they show an 8 ohm impedence.
>>
>>1. If I put one speaker on each side of an amp each side would see 8 ohms.
>>2. If I were to use a cross over and bi-amp the two speakers, using the same amp and running the
>>high ends on one channel and low ends on the other it would show the amp a 4 ohm load on each
>>channel.
>
>So they whole system is a monaural system? Is the crossover line level or is it
>speaker level? Firstly, no speaker has a simple resistive 4 ohm or 8 ohm
>load, and with a filter (crossover), one channel would have a very large
>load impedance at low frequencies, and the other at high. I'm not sure what
>the point is here, but I'll play along.
>
The point of the question was this:
I have a mono P.A. cluster, using 4, two-way boxes (each have their own internal passive
cross-over. They are powered by two Crown MT1200 amps. Each box is attached to an individual
output of the amp (stereo mode but run with mono signal)
I am investigating the advantages of using an active crossover, grouping the high ends on one amp
(two drivers per channel), and the lows on the other.
I am exploring what relevant effects this has with damping, etc.
David I. Baldwin
please. I did look there first.
Richard D Pierce
Richard D Pierce <DPi...@world.std.com> wrote:
>Here's an excerpt from a post I made a while ago on an analysis of the
>utter myth of damping factor:
In cleaning up the original post, I added a typo. The correct equation for
the corrected Qe SHOULD be:
Re + Rg
Qec' = Qec ---------
Re
Sorry for the confusion this might have caused.
Jan-willem Stekelenburg
> Dick Pierce, would you put your explanation in the rec.audio faq,
Where can I get this faq, please?
met vriendelijke groeten,
Jan-Willem Stekelenburg
>------------------------------------------------------------------------<
> STEM Musical Engineering | internet: j...@tess.nl <
> Maarssen, The Netherlands | fax : +31 346 552205 <
>------------------------------------------------------------------------<
... As irritating as a teenage Vulcan.
John Busenitz
>cross-over. They are powered by two Crown MT1200 amps. Each box is attached to an individual
>output of the amp (stereo mode but run with mono signal)
>
>I am investigating the advantages of using an active crossover, grouping the high ends on one amp
>(two drivers per channel), and the lows on the other.
>
>I am exploring what relevant effects this has with damping, etc.
I see. I would at least try it out, and I would guess that you would
get good results. It sounds like a good solution to me. Make sure,
of course, to select a filter alignment that works well with the
drivers. Any old crossover won`t work, as I'm sure you are aware.
Unless the DF gets insanely low, differences in sound cannot be
contributed in DF change, as all losses are dwarfed and made
negligable by the VC DCR.
Good luck.
tINY
>Here's a senario,
>
>I've got two speakers (each is a two way box using an internal passive crossover) and they show an
>8 ohm impedence.
>
>1. If I put one speaker on each side of an amp each side would see 8 ohms.
>2. If I were to use a cross over and bi-amp the two speakers, using the same amp and running the
>high ends on one channel and low ends on the other it would show the amp a 4 ohm load on each
>channel.
>
>(using 4 ohms rather than 8 ohms).
>
>Does this sound alright?
Backwards (and irrelevant if you have a solid state amp), if I understand the
term properly. It seems to me that the problem lies in the voltage drop across
the source resistance (and cable resistance). This would cause less than the
full voltage to be fed back to input of the amp when the speaker generates
voltage.
Anyone who knows for sure care to comment.
-tINY
tINY
>The point of the question was this:
>
>I have a mono P.A. cluster, using 4, two-way boxes (each have their own internal passive
>cross-over. They are powered by two Crown MT1200 amps. Each box is attached to an individual
>output of the amp (stereo mode but run with mono signal)
>
>I am investigating the advantages of using an active crossover, grouping the high ends on one amp
>(two drivers per channel), and the lows on the other.
>
>I am exploring what relevant effects this has with damping, etc.
Unless you are designing speakers for the "Gold Bricks"
crowd, forget damping factor.
The best set-up (probably) will be all four horns in a
series-parallel conection on one amp, bridged, and all four LF
drivers on the other amp. (I think the MT will drive 2 ohms
bridged, but you better check)
Better yet, go buy a 200 to 600 W x 2 Ch amp for the horns
and use both MT1200's for the woofers (if they can take that kind
of power). The idea here is that the horns are MUCH more efficient
than the typical direct radiating woofer, and require less power.
It is not unusual for HF horns to be 12dB more efficient than
direct-radiating LF drivers. This is 4x the power.
You may consider disconecting two of the horns (unless you
need the coverage) and splay the horns so that the -6dB points
just overlap. This will reduce comb filtering.
-tINY
Chris Christensen
[text edited]
>>I have a mono P.A. cluster, using 4, two-way boxes (each have their own
>>internal passive cross-over.
>>I am investigating the advantages of using an active crossover, grouping the
>>high ends on one amp (two drivers per channel), and the lows on the other.
Good Sound idea.
>>I am exploring what relevant effects this has with damping, etc.
Just what does the fact that you want to change the system configuration have
to do with _damping factor_
If the internal crossovers are really bad, ie high impedance of inductors, etc.
you might realize a difference in _caculated_ damping factor.
The facts are that, the voice coil DC resistance is dominant term in the
equation.
>I see. I would at least try it out, and I would guess that you would
>get good results. It sounds like a good solution to me. Make sure,
>of course, to select a filter alignment that works well with the
>drivers. Any old crossover won`t work, as I'm sure you are aware.
>Unless the DF gets insanely low, differences in sound cannot be
>contributed in DF change, as all losses are dwarfed and made
>negligable by the VC DCR.
Commercial active crossovers come in several different slopes. The most
popular slope is 24dB per octave. I suggest that you go for a Rane AC22.
While the 24dB per oct. slope of the Rane isn't optimum for all system designs
it won't cause any problems.
Back to the topic, I believe that if you Bi-amp the system you will net an
audible quality increase. NOT because of damping factor, because you have
bypassed a poor crossover.
--
Just My opinion, worth the price paid and not a reflection of my employer.
D.R. "Chris" Christensen chr...@shasta.gvg.tek.com
Grass Valley Group Inc. 916-478-3419 Voice 916-478-3887 FAX
P.O. Box 1114 Grass Valley, California, 95945
Chris Christensen
>In article <dibaldDC...@netcom.com>,
>David I. Baldwin <dib...@netcom.com> wrote:
WHEN ARE YOU PEOPLE GOING TO LEARN!
It only took be 3-4 times to understand what Dick has been saying about Damping
factor!
