Damping Factor

[Guest Sime]

It’s something that doesn’t get talked about a lot, it’s specs, if ever mentioned, can vary massively from amp to amp. How important of a specification is it, what situations/speaker requirements may want someone to seek a high damping factor, and if it is a necessity, what are the rating ranges to be looking for. 

I know my mono’s are rated at 800, is that overkill for an 805, or could they improve from a higher spec?

 

Is it as important as any other amplifier spec?

 

cheers. 

[Guest]

 My favourite topic 

[VA50]

To my understanding dampening factor refers to an amplifiers ability to stably control the movement of a speaker. A large speaker with a wild impedance curve for example will be more suited to an amplifier with a higher dampening factor as it will be able to more accurately control the movement of the speaker. You could look at the dampening factor as like a brake that can stop the movement of the speaker i guess.  

 

I'm sure others will chime in with a better explanation soon. 

[frankn]

Have a read through the posts in this thread: Joe Rasmussen, Custom Analogue Audio, JLTi

under the Melbourne Audio Club Inc. 

There are some differing opinions. However, it seems that once you get over 50-100 it doesn't mean much. 

[Tweaky]

Depends what sort of speakers you are using, you'd certainly want a higher damping factor than 100 with a Bass reflex speaker with a woofer of 8" or bigger.

 

The bass should show a lot less overhang and generally sound taunter, it's one of the reasons I [and many others] don't use Tube amps with Harbeth SHL5's, as generally, tube amps have a lower damping factor than Transistor amps....of course, there are always exceptions.

[MLXXX]

1 hour ago, frankn said:

There are some differing opinions. However, it seems that once you get over 50-100 it doesn't mean much.

Yes, though perhaps not even 50 is needed, according to this article: http://www.hi-fiworld.co.uk/index.php/amplifiers/75-amp-tests/149-damping-factor.html

 

[I was surprised at the representative figure adopted for speaker lead resistance, of 0.3 Ohms.  That struck me as being on the high side.]

[BATMAQN]

2 hours ago, Tweaky said:

 

The bass should show a lot less overhang and generally sound taunter, it's one of the reasons I [and many others] don't use Tube amps with Harbeth SHL5's, as generally, tube amps have a lower damping factor than Transistor amps....of course, there are always exceptions.

Really?? 

I’m using an Earl Weston with SHL5 getting pretty good results here.

 

[Zaphod Beeblebrox]

Funnily enough, I am presently writing a treatise on damping factor. I should move it along a bit, as there is a huge amount of misunderstanding about the issue. Here are some very broad generalisations:

 

* The term 'Damping Factor' is quite misleading, as it refers to a ratio of the output impedance of the amplifier, relative to an 8 Ohm load. Many speakers are not 8 Ohms. Many speakers that are 8 Ohms, are not 8 Ohms over much of the frequency range. The important figure for consumers is actually 'Output Impedance'.

 

* There is not much point in trying to achieve a DF of more than about 20, OVER THE FULL FREQUENCY RANGE. For a speaker with a 4 Ohm minimum impedance, that means around 0.2 Ohms OP Z (Output impedance) will usually be entirely adequate. More is better, but the law of diminishing returns sets in pretty quickly, once the DF exceeds 10 or so. 

 

* A high OP Z (low DF) can affect the frequency response of a loudspeaker, but not only at the bass end, but also at the HF end too. This can be a particular problem with ESLs, where HF impedance can fall to very low levels (see graph). I have also included an impedance curve of an electrodynamic speaker. Both show how important a low OP Z can be. In the case of the ESL speaker system, the OP Z should less than 0.05 Ohms, whilst in the case of the Kappa 9, the OP Z should be less than 0.8 Ohms. These figures correspond to damping factors of around 160 and 100 respectively. It is important to note that these are extreme figures, but they are not unheard of. 

[davewantsmoore]

8 hours ago, Sime said:

How important of a specification is it

Most things you will read written about it completely misrepresent it.    It is really really misunderstood by most people.

 

Ignore any reference to damping factor being "control" or "grip" .... or providing some type of other "intangible" performance increase.   This is an obvious tell that the person doesn't understand what they're on about.

