Damping Factor - Revisited

[damacman]

So, let's all learn something (me included). I have, like many of you, developed a preference for a certain brand of amplifier when using it to drive subwoofers. For me, that brand has been Crown. It just seems like the Crown amplifiers have a control over the drivers that other amps (including McIntosh) can't seem to match.

One would be inclined to draw the conclusion that damping factor is why - the typical Crown having a far greater damping factor than the typical McIntosh. Over the years, McIntosh (Gordon Gow specifically I believe) has done demonstrations in dealer show rooms to prove this is a fallacy. Unfortunately, I never saw any of those demos - maybe c_dk or twiiii have.

So, is damping factor the reason or is it something else? After all, a moving loudspeaker obviously exerts back EMF on the output section of the amplifier driving it . . .

A really good friend of mine owns a company at which he performs competitive analysis on power amplifiers for a living. Once a firm believer in damping factor, he now laughs at the idea. His logic is sound. First, consider the formula for damping factor:

DF = Load Impedance / Amplifier Output Impedance

[Refer to this white paper from Crown if you like - http://www.crownaudio.com/media/pdf/amps/damping_factor.pdf ]

So, if we've got the typical 8 Ohm speaker, you can see that DF improves as Amplifier Output Impedance decreases. Consider the following examples:

MC2300 - According to the manual, Damping Factor is 14 at the 8 Ohm Output tap therefore the output impedance at the 8 Ohm tap would be .57 Ohms.

Crown MT2400 - According to Crown, Damping Factor is "better than 1000 from 10 Hz to 400 Hz", therefore the output impedance would be .0008 for those 5+ octaves. [It is worth noting that Crown doesn't specify which impedance this correlates to.]

In my listening experience, the Crown definitely seems to have a distinct edge over the McIntosh when driving subwoofers. But why? Let's follow my friend's logic now. According to him, one has to factor DCR of the load to compute the actual damping factor. This certainly makes sense.

Let's consider the subwoofers I use (two 18" with Dual 4 Ohm VCs). According to my Fluke 87 V DMM, the DCR of a single woofer with both coils in series (8 Ohms) is 7.0 Ohms.

Let's assume the DCR of the speaker wire is negligible (it really isn't) - how can the DCR of the voice coil(s) of the load be ignored? If we use DCR as a factor in computing DF, then it's a wash:

MC2300 - (.57 Ohms + 7.0 Ohms) / 8 Ohms = .95 DF
MT2400 - (.0008 + 7.0 Ohms) / 8 Ohms = .88 DF

Obviously, Damping Factor must change with frequency as the impedance of the load is constantly changing with frequency. I actually do have an impedance meter and can tell you the exact impedance of these woofers at any frequency between 20Hz and 400Hz . . . but, is that even necessary here?

I have more to share, but first - let's hear from some of the gurus. I'm as eager as anyone to get to the bottom of this.

 

[Beowulf62]

There are many factors that influence the damping of a driver apart from the output impedance of the amplifier. For example, my 15" JBL driver has a 13 lb magnetic assembly and a flux density of 12,000 gauss. This provides inherently good damping. The driver is mounted in a folded back-loaded corner horn and this also provides damping as the cone is coupled to the mass of the air in the room very effectively. I use tubes amplifiers with negative feedback that do not have a "very" low output impedance and this sounds just right. If I use a transistor amplifier with a lower output impedance, the sound becomes dry and too "tight" for my taste. However, that may be just what another listener prefers. Past a certain point, increasing damping is not better, I believe it is simply a matter of preference.

 

[ConradH]

This argument/analysis has been stated for many decades, but doesn't seem to match actual practice. Maybe it's all in our heads! Being part of the "if you can hear it, I can measure it" club, I'd find it interesting to probe the waveforms at the subwoofer and even look at the acoustic output. Maybe use tone bursts at various frequencies, or a simulated drum impulse. Once you prove that there's a voltage difference between the amps for the same signal, it should be easy to track down the cause.