Look at it again:
[Poster notes]
>>From standard wire tables, ten feet of 16 gauge speaker cable (20 ft. of
>>wire total) has .08 ohms of resistance, and ten feet of 12 gauge has .032
>>ohms of resistance. On simple terms, if your amp has .01 ohms output
>>'resistance', the wire dominates the damping.
[The great one speaks]
>Wrong, if you amp as 0.1 ohms output resistance, your cables have .08
>ohms resistance, and your speaker's voice coil has 6.5 ohms resistance,
>the voice coil dominates the damping. If fact, this is true of pretty
>much every situation, save some very rare pathological cases.
Damping factor is only a consideration when you use a transformer. That
would be in tube amps and 70.7V line systems.
[poster tries to find a point]
>>Connectors also add
>>resistance. With 8 ohm speaker load and 0 ohm wire, the damping factor
>>would be 800; with the 16 gauge wire the damping drops to 88. With 4 ohm
>>load, the damping would be 44. This is a simplified calculation, a
>>number of things are left out that reduce the effective damping even more.
[the great one retorts]
>Yeah, like the DC resistance of the voice coil, which has been shown time
>and time again to be the dominant resistance in the entire chain (see
>Small, et al).
[I drool on]
I am a total idiot and this makes sense to me.
Speaker is 6.5 Ohms DC resistance.
Amp and speaker cable= .18 Ohms.....
What's so difficult about that picture?
And just what is _so_ important about damping factor?
Once again, It might be important if you are talking a tube amp. But, so far
every model shown in this discussion has been a semicondictor amp. The
discussion is moot!
Arther Dent
>
> Just to head off on a bit of a tangent, I've seen incredible prices for
> very short lengths of audiophile quality speaker cable. I'm not
> questioning that something made out of gold and hand made etc etc should
> be priced so high, just whether the effect on the sound is worth it.
>
> Can you hear the difference between extremely pricey speaker cable and a
> "standard" OFC type cable?
I can, yes. If you can, get some "pricy" cable. if not, stick with OFC.
>
> Is there enough of an audible difference to merit spending hundreds of
> $$$ on cable alone, especially given the recent revelation (to me anyway)
> that damping factor is not as important a quantity as I first thought?
Well, only YOU can justify if a slight difference is worth the $$ (dosen't
even have to be hundreds, many $50 interconnetcts sound better to me, than
some $500 ones)
Arther.
Espen Braathen
: Could someone (who speaks english) please explain exactly what Damping Factors are?
: I know that it is effected by somethings like speaker cable resistance, etc.
Power-amplifier output impedance is usually a very low value,
between 0,05 and 0,5 ohms for solid-state designs. Output impedance is
related to an amplifiers damping factor - the ability to control
cone motion in the loudspeaker. The higher the damping factor, the better.
High output impedance = low damping factor and vice versa.
Resistance in the loudspeaker cables will increase output impedance
and therefore decrease the damping factor.
EspenB
Dean Aldridge
very short lengths of audiophile quality speaker cable. I'm not
questioning that something made out of gold and hand made etc etc should
be priced so high, just whether the effect on the sound is worth it.
Can you hear the difference between extremely pricey speaker cable and a
"standard" OFC type cable?
Is there enough of an audible difference to merit spending hundreds of
$$$ on cable alone, especially given the recent revelation (to me anyway)
that damping factor is not as important a quantity as I first thought?
What are the factors other than damping, that merit spending big $$$ on
speaker cable? What's wrong with grabbing some mains rated twin flex
from any old piece of equipment and using that for speaker cable? As
long as it can handle the power of your amp, would there be much/any
audible difference?
thanks in advance
/Dean
Daniel Maverick Falkoff
I haven't been following this thread, but will take the risk
(of repeating common knowledge). This whole issue
of needing special thick low resistence, low inductance
speaker wires would be almost irrelevant IF ONLY the
power amplifier employed REMOTE SENSING, like
all reasonable power supplies do.
Greg J Szekeres
Daniel Maverick Falkoff <d...@c-c.com> wrote:
.>
.>I haven't been following this thread, but will take the risk
.>(of repeating common knowledge). This whole issue
.>of needing special thick low resistence, low inductance
.>speaker wires would be almost irrelevant IF ONLY the
.>power amplifier employed REMOTE SENSING, like
.>all reasonable power supplies do.
What kind of wire would you use for the sense lines.
greg
Henry Pasternack
: The reason it's an important number is that speakers, and especially
: woofers, don't stop moving as soon as the amp stops sending signal. The
: inertia of the driver cone keeps the cone moving, and now all of a sudden
: the speaker turns into something of a "microphone" in the sense that the
: motion of the cone will induce some current flow in the voice coil of the
: speaker. This current is going to get back to the amp, and an amp with a
: lower output impedance is going to do a much better job of "resisting"
: this current (and therefore resisting the continued motion of the cone)
: than a model with higher output impedance. The net result is better
: control of LF drivers, i.e. "tighter" bass.
Unfortunately, this is a misleading oversimplification, albeit
a very common one. The implication is that by increasing the
damping factor to infinity, the amplifier will perfectly
"control" the motion of the loudspeaker cone, which will follow
exactly the voltage variations at the amplifier output terminals.
This is not at all what happens in real life.
In fact, as Shankar Ramakrishnan points out, in a proper
loudspeaker system, the system bass damping factor is on the
order of unity, usually a bit lower. This is the system damping
factor, mind you, and assumes a very low source impedance. The
DC series resistance of a typical 8 Ohm woofer is on the order of
5 Ohms, and even with zero amplifier output impedance, the
amplifier will never "control" the cone motion. In fact, all
conventional loudspeaker drivers are resonant systems. The sound
pressure output is not proportional to cone displacement, but to
velocity. For constant SPL output, the peak cone displacement
varies inversely with frequency. The amplifier imparts energy
to the cone, but the actual motion depends very much on the resulting
resonant motion, which is determined primarily by the spring and
damping characteristics of the driver, and to a lesser degree on
the damping effect of the air itself.
I once wrote a long article explaining all of this in greater
detail. Unfortunately, I don't have a copy here now. In that
article, I said, and I think it was justified, that the amplifier
does not so much control the cone as it gives it "hints" how it
should move. If the system damping factor were infinite, the low
frequency cutoff of the driver would be infinite and the driver
would produce no output.