 

33 minutes ago, Zaphod Beeblebrox said:

More is better

You say more is better....   becuase not enough will affect the frequency response of a speaker (which is true).

 

The practical implication of this, is one should use the "correct" type of amplifier to drive their speaker (ie. the one recommended by the speaker design) .... lest you get a frequency response that the designer did not expect (and a different sound).

 

BUT .... this (the frequency response) is just the tip of the iceberg.

 

If we take a speaker, and drive it with a high or a low amplifier output impedance (aka damping factor) .... we will have a different frequency response.      If we CORRECT the frequency responses of the two systems to be the same  (not practical for many consumers, of course) .... and THEN we look at the differences, there is heaps going on - and the whole "voltage source = more damping = better" idea completely fails.

 

 

[Guest Sime]

Ok, when an amp designer has his topology design on paper, of course he’ll have all the performance goals set out during design, is damping one of those goals premeditated or is it something that only after the amp has been built, measured, that he’ll then see what the damping results are. Is it something designers are striving for as part of the designs?

[Zaphod Beeblebrox]

13 minutes ago, Sime said:

Ok, when an amp designer has his topology design on paper, of course he’ll have all the performance goals set out during design, is damping one of those goals premeditated or is it something that only after the amp has been built, measured, that he’ll then see what the damping results are. Is it something designers are striving for as part of the designs?

Damping factor was not something anyone mentioned until the early 1970s. The reason was that manufacturers were scrabbling for numbers to prove that their amplifier was better than the opposition's amplifier. A classic case are the Phase Linear amps of the early 1970s. The claimed DF was a phenomental 1,000. Trouble is two-fold:

 

1) The DF of 1,000 is largely achieved by using massive amounts of global NFB. Large amounts of poorly applied global NFB can introduce other, more serious problems. And, as I previously stated, needlessly large anyway. 

2) The DF was only applicable at frequencies below 1,000Hz. At high frequencies, the DF was much, MUCH lower. Thus, the frequency response, when driving ESLs or other loads that present low impedances at HF, was compromised.

 

In reality, manufacturers make/made a big deal about DF was because they want people to be impressed by numbers that don't really mean all that much. As I showed earlier, a DF of around 150 will be entirely adequate for even the most brutal speaker loads.

[Addicted to music]

20 minutes ago, Sime said:

Ok, when an amp designer has his topology design on paper, of course he’ll have all the performance goals set out during design, is damping one of those goals premeditated or is it something that only after the amp has been built, measured, that he’ll then see what the damping results are. Is it something designers are striving for as part of the designs?

No

[frankn]

@Sime - I think that as you see from various responses here, and also if you do some further research there are differing opinions of what matters with damping factor. Different engineers, designers, companies have their opinions on what is their preferred way. One good take away is don't get 'hung up' on numbers, and biggest isn't the best. If your system sounds good, no obvious problems then it's good. Nothing is perfect - doesn't exist.. Your room response will overwhelm any small issues you don't have with amplifier/cable/speaker damping issues. 

[davewantsmoore]

10 minutes ago, Zaphod Beeblebrox said:

As I showed earlier, a DF of around 150 will be entirely adequate for even the most brutal speaker loads.

My only point is that it's very enlightening to unpack what "entirely adequate" means .....   others will hear "adequate", and still think of the "more is better" thing.

 

 

The shame, is that hiding behind behind "damping factor" .... is actually all things about how a speaker works.   Why does it move?  What are the thing that can result in distortion?

 

The low output impedance / high damping factor .... quickly confuses people into thinking that voltage drives a speaker.    It doesn't.   The speaker moves because of current flow.

[Guest Sime]

11 minutes ago, frankn said:

@Sime - I think that as you see from various responses here, and also if you do some further research there are differing opinions of what matters with damping factor. Different engineers, designers, companies have their opinions on what is their preferred way. One good take away is don't get 'hung up' on numbers, and biggest isn't the best. If your system sounds good, no obvious problems then it's good. Nothing is perfect - doesn't exist.. Your room response will overwhelm any small issues you don't have with amplifier/cable/speaker damping issues. 

I think DF is more of a curiosity to me more than anything. 