 

[dewickt]

As I have had it explained to me to get the best sound with a high damping factor anplifier you need a low compliance speaker and the reverse with a low damping amplifier a high compliance speaker. This seems to make sense and in practice is borne out by the results, a speaker that tends to boom is usually low compliance, subs and ported bass reflex, like high damping control. The problem as I have found is that once the damping is over 10, or so, it is enough to control 'most' speakers, thus the numbers become meaningless.

 

[robgmn]

What type of speakers do those amps usually drive (acoustic suspension, PA, ported, etc.)?

 

[damacman]

In my application, the enclosures for the woofers are vented. Like Beowulf and Terry allude, I believe enclosure design and the type of driver are larger factors than DF.

ConradH - you're thinking is sound. I see Grommet has been reading up . . .

 

[Beowulf62]

A bit of a read...but here are the basics:

Damping & Q
Before we define Q, let’s think about a car again. All automobiles have a suspension which prevents bumps and dips in the road from jarring the passengers. However, a car cannot use a spring suspension by itself because the car would continue to bounce (resonate) after each bump or dip. The bouncing needs to be suppressed and this is what the shock absorbers or dampers do. They suppress the unwanted bouncing after the springs have absorbed the initial impact of the bump or dip. In similar fashion, a speaker driver needs to be damped. Otherwise, its diaphragm will tend to vibrate excessively at its resonance frequency.

Several things serve to damp the driver. These include the suspension (surround and spider), motor (voice coil and magnet circuit) and amplifier output resistance. Achieving an optimal amount of damping can sometimes be a challenge and it almost always requires a consideration of the type of box in which the driver will be used. If the driver is not designed properly, it may be underdamped or overdamped. An underdamped driver will have ripples in its response and it will reproduce transient signals poorly. An overdamped driver will have a reduced low-frequency response.

"Q" is the term used by speaker designers and engineers to describe how well damped a driver is. However, Q is not damping. Instead, it is resonance magnification which is the exact opposite of damping. So a higher amount of damping results in a lower value of Q and visa versa. Let’s return to our earlier graph (see attachment below)

The driver which was used to create this graph has a total Q (Qts) that is super low at only 0.19. This low Q value results primarily from an extremely strong magnet and this driver by itself would be considered severely overdamped. To get a maximally flat response out of it we had to counteract some of the damping by stiffening the overall compliance. We did this by putting the driver in a relatively small box having an air volume with a much smaller compliance than the driver suspension. This caused the overall compliance to decrease considerably and the total Q (Qtc) to increase to 0.7 as shown in the green curve in the graph. A Q of 0.7 is considered by many to be the ideal value because it balances driver damping with a smooth low-frequency response and a reasonably good transient response.

The red curve was created with a huge box whose volume was over 100 times larger than the maximally flat box. As a result, the box has a very high compliance and so it could not stiffen the compliance of the driver much at all. In fact, the total Q was 0.2 which is just slightly greater than the driver, itself. Because the total Q is allowed to remain so low, the design is considered overdamped. Notice how it produces less low-frequency amplitude between 95 and 400 Hz. You might consider its greater output below 95 Hz to be an advantage but this is probably not the case because the driver would have to move a tremendous amount of air at these low frequencies and its maximum excursion (Xmax) may not be large enough to handle this. If it is not, it will at best have higher distortion and at worst become damaged.

The blue curve was created with an ultra small box whose volume was over 25 times smaller than the maximally flat box. As a result, the box has a very low compliance and so it stiffened the compliance of the driver by a very large amount. In fact, the total Q was 3.0 which is quite high and causes the box to be underdamped. Notice how it produces a 9 dB response peak that is centered at 580 Hz. The peak occurs at this frequency because the low compliance of the box has in effect shortened the spring, raising the resonance to 580 Hz. The peak is 9 dB high because the driver lacks sufficient damping to overcome its resonance. This means that the driver will ring like a bell at 580 Hz. Needless to say, it will also have poor transient response at this frequency.

Note: The compliance of the driver and box are in series with each other but they sum in parallel like capacitors so the sum is always less than the lowest of the driver or box compliance values. In other words, the lowest compliance dominates the system. With the red curve, the box had a compliance that was much higher than the driver and so the driver’s compliance was dominant. Therefore the total compliance was not influenced much by the box. However, the box of the blue curve had a compliance that was much lower than the driver, and being the lower value it became the dominant one and caused the total compliance to appear to be much lower.