This is all rather counterintuitive, and it contradicts a very
popular notion of the relationship between bass response, back
EMF, feedback, and amplifier output impedance. The correct tuning
of the loudspeaker depends on adjusting the damping factor so that
the low frequency response is extended without being peaky. The
fact that the phase response of a loudspeaker varies rapidly near
the LF cutoff is a consequence of the resonant tuning. As with
most engineering problems, speaker design is a compromise between
many competing priorities including wide bandwidth, flat response
(both in phase and amplitude), high sensitivity, and high power
handling. The whole business is surprisingly complex.
By the way, although I think Shankar has the right idea, I
disagree when he says the amplifier damping factor depends on the
total series resistance in the circuit. Incorrect as it may be,
amplifier damping factor is standardly defined precisely as 8
Ohms divided by the output impedance.
-Henry
Scott Dorsey
>Just to head off on a bit of a tangent, I've seen incredible prices for
>very short lengths of audiophile quality speaker cable. I'm not
>questioning that something made out of gold and hand made etc etc should
>be priced so high, just whether the effect on the sound is worth it.
>
>Can you hear the difference between extremely pricey speaker cable and a
>"standard" OFC type cable?
I can't tell you what you can hear. But I can tell you what I can hear. I
did some short tests, using a Citation II amplifier and my Magnapan speakers,
with about ten feet of cable from one to the other. The Maggies are pretty
close to a 6 ohm resistive load, with a few warts here and there on the
plot. The Citation II is a tube amplifier with oversized output transformers,
and I was using the 4 ohm taps on the transformers. (Do note that the
characteristic output of the amplifier may well have accentuated differences
between cables, but by the same token the flat Z/frequency of the speakers
should have minimized them).
Differences in gauge were very obvious; the 22 AWG cable sounded a lot worse
than the 10 AWG cable. This wasn't so surprising, since the resistive loss
easily accounts for anything here. However, the difference between the
10 AWG and 8 AWG cables with the same stranding were also pretty easy to
tell, which amazed me.
Surprisingly, though, the differences in stranding were also very clear,
with the more finely stranded cable sounding better to my ear.
Just as surprisingly, the differences between dielectric materials were
not audible at all. I might have been able to tell a difference between
a Teflon and a very lossy rubber insulation, but the difference was very
subtle and was probably in my head. (I still don't recommend getting
cables with vinyl or rubber insulation, though, because they tend to fail
a lot.)
I heard a difference between silver and copper cable, and the copper
sounded better. I didn't try any of the OFC or six-nines copper.
>Is there enough of an audible difference to merit spending hundreds of
>$$$ on cable alone, especially given the recent revelation (to me anyway)
>that damping factor is not as important a quantity as I first thought?
That depends. You're into the level of diminishing returns here. There
are some audible differences between cables, but just because it's more
expensive doesn't mean it sounds any better. Whether the difference is
significant enough to you depends on how much of it you notice and how much
money you have.
>
>What are the factors other than damping, that merit spending big $$$ on
>speaker cable? What's wrong with grabbing some mains rated twin flex
>from any old piece of equipment and using that for speaker cable? As
>long as it can handle the power of your amp, would there be much/any
>audible difference?
I would recommend getting a large cable with the finest stranding you
can find, and not worrying so much about it. While there might be an
audible difference, it won't blow you away to the point where you'll
want to spend a lot of money.
I will also recommend the ribbon speaker cables, like the Goetz M-1.
These won't hold up to PA system use, but they're strongly recommended
for studio monitors, and they're dirt cheap. Personally I am using
two pieces of government-surplus flat copper braid, fused together in
a plastic jacket intended to hold 120 film negatives. Cost me next to
nothing, sounds very fine, and slides under the rug without leaving
a bump.
--scott
--
"C'est un Nagra. C'est suisse, et tres, tres precis."
Scott Dorsey
>
>I haven't been following this thread, but will take the risk
Yes, but this presupposes that special low resistance, low inductance
cable _is_ needed, and that the losses in the remote sensing lines will
be less significant.
There actually are amplifiers which use current feedback to reduce such
effects; the Electro-Voice amplifiers designed around the Wiggins Circlotron
output stage used current feedback which was adjustable so you could change
the damping factor. Also, Fisher made several amplifiers with Z-Matic,
which was his trademark for a similar current-feedback scheme.
In actuality, with modern low-Z output semiconductor amplifiers, it's not
half the big deal that it was in the past.
Daniel Maverick Falkoff
>From: szek...@pitt.edu (Greg J Szekeres)
>Subject: Re: speaker cable, was Re: Damping Factors
>Date: 26 Jul 1995 16:34:05 GMT
>greg
Anything you'd use for a microphone, telephone, or coaxial
signal, with suitable termination should be workable. The signal
at the speaker could be attenuated with resisitors and returned
balanced, ideally. The best feedback (I'm NOT advocating)
would be a microphone listening to the speaker. Too many
problems. But this remote sensing is really practical, I think.
Richard D Pierce
Mark S. Gahr <iau...@Walden.MO.NET> wrote:
>Richard D Pierce (DPi...@world.std.com) wrote:
>: In article <dibaldDC...@netcom.com>,>: >From standard wire tables, ten feet of 16 gauge speaker cable (20 ft. of
>: >wire total) has .08 ohms of resistance, and ten feet of 12 gauge has .032
>: >ohms of resistance. On simple terms, if your amp has .01 ohms output
>: >'resistance', the wire dominates the damping.
>
>: the voice coil dominates the damping. If fact, this is true of pretty
>: much every situation, save some very rare pathological cases.
>
>even mentioned (I _could_ be blind!) that damping factor is another
> one of those frequency related topics. These devices are not purely
>resistive.
Well, first, let me state that if you follow the analysis, the standard
definition is damping factor is essentially a useless piece if
information for determining, for example, the transient response of an
integrated loudspeaker-amplifier system, which it is supposedly intended
to do: it's utterly meaningless.
And, yes, you are right, the "damping factor", useless as it is, IS
frequency dependent for a variety of reasons:
1. The output impedance of the amplifier is seldom a frequency-
independent quantity, usually decreasing with increasing
frequency.
2. The resistive portion of the impedance presented by the loudspeaker
is also frequency dependent.
3. Etc...
There you have it: a useless but clearly and provably frequency-dependent
parameter!
Now, the only point where the action of damping is significant is where
the ratio between the motional impedance and the lossy portion of the
system is high, and that, for the vast majority of loudspeakers,
occurs only at and around the system cutoff frequency due to the large
motion of the cone at resonance. And there the lossy portions of the
system (those elements that can dissipate energy) include the amplifier
source impedance, the speaker cables, the crossovers, the connection
resistances and, oh yes, the DC resistance of the voice coil which, as we
have seen, totally swamp all other effects.