[Joe Rasmussen]

This is a myth that will never go away. The maths does not support it and never has, both Neville Thiele and Richard Small knew that (and still does).

 

Damping is real.

 

But 'Damping Factor' is a description of a mechanism that does not exist. 

 

Let go science for a second. In a speaker (let us think of a sealed box for simplicity), there are two kinds of damping, Electrical damping and Mechanical damping.

 

If 'Damping Factor' is for real, what kind of damping can the amplifier affect? Clearly it is Electrical damping.

 

So we are not arguing against damping here, damping is very real. But it is a question of how damping works. 

 

Here is the maths that defines the Electrical damping of a loudspeaker, right out of the Thiele-Small textbook.

 

Qes is Electrical damping and we can define thus:

 

 

HiFi World has the popular explanation for 'Damping Factor' - but ask yourself, is this explanation compatible with the above equation that is set in cement:

 

 

"An amplifier presents a low output impedance to a loudspeaker and it similarly will stall the bass cone if a signal abruptly ceases, causing it to stop rapidly, rather than flap back and forth in uncontrolled fashion as it settles down. An amplifier with low output impedance grips the bass cone and produces ‘tight’ sounding bass, free of waffle and overhang it’s said, and our listening tests confirm this. High output impedance (low damping factor) valve amplifiers in particular produce soft and bloated bass with weakly damped loudspeakers. But audio is never simple and the picture turns out to be a little more complex, albeit interesting for what it suggests." HiFi World

 

No, there is nothing here that is compatible with the above scientifically correct equation. And even the article in HiFi World seems to admit there is more to it than that, but goes on to a less than fuzzy conclusion about the number that is being adequately high enough etc.

 

Look at their own example above. The total Re = 0.1+0.3+0.66+6.3 = 7.36

 

What has happened here according to the equation?

 

 

Now Re has been increase by dividing 7.36/6.3 = 1.168

 

You have decreased the Electrical damping because it has increased by 16.8% (higher number means less damping - for those who need to know).

 

Nothing else has happened, absolutely nothing!

 

The higher Re of 16.8% has increased the Qes by 16.8% and hence the Alignment of the speaker system has changed. The Alignment is now less damped.

 

What does the amplifier contribute to this. Tube amps can be a lot, solid state not as much. But both lessens the damping, just be different degrees.

 

There is no mysterious back-EMF being absorbed by the amplifier and a brake that 'stops' the cone, this mechanism does not exist. But you have changed the damping of the alignment.

 

The Amplifier has no idea what makes up the 7.36 Ohm, it can only see the total sum of 7.36 Ohm and cannot see the divide between the Re of the driver as a separate entity to the added 1.03 Ohm. This is important, because the 'Damping Factor' assumes that something is seeing the divide between 6.3 Ohm and 1.06 Ohm making up the whole 7.36 Ohm, but nobody can explain how the system can be aware of that divide and only see the sum of series elements. That alone tells us that there is a problem dividing the speaker's impedance with the amplfier's output impedance. Such a division makes no sense. In fact, you cannot divide at all and it is meaningless, the only thing you can do is add them into a single series figure/amount.

 

And that is indeed what Thiele-Small Parameters does, it makes no allowance other than to add and not divide.

 

We hear the change in damping in the speaker being varied by the output impedance of the amplifier, we hear a change in damping, that is not questioned.

 

Bottom line: Higher added on Re reduces the Electrical damping and that is what we hear. An Amplifier cannot have a negative impedance that can reduce the total Re and hence any real world amplifier can only make damping worse and not better. This is what we hear when we have a valve amplifier with high output impedance wired to a speaker that already has an un-damped alignment and then the amplifier makes it worse. The amplifier is changing the alignment because Qes is a key component that determines the damping of the alignment.

 

PS: If indeed the amplifier was capable of 'braking' cone movement and apply 'stopping' power, ask yourself if this is really possible when the current produced by such a back-EMF component can be absorbed if the that current has to go, it has to travel through the Re of the driver itself? The Re is in fact a current limiter that stops it being potentially absorbed by anything that might be considered a 'short' - and another myth is that the output amplifier in ideal terms is a 'short' and be a brake of sorts - but that would lead towards another discussion. It has to do with the nature of a voltage source.