 

[c_dk]

While I had the MC7200 schematic out....dampening is greater than 200, MC7300 greater than 40.

 

[damacman]

c_dk - do you remember any of Gordon Gow's presentations on this topic?

 

[kevzep]

While I had the MC7200 schematic out....dampening is greater than 200, MC7300 greater than 40.

Now this is where it gets interesting, according to the mathematical equation, an amplifier cannot have an effective damping factor of more than about 20 because once you consider the speakers DC resistance, and the cable resistance, all that precious damping factor is ........GONE.....

Crown have their amplifier design philosophy around very high damping factors.....and frankly I can completely hear what they are talking about.....

Now there is a lot of really good info here in this thread in the damping of speakers, but what we are really after is an explanation as to why we (damacman and myself) can hear a more punchy tighter lower and more sustained, yet controlled bass sound from our subs using a Crown (the Macro-tech 5002VZ in my case) and yet when we use another amplifier such as a MC2300, MC352, MC452, MC2205, Carver PM1.5, Sansui (various power-amps) Perreaux, etc etc, that the bass is a bit disappointing compared to the Crown.....I might add that I also had success using a QSC Powerlight 4.0...
The 4.0 has a damping factor of greater than 500 and the Crown greater than 1000.....
It is also interesting to note both the Crown and the QSC are both bridged amplifiers
Now obviously the speaker plays a part in that obviously, but so does the amplifier, and its the amplifiers influence which is what I want to know about....
I gets confusing when you consider the calculations around effective damping factor....

The bass in my system is absolutely astounding, and it is usually the first thing people say about my system when they haven't heard it before, one of my close friends was over yesterday and we were listening to a few tunes at a moderate level, he has heard my system many many times, and he commented again on how fantastic the bass is in my system.....
Of course I want people to say, "wow, that MC452 really sings, which they do, but they comment on the bass first"
Take that Crown away or the QSC away, and that magic bass is gone!! I **** you not!!

Unfortunately amplifier manufacturers have used both sides as marketing tools which further confuses the issue......
I can hear it, but the Maths says I cant? So what is it if it is not the amplifiers damping factor?

 

[whoaru99]

Ya wanted to learn something....

Here ya go: http://www.audioholics.com/audio-amplifier/damping-factor-effects-on-system-response

Damping factor is a potentially meaningful factor, but most likely not in the way you're expecting (until reading the above link).

 

[whoaru99]

Take that Crown away or the QSC away, and that magic bass is gone!! I **** you not!!

Don't doubt it at all.

However, none but the biggest, baddest consumer amps (that most people do not have) compare to even moderately sized pro amps, let alone big guns, in delivering power. I think that's where it's at, not the DF, per se.

I think I posted this once before, comments from the late Brian Cheney of VMPS, regarding the bass performance of a QSC pro amp he used for the low end on CES demo.

This is in regard to a QSC PL236 (the exact one I happen to own, none the less).

...the amp provided detail, definition, extension and dynamic impact which put every other audiophile amp--Class A, AB, C, and D--I have ever tried to shame. Yes that includes big Krells and Levinsons, my Plinius 250/IV, the BBQ, and everything else. Everything else.

He didn't like it full range in one scenario, but did acknowledge running it full range in some other situations with no concerns at all for the full range sound quality.

 

[robgmn]

Kevzep,

Can you list the specs of the Crown and various amplifiers that you have compared against the Crown?

I'm thinking about the amp in my sub and if DF would matter a whole lot.
It's an isobaric setup and the woofer will hardly budge if you try to manually push it in.
Not much slop to allow for any feedback there...