At higher frequencies, you could well have a significant cabinet or cone
resonant mode, but, at those frequencies, the voice coil is no longer in
control of the cone or enclosure anyway, so no amount of damping is going
to havbe any appreciable effect.
Damping factor is little more than another linchpin in the war of useless
specsmanship.
John Busenitz
>loudspeaker system, the system bass damping factor is on the
>order of unity, usually a bit lower. This is the system damping
>factor, mind you, and assumes a very low source impedance. The
>DC series resistance of a typical 8 Ohm woofer is on the order of
>5 Ohms, and even with zero amplifier output impedance, the
>amplifier will never "control" the cone motion.
Would it then be more correct to use the term "damping ratio"
to refer to the ration of the amplifier output impedance to
the loudspeaker impedance?
Richard D Pierce
John Busenitz <buse...@bobeck.ecn.purdue.edu> wrote:
>> In fact, as Shankar Ramakrishnan points out, in a proper
>>loudspeaker system, the system bass damping factor is on the
>>order of unity, usually a bit lower. This is the system damping
>>factor, mind you, and assumes a very low source impedance. The
>>DC series resistance of a typical 8 Ohm woofer is on the order of
>>5 Ohms, and even with zero amplifier output impedance, the
>>amplifier will never "control" the cone motion.
>
>Would it then be more correct to use the term "damping ratio"
>to refer to the ration of the amplifier output impedance to
>the loudspeaker impedance?
A ratio is, in essence, a factor, so you're playing with, at best,
semantics that do little to clarify the point.
The real meat of the matter is that the assertion by those espousing
damping factor as an important number is that, somehow, damping factor is
a measure of the amplifier's ability to control the transient response of
the loudspeaker system. That assertion is provably false and patently
absurd, the strident and ill-informed protestations of proponents of the
concept to the contrary notwithstanding.
The ability of a woofer cone to stop oscillating is essentially due to
the ratio of the total motional reactance (the energy-storage part of
the system) to total resistances (the energy dissipating part of the
system).
Let's ignore, for the moment, mechanical losses (friction in the surround,
for example) and acoustical losses (absorbtion, the resistive portion of
the radiation impedance, etc.), since 1) they are NOT affected by the
amplifier or any other electrical considerations and 2) they are small
compared to the electrical losses anyway. This measure of the ratio
between the motional reactance and the losses has a name, it's called
"Q", and this is a measure of the transient response of the system.
So let's say we have a system that has a motional reactance of Xw and a
total electrical loss of R. What is the relative effect of changing the
two loss mechanisms under discussion, the amplifier output resistance and
the voice coil resistance?
Well, the electrical loss R has two parts: Re, the voice coil resistance
and Rg, the amplifier output resistance. So the expression of the system
Q now becomes:
Q prop Xw (Re + Rg)
So, indeed, changing Rg WILL change the Q, and thus the transient
performance of the system. But by what magnitude? Well, that depends upon
the relative magnitudes of Re and Rg. And it is here the falacy of damping
factor as a usable measure of transient behaviour is revealed.
Re, as mentioned elsewhere, is several ohms. The typical value I have seen
for "nominal" 8 ohm driver is on the order of 6.5 ohms (5 Ohms, as
mentioned elsewhere, is unusually low, but not unheard of). So let's pick
6.5 ohms.
Rg, the amplifier output resistance, is, of course, stated by damping
factor. Let's look at some extreme numbers, a damping factor of 50,000 vs
a damping factor of 50, a difference of 1000:1. Damping factor proponents
would have us believe that such a ratio will lead to tremendous
differences in transient behaviour, "damping" of the loudspeaker system.
However, this is simply not so.
The difference in the TOTAL losses is, remember, Re + Rg. In one case,
with a damping factor of 10,000, Rg = 0.00016 ohms, while in the case with
a damping factor of 50, Rg = 0.16 ohms. Big difference, yes?
No, because it is the TOTAL loss that is important. In one case, the total
loss is equal to 8 + 0.00016 = 8.00016 vs 8 + 0.16 = 8.16. Now, the
difference is NOT on the order of 1000:1 or 100,000%, but 8.16:8, or all
of 2%!. Where did the tremendous advantage of a damping factor of 50,000
disappear to? Well, it crawled back into the hole of uselessness where it
belongs.
Well, you might argue, that 2% MIGHT be important! It might, except for
the fact that that 2% variation in losses, resulting in a 2% variation in
Q, is smaller BY A LOT than the the production allowances on voice coil
resistance, suspension losses and compliance, cone mass, and variation in
all of these dues to normal and SMALL changes due to temperature,
humidity, pressure and aging.
Call it damping factor, call it damping ratio, call it whatever you like.
But realize that as a figure of merit, it measures nothing, reveals
nothing per se about the transient performance of the combined
amplifier/loudspeaker system, which is what is claimed is the point to it
all.
jj, curmudgeon and all-around grouch
Well, it's not that simply, unfortunately.
First, many "audiophile" amplifiers have little if any feedback, but they
are the same market as unobtanium cables.
Second, many speaker systems have what one might call "remarkable" impedence
characteristics, that would couple with cable resistance to provide some
really very interesting new poles in the feedback loop. (I note that many
power supplies have very complex compensation in order to ensure stability,
and compensation that results in nothing NEAR a flat frequency response, while
power supplies very often face a more polite and less reactive load than
power amplfiers.
Third, in order for remote sensing to be useful, one must concede the use
of high-feedback designs.
Now, personally, I've got no problems with high-feedback designs that are
done right, I have no use for bizzare and archaic tube amplifiers that
go back to before the dawn of symmetryic amplification, and so on,
but I'm not your market, the audiophile is your market, and he (very rarely
she) may hold very strong, unsupported opinions about symmetry,
feedback, or wiring.
--
Copyright alice!jj 1995, all rights reserved, except transmission by USENET and like facilities granted. Said permission is granted only for complete copies that include this notice. Use on pay-for-read services or non-electronic media specifically disallowed.
j...@alice.att.com Member HASA - Atheist Scum Division, curmudgeon caucus - Numquam Nihil Preparandum
Mark S. Gahr
: In article <dibaldDC...@netcom.com>,
: David I. Baldwin <dib...@netcom.com> wrote:
: >From standard wire tables, ten feet of 16 gauge speaker cable (20 ft. of
: >wire total) has .08 ohms of resistance, and ten feet of 12 gauge has .032
: >ohms of resistance. On simple terms, if your amp has .01 ohms output
: >'resistance', the wire dominates the damping.
: Wrong, if you amp as 0.1 ohms output resistance, your cables have .08
: ohms resistance, and your speaker's voice coil has 6.5 ohms resistance,
: the voice coil dominates the damping. If fact, this is true of pretty
: much every situation, save some very rare pathological cases.
: >Connectors also add
: >resistance. With 8 ohm speaker load and 0 ohm wire, the damping factor
: >would be 800; with the 16 gauge wire the damping drops to 88. With 4 ohm
: >load, the damping would be 44. This is a simplified calculation, a
: >number of things are left out that reduce the effective damping even more.
: Yeah, like the DC resistance of the voice coil, which has been shown time
: and time again to be the dominant resistance in the entire chain (see
: Small, et al).
: --
: | Dick Pierce |
: | Loudspeaker and Software Consulting |
: | 17 Sartelle Street Pepperell, MA 01463 |
: | (508) 433-9183 (Voice and FAX) |
You win, Dick. But have I missed something in this thread? No one has
even mentioned (I _could_ be blind!) that damping factor is another
one of those frequency related topics. These devices are not purely
resistive.
Food for thought if you're not very hungry,
Mark S. Gahr
Integrated Audio Systems
Sound & Stage Supply
Mark S. Gahr
: Subject: Re: Damping Factors
: From: Gerral Hubbard, ger...@fastlane.net
: Date: 22 Jul 1995 13:05:22 GMT
: In article <3uqt2i$o...@dfw.nkn.net> Gerral Hubbard, ger...@fastlane.net
: writes:
: >Here's a senario,
: >
: >I've got two speakers (each is a two way box using an internal passive
: >crossover) and they show an 8 ohm impedence.
: >
: >1. If I put one speaker on each side of an amp each side would see 8
: ohms.
: Assuming you have a stereo amp, yes.
: >2. If I were to use a cross over and bi-amp the two speakers, using the
: same
: >amp and running the high ends on one channel and low ends on the other
: it >would show the amp a 4 ohm load on each channel.
: Would it? How did you figure that out? It looks like you're saying 4
: ohms is equal to 2 x 8 in parallel (which it is), but how do you get 8
: ohms for the woofer on its own if the resultant impedance of the box is
: also 8 ohms? It boggled my brain trying to figure out how that wold work.
: >Advantages of bi-amping aside; I would have a better sound because of
: the
: >higher damping factor (using 4 ohms rather than 8 ohms).
: >Does this sound alright?
: Not to me. Apart from the fact that the sound would be in mono (unless
: you're getting another stereo power amp), I always thought that higher
: speaker impedances tended to give rise to better damping by the
: amplifier; more impedance gives it "more to hold onto", in a way. I've
: only seen amplifiers' damping factors go down when using lesser ohmages.
: Unless I've totally misunderstood the question I'd be tempted to forget
: the whole thing.
: /Dean
Of course, putting another layer of reactive components between the
driver and amp may well be worse for the damping factor overall than the
four-ohm load.
BTW, thanks to everyone for saying it right. I have a friend (good mix
engineer, too) that I always quip with "It's not a measure of how wet the
amplifier is".
tINY
>
>I haven't been following this thread, but will take the risk
>(of repeating common knowledge). This whole issue
>of needing special thick low resistence, low inductance
>speaker wires would be almost irrelevant IF ONLY the
>power amplifier employed REMOTE SENSING, like
>all reasonable power supplies do.
Sounds like a great idea. I'll build a kelvin connection circuit
into the active x-over I've been planning on for a while.
Look for a report (and maybe a web site for schematic) in a few
months.
-tINY
John Byrns
wrote:
> Unfortunately, this is a misleading oversimplification, albeit
> a very common one. The implication is that by increasing the
> damping factor to infinity, the amplifier will perfectly
> "control" the motion of the loudspeaker cone, which will follow
> exactly the voltage variations at the amplifier output terminals.
> This is not at all what happens in real life.
>
>--- Snip, Snip ---<
>
> -Henry
Couldn't say away? Welcome back Henry.
Regards,
John Byrns
James Adrian
It's all been done before - there have been several amplifier designs
that utilise some form of four terminal sensing or current feedback to
improve the damping factor etc. One interesting example is the
Delta-Omega version of the M600/7560 manufactured by Crown International
- this didn't use four terminal sensing but something more akin to the
velocity feedback found in motor drive amplifiers.
What I want to know is why don't loudspeaker manufacturers look into
tailoring their loudspeakers performance and response to constant-current
(transconductance) drive. In vibration research it is common practice to
drive the *shakers* in constant-current mode. The theory behind this is
that the strength of the magnetic field is directly proportional to the
current flowing through the wire/coil producing it and not to the voltage
applied across it (remember physics at school?). This overcomes the
effects of many of the dynamic and heat variation problems associated
with voltage drive.
Now don't get me wrong - it's not without it's problems external acoustic
loading becomes a factor and it works best/is more practical at low
frequencies - but it's got to be worth a try.
Over to you manufacturers (unless I get there first!)
Adrian
Arsenio Novo
Re: Damping Factors (27 Jul 95 18:34:30)
Dp> So let's say we have a system that has a motional reactance of Xw and
Dp> a total electrical loss of R. What is the relative effect of changing
Dp> the two loss mechanisms under discussion, the amplifier output
Dp> resistance and the voice coil resistance?
Dp> Well, the electrical loss R has two parts: Re, the voice coil
Dp> resistance and Rg, the amplifier output resistance. So the expression
Dp> of the system Q now becomes:
Dp> Q prop Xw (Re + Rg)
Dp> So, indeed, changing Rg WILL change the Q, and thus the transient
Dp> performance of the system. But by what magnitude? Well, that depends
Dp> upon the relative magnitudes of Re and Rg. And it is here the falacy of
Dp> damping factor as a usable measure of transient behaviour is revealed.
Dp> --
Dp> | Dick Pierce |
Dp> | Loudspeaker and Software Consulting |
Dp> | 17 Sartelle Street Pepperell, MA 01463 |
Dp> | (508) 433-9183 (Voice and FAX) |
The two resistances of which you speak are a certain distance apart
temporally. This means at some moments they may even effectively have
opposite signs and partially cancel one another. This may also be
partly aided by the anti-resonant design of the enclosure.
The result is that the total instantaneous system resistance may even
come close to zero. Thus at that particular instant in time one might
in essence observe a nearly totally reactive impedance with close to
zero ohms effective resistance.
This then is the moment the lower damping factor will make a
difference. The nearly solely reactive impedance will draw large
currents particularly if also under resonant conditions. The voice
coil resistance alone won't limit the instantaneous current drawn by
the speaker system.
It seems to me you are assuming a steady-state response which is the
usual simplification. Thus I believe you are neglecting to take into
account the time variable effects on the total effective system
resistance.
Arsenio
... 186,000 miles/sec: Not just a good idea, its the LAW.
~~~ ReneWave v1.00.wb2 (unregistered)
--
| InterNet: Arseni...@mba.org
| Standard disclaimer: The views of this user are strictly their own.
Richard D Pierce
Arsenio Novo <Arseni...@mba.org> wrote:
> Dp> Well, the electrical loss R has two parts: Re, the voice coil
> Dp> resistance and Rg, the amplifier output resistance. So the expression
> Dp> of the system Q now becomes:
>
>temporally. This means at some moments they may even effectively have
>opposite signs and partially cancel one another. This may also be
>partly aided by the anti-resonant design of the enclosure.
Huh? WHere did THAT come from?
The two resistances are separated by a wire, over which the signal
travels at the speed of light. They are also RESISTANCES, in the vats
majority of amplifier designs, the output resistance is ALWAYS positive
(the exception is specially designed amplifiers that have negative output
resistances, none of which exist on the market today). The resistance of
the voice coil IS ALWAYS POSITIVE, there are NO EXCEPTIONS to that!
>The result is that the total instantaneous system resistance may even
>come close to zero. Thus at that particular instant in time one might
>in essence observe a nearly totally reactive impedance with close to
>zero ohms effective resistance.
>
>This then is the moment the lower damping factor will make a
>difference. The nearly solely reactive impedance will draw large
>currents particularly if also under resonant conditions.
The resonance is a parallel resonance, so it's impedance is at a maximum,
NOT a minimum at resonance.
>The voice
>coil resistance alone won't limit the instantaneous current drawn by
>the speaker system.
SHow us how.
>It seems to me you are assuming a steady-state response which is the
>usual simplification. Thus I believe you are neglecting to take into
>account the time variable effects on the total effective system
>resistance.
It seems to me that your are making some assumptions that are utterly
unsupported by reality. Sure, the voice coil resistance is time variant,
because it is temperature variant, and the tempco of all materials used
for voice coils is, indeed, positive.
There is not a single shred of evidence suggesting any of the mechanisms
you are suggesting exist in loudspeakers, nor is there any theoretical
model that predicts these effects.
--
| Dick Pierce |
| Loudspeaker and Software Consulting |
| 17 Sartelle Street Pepperell, MA 01463 |
Henry Pasternack
: Would it then be more correct to use the term "damping ratio"
: to refer to the ration of the amplifier output impedance to
: the loudspeaker impedance?
Probably it would be better just to quote output impedace at
midband. Or, if you want big numbers, use output conductance
instead.
-Henry
Kurt Strain
about "damping" for the RM-9. The three different gain settings have
the following published characteristics:
POS. GAIN DAMPING FEEDBACK INPUT FOR 100W
---- ---- ------- -------- --------------
LO 27dB 0.7dB 19dB 1.2V
MED 32dB 1.5dB 14dB 0.8V
HI 36dB 2.5dB 10dB 0.4V
0.7 dB damping? What does that mean? If it were the ratio of speaker
impedance to output impedance, then:
0.7dB = 20*log(DF) ==> damping factor = 1.08. Likewise, for 2.5dB damping,
damping factor = 1.33. That's amazingly low, if true. I don't buy that.
Is this not referring to the simple definition for damping factor, and
is instead a ratio of something else, like ratio of different output
impedances without regard to speaker impedance? It's a tube amp, BTW.
Kurt
Richard D Pierce
Arsenio Novo <Arseni...@mba.org> wrote:
>Mr. Pierce was describing the coil resistance in series with the
>reflected equivalent mechanical resistance however in reality the
>voice coil is not constructed as the simplified system model would
>suggest.
>
>The voice coil's resistance is actually in parallel with the other
>distributed impedances which include all the reflected equivalent
>mechanical components and motor emf (this latter is further affected
>by magnetizing currents, the pole piece BH curve, and it's resultant
>hysterisis).
Wrong. Absolutely, positively, provably dead wrong. The driver's DC
resistance is, in ALL cases, in series with the relfected mechanical
equivalents. Let's look at the implications of your assertion, Arsenio.
Here's what you propose as the electrical equivalent of the system, for
the simplest case, the fundamental mechanical resonance:
o--------+-------------+
| |
| +-----+-----+
| | | |
Re Lces Cmes Res
| | | |
| +-----+-----+
| |
o--------+-------------+
Versus the "standard" model that I propose, that Small proposes (fig
4, "Direct-Radiator Loudspeaker System Analysis," JAES June 1972), that
Theile proposes (fig 4, "Loudspeakers in Vented Boxes," JAES May 1972),
that Locanthi proposes (figs 2, 8, 10, and 18, "Application of Electrical
Circuit Analogies to Loudspeaker Design Problems," JAES 1971 Oct), that
Beranek proposes (fig 7.2c&d, "Acoustics"), that Kinsler proposes
("Fundamentals of Acoustics" 'Transduction'), that even MacLachlan
proposes ("Loudspeakers," 1954), which looks like:
o------------+
|
Re
|
+-----+-----+
| | |
Lces Cmes Res
| | |
+-----+-----+
|
o------------+
These are two VERY different models that make some specific predictions of
behaviour. Your model, when analyzed, predicts that the impedance of the
driver at DC is dominated by the equivalent inductance of Lces (the
electrical inductive equivalent of the suspension compliance) which leads
to an impedance of 0. At infinite frequency, it's dominated by Cmes (the
electrical capacitive equivalent of the cone mass) which leads to an
impedance of 0. At the resonant frequency, proportional to the reciprocal
of the square root of the product of Cmes and Lces, it's dominated by the
parallel combination of Re (the DC resistance) and Res (the electrical
equivalent of the suspension losses).
So that in a typical 6 1/2" woofer, we should measure the following three
impedances:
DC resistance 0 ohms
At resonance 6.5 ohms || 32 ohms = 5.40 ohms
At inf freq 0 ohms
Whereas the model I and others propose is dominated at DC by Re (that's
uh, the DC resistance of the voice coil). At infinite frequency, it's
dominated by the DC resistance of the voice coil. At resonance, it's
dominated by the series combination of Re and Res.
So, in THIS case, we should measure the following three impedances:
DC resistance 6.5 ohms
At resonance 6.5 ohms + 32 ohms = 38.5 ohms
At inf freq 6.5 ohms
So much for competing theories, what does REALITY tell us?
Well, in fact, we can't measure infinite frequency, so let's, for the
moment, look at the first two points.
Indeed, when we measure impedance BY ANY ACCEPTED MEANS YOU CAN PROPOSE,
be it steady state, transient, broadband noise and filtering, whatever,
we find the following:
DC resistance 6.5 ohms
At resonance 38.5 ohms
Arsenio, your theory is dead: it utterly fails to predict anything
remotely resembling reality.
>This why it seems to me the simplified model tends to break down when
>one needs to consider the effects of the amplifier's damping factor.
But your model suffers from an even greater, far more deadly flaw: it's
completely wrong in that it fails to predict what the impedance of the
driver is. So of what further use could it be? None.
>For the two equivalent resistances in question have an additional time
>dependent relationship. In the case of the simplified transformer
>model the mutual coupling between primary and secondary windings is
>practically instantaneous since in essence it operates roughly at the
>speed of light.
B*llsh*t! it's limited by the bandwidth of the transformer, which is
dependent upon the equivalent inductance and the parasitic capacitance.
>Not so for the loudspeaker system in which this time difference is on
>the contrary appreciable. Consider that at best the piston displaces
>air at less than the speed of sound which is but a mere fraction of
>the speed of light after all.
B*llsh*t! it's limited by the badwidth of the driver system, which is
dependent upon the system mass, compliance, radiating area and so on.
>So I conclude that the mutual coupling
>between the electrical voice coil and the mechanical system cannot be
>considered instantaneous as the simplified model would tend to suggest.
And you conclude wrong. Your model is flawed, you concept of
"instantaneous coupling" is wrong, your tranformer analogy is wrong. All
wrong because they fail to predict even remotely even the most rudimentary
behaviours of real loudspeakers
Go get Small, and run through the analysis. See how the models DOES
predict the beahviour of real systems. Go understand the fallacy of your
"time dependent resistances." Then try again.
Or, if you think you're right (and, who knows, you MIGHT be), then
provide us with something other than a vague theory based on incorrect
notions of instantaneous this and time dependent that. Provide us with
supporting citations from the literature. Provide us with actual evidence
supporting the predictions of your model. Then be prepared to defend that
model in the face of overwhelming evidence to the contrary. I have about
3,000 impedance curves, measured with both steady-state and transient
methods, not a single one of which behaves even remotely as you predict.
Arsenio Novo
Re: Damping Factors (28 Jul 95 23:49:26)
Ya> From: yat...@babbage.csee.usf.edu (Randy Yates)
Ya> Newsgroups: rec.audio.pro,rec.audio.tech
Ya> Subject: Re: Damping Factors
Ya> Organization: University of South Florida
Ya> <3v6a8d$i...@bmtlh10.bnr.ca>
Ya> In article <3v6a8d$i...@bmtlh10.bnr.ca>, hen...@bnr.ca says...
Ya> Here is a theory that attempts to reconcile the phenomena in question:
Ya> Granted that even an amplifier with infinite damping factor will not
Ya> cause a loudspeaker cone to follow exactly the voltage output.
Ya> . . . So finite amplifier damping
Ya> factors and speaker back emfs are indeed things that degrade a
Ya> system's response, and the elimination of the former removes the
Ya> problem of the latter.
Ya> . . .
Ya> % Randy Yates % "...the answer lies within your soul
Ya> % EE/Mathematics Student % 'cause no one knows which side
Ya> % University of South Florida % the coin will
Back emfs aren't what degrades the system's response. They are in
fact intrinsic to the speaker's operation. The fundamental question
of "what is a speaker" can be answered simply by "it's a transducer".
A "transducer" is a device intended to couple energy efficiently
between two mediums via an impedance adaptation or transformation. To
use an analogy, it is somewhat akin to a simple electrical a.c.
transformer where the secondary winding's characteristics reflect back
into the primary and vice-versa via a magnetic mutual coupling.
However in the case of the loudspeaker the secondary is a mechanical
system consisting in its simplest form of a moving piston and an air
mass. The equivalent electrical counterparts of that mechanical
system are reflected (although via motor action) back into the voice
coil (or primary) just like in the transformer analogy.
Mr. Pierce was describing the coil resistance in series with the
reflected equivalent mechanical resistance however in reality the
voice coil is not constructed as the simplified system model would
suggest.
The voice coil's resistance is actually in parallel with the other
distributed impedances which include all the reflected equivalent
mechanical components and motor emf (this latter is further affected
by magnetizing currents, the pole piece BH curve, and it's resultant
hysterisis).
This why it seems to me the simplified model tends to break down when
one needs to consider the effects of the amplifier's damping factor.
For the two equivalent resistances in question have an additional time
dependent relationship. In the case of the simplified transformer
model the mutual coupling between primary and secondary windings is
practically instantaneous since in essence it operates roughly at the
speed of light.
Not so for the loudspeaker system in which this time difference is on
the contrary appreciable. Consider that at best the piston displaces
air at less than the speed of sound which is but a mere fraction of
the speed of light after all. So I conclude that the mutual coupling
between the electrical voice coil and the mechanical system cannot be
considered instantaneous as the simplified model would tend to suggest.
Arsenio
arseni...@mba.org
... The Idea is the immortal virus.
John Byrns
> I am requesting an explanation for the strange spec Music Reference has
> about "damping" for the RM-9. The three different gain settings have
> the following published characteristics:
>
> POS. GAIN DAMPING FEEDBACK INPUT FOR 100W
> ---- ---- ------- -------- --------------
>
> LO 27dB 0.7dB 19dB 1.2V
> MED 32dB 1.5dB 14dB 0.8V
> HI 36dB 2.5dB 10dB 0.4V
>
> 0.7 dB damping? What does that mean? If it were the ratio of speaker
> impedance to output impedance, then:
>
> 0.7dB = 20*log(DF) ==> damping factor = 1.08. Likewise, for 2.5dB damping,
> damping factor = 1.33. That's amazingly low, if true. I don't buy that.
Just a guess, but maybe it refers to the change in output voltage,
expressed in dB's, between no load and the rated load. I am doing this in
my head, so my math may be off, but that would make your "damping factor =
1.33" be a damping factor of 3. The "damping factor = 1.08" case is too
hard to do in my head, so I will leave it as an exercise for the student,
It's a bigger number in any case. This is just a guess as to what it might
mean.
Regards,
John Byrns
Bill Claussen
Ok guys,
For simplicity, lets forget about the cable resistance, and just look
at the philosophy of Damping Factor. I know most manufactures like to
state a nice big number for damping factor, but it's rather mis-leading.
The usual formula that most people use is the speaker impedance (lets
use 8 ohms in this case) divided by the internal output impedance of the
amplifier(lets say .1 ohms). So, using the formula:
Zs (speaker impedance) 8
---- = --- = 80 (damping factor)
Zout (amplifier output impedance .1
The above seems to be the most popular, probably because of the neat
big numbers you get......but it is wrong.
If you figure the True damping factor, you also have to include the
voice coil resistance(in this case, 6 ohms). So, using this new formula:
Zs (speaker impedance) 8 True
---------- = ------ = 1.31 (damping factor)
Zout + Rvc (voice coil resistance) .1 + 6
As a matter of fact, the absolute maximum damping factor you could get with
a 8 ohm speaker with a 6 ohm voice coil resistance and an amplifier output
impedance of 0.000000000000 ohms (infinity) is 1.33.
So, from what I read in the previous post, it seems that the RM-9 amplifier
states the True damping factor, not the fictitious one.
Comments???
Bill Claussen
PS. Gee, Mr. Peter Larson, isn't it amazing that some people CAN use
e-mail, and the internet, and spell correctly....all at the same time! WOW
Henry Pasternack
: For finite damping factor, the amplifier output voltage will vary from the
: ideal depending on the effective load impedance (effective implying that the
: instantaneous load impedance depends on such things as back emf). So finite
: amplifier damping factors and speaker back emfs are indeed things that degrade
: a system's response.
It's important to remember that "back emf" is just a result of electro-
mechanical energy storage in the speaker driver. Looking into the speaker
terminals, this appears as a reactive component to the amplifier. The ability
of the amp to drive a reactive load is important. But there is nothing
magic about "back EMF", despite weird claims to the contrary by audiophile
pundits. I once had a long argument here with a fellow who claimed a feed-
back amplifier gives better damping of a large woofer than an ideal voltage
source without feedback, or a dead short circuit. He claimed to have done
measurements, but, of course, he was wrong. Never admitted it, though, but
I learned a lot from the argument.
-Henry
James Adrian
>with infinite damping factor is simply an ideal voltage source.
Not quite, but I did suggest that people should look into
transconductance-mode amplifiers instead of relying on
constant-voltage-mode to drive what is - in essence - a current dependant
motor and that's not so far of the mark!
:-)
Adrian.
Richard D Pierce
John Byrns <ate...@email.mot.com> wrote:
>
>But couldn't we use feedback to create a negative resistance source which
>would cancel the DC resistance of the voice coil, yielding higher damping?
Absolutely. In fact, A. Neville Thiele (of Thiele/Small fame) made just
such a proposal in 1961. Negative output impedance amplifiers, in fact,
WILL circumvent the problems associated with the fact that the voice coil
DC resistance is THE dominating loss factor in loudspeakers.
But...
>Just a quick thought that may be all wet, but it poped into mind as I read
>your posting, as somthing that an active feedback circuit can do that a
>perfect voltage source can't. I'm not a speaker expert though, so don't
>know if this would help the damping. And of course I am not sure why we
>would want to increase the damping anyway, from what little I know about
>speaker theory.
We would want to increase the damping if it is not already adequate, and
that is a problem with certain VERY SPECIFIC systems, NOT WITH
LOUDSPEAKERS IN GENERAL. Loudspeaker designers have PLENTY of options
available to them to achieve optimal damping at resonance (the only place
where it's going to make any difference electrically) using normal
amplifiers without having to resort to "exotic" (at least as far as the
market is concerned) tricks such as negative output impedance.
If fact, if a speaker has been designed (and it's not rocket science to do
so) to have optimum (as defined by the designer) damping using
conventional amplifiers, it's certainly arguable that the system will be
as badly misaligned using it with an amplifier output impedance of -2 ohms
(it will now be overdamped) as with an amplifier with an output impedance
of 2 ohms.
The fact is that the vast majority of speakers are designed to work with
conventional, low output impedance amplifiers (at least, those that are
designed in any way to begin with). Small changes in source impedance,
whetherr they are positive or negative, whether they are due to wires or
what, have negligable effects on the transient performance of the system.
Introduce an amplifier with a large negative output impedance, and you
effectively obsolete every loudspeaker on the market. Those that it might
rescue probably have far more serious problems anyway. If the system
works NOW with optimum system damping (and many, many do), what problem
does negative output impedance fix? If the designer has achieved a system
Q of 0.707 in a closed box using a conventional amplifier, what problems
exists that negative output impedance fixes?
C.M. Hicks
>John Byrns <ate...@email.mot.com> wrote:
>>
>>But couldn't we use feedback to create a negative resistance source which
>>would cancel the DC resistance of the voice coil, yielding higher damping?
In our local newspaper this evening I spied, in the "For Sale" column
for a
Modern Upright Piano, Overdamped, #950
...well, it made me chuckle!
Christopher
--
=======================================================
Christopher Hicks http://www.eng.cam.ac.uk/~cmh
c...@eng.cam.ac.uk Voice: (+44) 1223 3 32767
=======================================================
John Byrns
wrote:
> Randy Yates (yat...@babbage.csee.usf.edu) wrote:
But couldn't we use feedback to create a negative resistance source which
would cancel the DC resistance of the voice coil, yielding higher damping?
Just a quick thought that may be all wet, but it poped into mind as I read
your posting, as somthing that an active feedback circuit can do that a
perfect voltage source can't. I'm not a speaker expert though, so don't
know if this would help the damping. And of course I am not sure why we
would want to increase the damping anyway, from what little I know about
speaker theory.
Regards,
John Byrns
BiasBob
power amp, which would improving damping, I would think. Very important
is the guage of the wire, subjectively more that one would think. I used
0000guage wire to the
woofers[solder connections], and the difference was very noticable over 12
guage.
TAD woofers, bryston pwr amps.
Lon Stowell
>Absolutely. In fact, A. Neville Thiele (of Thiele/Small fame) made just
>such a proposal in 1961. Negative output impedance amplifiers, in fact,
>WILL circumvent the problems associated with the fact that the voice coil
>DC resistance is THE dominating loss factor in loudspeakers.
Was this the basis for the "ACE Bass" used in some of the early
80's powered speakers and subwoofers? I could swear the term
ACE Bass came from Audiopro, but we are talking a decade or better
ago memory. I do remember some pretty impressive bass levels and
extension from some rather small drivers.