 

 

.

 

Edited October 12, 2017 by Joe Rasmussen

[Joe Rasmussen]

51 minutes ago, Zaphod Beeblebrox said:

 

1) The DF of 1,000 is largely achieved by using massive amounts of global NFB. Large amounts of poorly applied global NFB can introduce other, more serious problems. And, as I previously stated, needlessly large anyway. 

 

 

Hear! Hear!

 

That means you can make the nominal DF any number you like. Just increase open loop gain (often bad for the sound, as we know) and increase the amount of feedback and you can almost come up with any DF number.

 

Ask the customer what DF number he wants, then just dial it in. Voila!  

 

But the question is, since a speaker's damping is entirely defined by its alignment, does that not in itself make a constructed DF number meaningless.

 

I note that many speaker designers never bring up DF and yet they do talk about damping and compatibility with amplifiers like tube SE which typically has an output impoedance of 5 Ohm or more. The best bass I have ever heard was an amplifier with an output impedance of 3 Ohm. Very poor DF would be the dismissive reaction of some. Until you play it for them into speakers that have a Bessel like alignment (they can handle extremely high output impedance, but they are too rare) and there is no lack of damping at all. 

 

But I am sure that the DF mythology will continue on its merry way. Even if the science of it is very dodgy - in fact non-existent. 

 

But in fact, I hardly even bother thinking about it - I left this behind decades ago, thanks to Richard Small setting me straight. Can't be bothered by something that does not exist. So I am no crusader.

[Tweaky]

13 hours ago, BATMAQN said:

Really?? 

I’m using an Earl Weston with SHL5 getting pretty good results here.

 

Actually I spoke to Earle Weston years ago about getting one of his tube amps to drive my SHL'5, and even sent him the tech details of the speaker from the Harbeth site.

 

I've still got the email reply somewhere, I'll have to search for it, but in any case, after looking at the SHL5's specs, he though it would be better if he tweaked the amp to suit it better [I'm pretty sure it was his Tempest integrated I was looking at getting], and I'm pretty sure it was the Damping Factor he was going to raise, not by much if I remember correctly, but raise in none the less.

 

In the mean time, I had tried the Harbeth's with some Cary and Audio Research Tube amps, and felt like the Tube sound with the SHL'5 wasn't the sound for me, very nice in a lot of ways, but just didn't control the low frequencies as well as a Transistor amp in my room.

[MLXXX]

9 hours ago, Joe Rasmussen said:

The Amplifier has no idea what makes up the 7.36 Ohm, it can only see the total sum of 7.36 Ohm and cannot see the divide between the Re of the driver as a separate entity to the added 1.03 Ohm. This is important, because the 'Damping Factor' assumes that something is seeing the divide between 6.3 Ohm and 1.06 Ohm making up the whole 7.36 Ohm, but nobody can explain how the system can be aware of that divide and only see the sum of series elements. That alone tells us that there is a problem dividing the speaker's impedance with the amplfier's output impedance. Such a division makes no sense. In fact, you cannot divide at all and it is meaningless, the only thing you can do is add them into a single series figure/amount.

I would see the conventional formula used to calculate an amplifier's Damping Factor for a particular impedance load as only that, a formula. 

 

For argument's sake let's create the term "Relative Power Cord Resistance Factor" and define it using the formula  (Load Resistance) / (Power cord resistance).  A higher figure would be better.

 

If we wanted we could instead define "Relative Power Cord Resistance Factor" with the slightly more complicated formula [(Load Resistance) + (Power Cord Resistance)] / (Power cord resistance).  Again, a higher figure would be better.  I'd note that for very low resistance power cords, the simple formula would give a very similar result to the more complex formula.

 

I don't think the simplicity of the conventional formula for the Damping Factor of a power amplifier should be regarded as an error. I myself have always assumed it to be an intentional simplification.

 

9 hours ago, Joe Rasmussen said:

There is no mysterious back-EMF being absorbed by the amplifier and a brake that 'stops' the cone, this mechanism does not exist.

I'm not sure I follow you here. A conventional loudspeaker can be used as a microphone by connecting its leads to a preamplifier. If you were to short those leads I would think the cone's very small movements when exposed to air pressure variations (sound) would be less than with the leads left open circuit. 

 

It could be I've missed your point. Could you explain it further? 

Edited October 13, 2017 by MLXXX

[allthumbs]

I have been spending sometime at the Harbeth User Group on the Harbeth site of recent past.  It is an interesting place with the discussions that take place in regards to topics such as Damping Factor and the advantages/disadvantages (real or imagined) between one amplifier and another.

 

Alan Shaw participates in the blog quite vigorously and has carried out some interesting tests that challenge commonly held opinions.

 

Worth a visit, there is plenty to read and a lot of science, insights into design principles and the Hi-Fi industry in general.

 

It has changed my mind on certain topics and provided a lot of food for thought.

[Joe Rasmussen]

49 minutes ago, allthumbs said:

 

Alan Shaw participates in the blog quite vigorously and has carried out some interesting tests that challenge commonly held opinions.

 

I have not read his forum, but I have become that aware that Alan Shaw's view is likely identical to mine. And yes, arguments against any DF notion can be attacked from so many angles and they all conclude that DF is basically bunkum (as in nonsense, as in making no sense at all). 

 

Think about this fundamental fact that an amplifier must have an output impedance. This mean that Qes can only be increased, never decreased. An increase in Qes is less damping.

 

So in fact the amplifier can only make damping worse and not better. For the amplifier to add damping, the output impedance would have to go negative for the simple reason that the loudspeaker sees its own DC resistance as a factor of damping. If Re (which is the same as the DC resistance) is 8 Ohm, then any increase above that caused by the amplifier, will erode the electrical damping available to the loudspeaker. If 8 Ohm goes up 10%, then so was the Qes  hence damping has been reduced by 10%.

 

The only way that the amplifier can add damping, it must reduce the DC resistance inside the Voice Coil. Somehow the amplifier connected has to reduce the Re of the Voice Coil. That is never going to happen.

 

Irrefutable Fact: The damping of a loudspeaker is entirely defined by its alignment and nothing else.

But I suspect I am aiming my blows into the air. This is one myth I suspect that will never go away.

 

BTW, if you know how things works, then a speaker with this impedance and electrical current phase angle would work the same with any impedance and hence proves DF to be irrelevant.

 

 

Impedance Blue and Current Phase Angle Red.

 

If those parameters can be achieved, the alignment will not change with any output impedance of the amplifier, even if the amplifier had infinitely high impedance.

 

And yes, the above is an actual measurement of an actual speaker - and the amplifier cannot add or subtract damping.

 

Edited October 13, 2017 by Joe Rasmussen

[Joe Rasmussen]

1 hour ago, MLXXX said:

I'm not sure I follow you here. A conventional loudspeaker can be used as a microphone by connecting its leads to a preamplifier. If you were to short those leads I would think the cone's very small movements when exposed to air pressure variations (sound) would be less than with the leads left open circuit. 

 

It could be I've missed your point. Could you explain it further? 

Don't get me wrong, a speaker/driver produces back-EMF and in fact there are three different types.

 

1. Motional back-EMF - mostly below 100 Hertz.

 

2. Microphonic back-EMF - broadly across many frequencies.

 

3. Inductive back-EMF.

 

In fact, if a driver produced no back-EMF, then if the DC resistance (Re) of the Voice Coil was 6 Ohm, then the impedance would be flat 6 Ohm from DC to infinity. Of course that is not the case.

 

So if the impedance at any point, at any frequency, is above 6 Ohm, this is caused by back-EMF.

 

BUT:

 

You actually put your finger on it when saying:

 

"If you were to short those leads I would think the cone's very small movements when exposed to air pressure variations (sound) would be less than with the leads left open circuit."

 

Ah, but think about it. Can you really short the leads of a speaker (basically its terminals) and get a real short? No you can't, for the simple reason that the 6 Ohm DC resistance (Re) we used above, prevents you from really shorting out the leads. The energy is being absorbed inside the Voice Coil by the Re, not by the amplifier 'short'. Hope you can see that.

 

Indeed, the amplifier can't absorb anything at all. I know this is difficult to grasp. 