 

[twiiii]

I didn't read all the words, but all you guys are driving your speakers directly, no passive crossovers. All the papers written about damping factor from Mac assume we have passive crossovers. The passive crossover puts a real wedge in getting optimum performance form a woofer that isn't critically damped on its own. If we had perfectly controlled woofers we wouldn't need an extremely high damping factor. Listen to ML-2c with a Crown, a Mac, or a Hafler. They sound very close to each other. You take those same amps and drive my 4 Bozak CG woofers in parallel and it was like "who's been playing with my Crown EQ2 parametric equalizers. Then throw in a 7200 instead of a 7150 and there is another change. So you guys have pretty much got it right as far as I am concerned. A horn speaker may prefer A while a ported speaker C while an infinite baffle B and a air suspension speaker D amplifier. Feed back and how it is used and influenced by the speaker load is a factor too. The lower the internal impedance of an amp the less effect the speaker has. Just look at some of the tests Stereophile does. They use an imaginary reactive load to test an amps frequency response. Amps like Cary tube amps the response wanders +/- 3 db, a 275 less than +/- 1 db, an Accuphase less than 2/10s of a db a 501 around 3/10's of a DB. Well we can hear that difference. And as our speakers are reactive and the enclosure has a great affect on the sound and loading that influences the amp under question,too.

I am 68 now and have been dealing with these issues since I was 17. We are no closer to the answer than we were in the early 70's. Speaker cone materials and adhesives and Macs shorting ring now used by JBL, etc etc has reduced distortion and increased accuracy in the bass. But unless you can afford to have 32 critically damped woofers and thousands of watts in power, having a horn configured properly with the right amp is the only economical solution. If you can afford more power, more space and woofers, ported, reflex seem to be the next logical choice. In my case I choose to give up some dynamic range, have large speaker enclosures, and moderate sized amps with a 200 damping factor. It works for me.

But I would love to try two JBL 15 " woofers in a sealed box flanking an Altec 604 in its latest version From GPA, bi- amped. First with the JBL's replacing the 604 below 80hz and then as addition at 80hz using a 6 db slope as second choice to compare. The enclosure would be a reinforced closet of about 128 cu ft each. Made from brick or cinder block and well padded, no ports. What do you think. One my 207s with 6 channels for the 6 speakers and a Bryston 10 B crossover should work. Its about # 15 on my bucket list. Whats first has 8 wheels. is 45 ft long and built in Canada. But it will have to have a Mcintosh sound inside.

 

[damacman]

Looks like we're all on the same page. I think the charts in the URL whoaru99 provided are quite telling.

My friend, referenced in the original post, attended a seminar at which a speaker was close mic'd and the output of the mic fed to a digital storage o-scope. The amplifier drove the speaker with tone bursts and the actual response of the speaker captured by the mic and displayed on the scope, which included all the decay. The demonstrator then inserted resistors (between .1 and 1 Ohm) between the output of the amplifier and the speaker, to intentionally lower the effective DF, like kev did, and captured the results. The resulting waveforms were compared and said to be identical to one another, no matter the value of resistor being used.

I didn't see this demo in person, but it would be easy enough to replicate. So, if we've ruled out DF, then what is is that causes one amp to sound tighter, more accurate, or faster than another on the bass?

 

[Beowulf62]

With bass frequencies, we are moving a lot of air to get any significant SPL. One can do this with large cones moving a short distance...or smaller cones moving a large distance. However with the latter, distortion increases and as the cone must move further, it needs a higher compliance suspension. This means the cone is not heavily "damped" in its movement. Hence an amplifier with a higher damping factor will be optimal.

Now...if we couple the large cone to a horn it will move a lot of air for a small movement, the driver is "loaded" by a large mass of air and the suspension can be lower compliance. Hence an amplifier with a lower damping factor will be optimal. The rub here is that in order for a horn to reproduce very low frequencies, it must be very long and thus large. One way around this physical size problem is to fold the horn and use the room walls as an extension of the horn, e.g., the Klipschorn is an outstanding example of this design principle.

Do a search for "horn loaded loudspeakers" and you can see numerous images of how large these designs can be. In fact years ago audio nuts (a.k.a. horn lovers, such as myself) built rooms that had enormous horns built into the basements below or walls (how about a 32' horn!) that would energize deep bass in the entire listening space with very little amplifier power...compared to today's "BIG" amplifiers.

 

[c_dk]

c_dk - do you remember any of Gordon Gow's presentations on this topic?

You would have to ask Skizo......he was usually in the audience while I was scurrying around hosting the event.