 

A loudspeaker is a current device (if driven from a voltage source, it has to convert voltage into current on the fly - it becomes a V/I converter), but with any back-EMF, the driver itself becomes a voltage source. So here we have Dr. Jekyll and Mr Hyde amply illustrated. From here it gets really complicated, but guess, where do you think the excess voltage can be absorbed? What will become the V/I converter here? 

 

Why can't a voltage source be regarded as a short? Simply because it can produce current at any phase angle - and a short cannot produce anything. So they are not the same.

 

Again, this is another little furphy. If you take a voltage source and put a large value resistor in series with it, hundreds of Ohm, then it become a current source. It can only produce current with a zero degrees phase angle. The ratio of the source versus the load needs to be a very high ratio relative to our 6 Ohm. Now as a thought experiment, we have a voltage source into a 6 Ohm load, but we put a ridiculously (this is only a thought experiment) low 0.1 Ohm DC resistance in parallel with the 6 Ohm load, again the ratio of the two must be high, now the voltage source can only produce current at near zero phase angle (this trick, not done to the same extreme, is what Wilson Audio does in their ridiculously expensive speaker, to flatten the load's current phase angle).

 

The question must be asked, how can a voltage absorb energy as some kind of brake, when the current phase angle cannot even be matched. Indeed, worse, the phase angle of the amplifier can be bossed around so easily. The back-EMF can do that to the amplifier and this has been well documented.

 

So how does this relate to DF? What sounds so simple and seductive explanation is not simple at all.

 

I am reminded of the 'simple' question that is not simple - if you bought a ball and bat for $1.10 and paid one dollar more for the bat than the ball, how much did the ball cost? Of course, it looks like a simple question and it is obvious that the ball cost 10 cents. Except the question was not simple at all and the ball cost 5 cents.'

 

Getit?

 

Cheers. 

 

 

[georgehifi]

12 hours ago, Joe Rasmussen said:

But the question is, since a speaker's damping is entirely defined by its alignment, does that not in itself make a constructed DF number meaningless.

 

No they can go hand in hand!

As a typical example Linn/Naim did a very good if questionable marketing ploy back in the 80's so they could sell just the LIN Isobaric/Naim 160 or 250 combos and all other amps would sound like c**p with the Linn Isobarics, which are a highly over damped design, but sound just magic with a Naim 250 or 160  which are a very low damping factor amps, this is of the time when Linn and Naim (Ivor and Julian) were in bed together and did all shows together and had the same retail outlets to dominate/monopolise with.

We put a ME150 (known for it's bass heft and control) on those same Isobarics, and we were all amazed that there was no bass, the weakling Naim 250 out bass'ed the bass master.

Today this sort of cheatery doesn't go on thank god and all amps usually sound like they are designed into most speakers. because if it were you could never change the speaker or the amp for another make, you are bound to them for life or you get rid of both to go back to normal. 

  

Cheers george  

Edited October 13, 2017 by georgehifi

[MLXXX]

5 minutes ago, Joe Rasmussen said:

Ah, but think about it. Can you really short the leads of a speaker (basically its terminals) and get a real short? No you can't, for the simple reason that the 6 Ohm DC resistance (Re) we used above, prevents you from really shorting out the leads. The energy is being absorbed inside the Voice Coil by the Re, not by the amplifier 'short'. Hope you can see that.

Most of the energy is absorbed/dissipated in the voice coil but that only happens because of the short circuit across the speaker leads.

When you close a switch, current flows through the switch but the switch dissipates very little. It is of course the load that absorbs/dissipates the bulk of the electrical energy.

What is the bottom line here: are you saying that in your view amplifier damping factor is an irrelevancy?

[Joe Rasmussen]

1 minute ago, georgehifi said:

 

No they can go hand in hand!

As a typical example Linn/Naim....

 

How does that the statement incorrect.

 

I will repeat it again:

 

The damping of a loudspeaker is entirely defined by its alignment and nothing else.

 

The amplifier can only add series resistance/output impedance and that alters the alignment. The only way the amplifier can avoid altering the alignment is if the output impedance is zero.

 

I am sorry George, the science is set in cement.