Gordon most often dealt with the important issues of turntable setup and room acoustics to my memory.

At dinner I would sometimes push him on various topics until Pam would send out stop sign signals when I might go to far.

 

[restorer-john]

So do the big Crowns and QSCs use greater levels of NFB than the 'audiophile' amps? Did the Pro scene ever go down that low TIM path where NFB had to be reduced or did they just push for greater power and control by sticking with high NFB? Greater numbers of output transistors and higher NFB can provide these super high DFs with perhaps more of the speaker's back EMF getting through the feedback loop essentially creating perhaps a more more unified speaker/amp interface?

I think an interesting test would be to take one of the super high DF pro amps and patch in various resistors in series (like the test mentioned), adjusting for level losses over the resistors and use a combination of both low mass and high mass bass driver speakers. I note tonebursts were mentioned in the test, obviously those tonebursts were in the low frequencies where we feel DF has the effects we think we all hear?

Very thought provoking, and if it were a simple exercise to adjust DF on the fly without compromising all other aspects of an amp, we could come to some conclusions very quickly.

 

[Valveglow]

I have, like many of you, developed a preference for a certain brand of amplifier when using it to drive subwoofers. For me, that brand has been Crown. It just seems like the Crown amplifiers have a control over the drivers that other amps (including McIntosh) can't seem to match.

One would be inclined to draw the conclusion that damping factor is why - the typical Crown having a far greater damping factor than the typical McIntosh. O
[Refer to this white paper from Crown if you like - http://www.crownaudio.com/media/pdf/amps/damping_factor.pdf ]

Crown MT2400 - According to Crown, Damping Factor is "better than 1000 from 10 Hz to 400 Hz", therefore the output impedance would be .0008 for those 5+ octaves. [It is worth noting that Crown doesn't specify which impedance this correlates to.]

In my listening experience, the Crown definitely seems to have a distinct edge over the McIntosh when driving subwoofers. But why?

Crown have their amplifier design philosophy around very high damping factors.....and frankly I can completely hear what they are talking about.....

Now there is a lot of really good info here in this thread in the damping of speakers, but what we are really after is an explanation as to why we (damacman and myself) can hear a more punchy tighter lower and more sustained, yet controlled bass sound from our subs using a Crown (the Macro-tech 5002VZ in my case) and yet when we use another amplifier such as a MC2300, MC352, MC452, MC2205, Carver PM1.5, Sansui (various power-amps) Perreaux, etc etc, that the bass is a bit disappointing compared to the Crown.....I might add that I also had success using a QSC Powerlight 4.0...
The 4.0 has a damping factor of greater than 500 and the Crown greater than 1000.....
It is also interesting to note both the Crown and the QSC are both bridged amplifiers

The bass in my system is absolutely astounding, and it is usually the first thing people say about my system when they haven't heard it before, one of my close friends was over yesterday and we were listening to a few tunes at a moderate level, he has heard my system many many times, and he commented again on how fantastic the bass is in my system.....
Of course I want people to say, "wow, that MC452 really sings, which they do, but they comment on the bass first"
Take that Crown away or the QSC away, and that magic bass is gone!! I **** you not!!

Unfortunately amplifier manufacturers have used both sides as marketing tools which further confuses the issue......
I can hear it, but the Maths says I cant? So what is it if it is not the amplifiers damping factor?

I use the "modest" Crown XTI 4000 and XTI 2000 amps to drive various woofers in four different systems. I have tried quite a few types of amps over the years to provide this serice, including the very good QSC 2402 Powerlight amp. Still, the Crowns have prevailed. This is true regardless of speaker configuation: DIY dual 12s, DIY dual 15s, pro cabinets with dual 15s each (ported) and dual 18s (Acoustic suspension). The prior reference to amps in bridged mode is interesting: although all my woofer amps have been bridged mono where they provide the greatest authority and control, I thought that the damping factor was halved when in bridge mode. Yes? No? In any case, that difference seems insignificant in my applications.

 

[BMWCCA]

Just checked the reference sheet for my Crown Studio Reference II and it shows damping factor >20,000 from 10 Hz to 400 Hz. That's a big number! :yes: