The folks at Krell told me years ago that damping factor was not really that important.
38 responses Add your response
In amps with output transformers, damping factor is limited by the magnetic hysteresis (lag) of the output transformers, to usually not more than 16. This includes all tube amps with output transformers, and certain ss amps with 'auto-formers' (like McIntosh ss amps.)
The main advantage of higher damping capability is in providing cleaner bass in cone woofers and subwoofers. I think 100 is plenty, unless your speaker cables are 50 feet long! Speakers with big voice coils and strong magnets (woofers) create a strong 'reactive load' (a counter-current back to the amp.) Damping is a measure of an amp's ability to overcome this reactive load; to make the cone follow the music signal precisely and reversing its direction of travel instantly when the polarity of the signal changes.
This is why subwoofer makers use ss amps, as well as why people who biamp their speakers use a ss amp for the bass, and a tube amp for the mid/highs. (A tube amp has plenty of damping to control midrange drivers and tweeters, while, many believe, offering a more nuanced presentation of the music, especially the midrange.)
Speaker Q (actually woofer Q) becomes an important issue primarily when it has a resonant peak at a certain (audible) frequency; because the cone becomes hard to control when excited by musical information at that frequency. Adequate damping can help keep the woofer's response flat, but more important IMO, good damping results in better bass transient response (detail.)
Vintage Mac gear (tube) has the lower damping associated with tube amps, although Mac's tube amp damping is always a couple of points better because of their unrivaled output transformers. Their ss amps, (except for a couple of models - can't remember which) all use autoformers, and although they have a bit better damping than their tube amps, I wouldn't buy one, new or vintage ;-) As for what the folks at Krell said: well, if you're using a Krell, you don't need to think about damping ;-)
Too high of a damping and the bass is thin, lacking in authority, and not very musical. Too high is relative depending on the speakers it is matched to. My tube amp has a damping factor of about 11, and the bass on the Legacy Whisper is stunning, better than the SS 600wpc Legacy amplifier also used with the Whisper. Damping is also like a "brake" on the woofer and too much braking and the woofer will not be able to move enough to get some body in the bass. A speaker and amplifier need to be matched with regard to damping and then magic occurs....I also use Bryston amplifiers and they match well with my speakers and their damping is over a hundred, and the bass is great. Numbers don't tell all the story....The biggest mistake is thinking that the higher the number, the better the control of the bass....not necessarily....jallen
Al, isn't damping simply the instant ratio of speaker impedance and the impedance of the amp to the back EMF generated by the speaker? The amp, best case is a dead short (not possible in fact) to this back emf. Take a bare woofer driver and short out the connections with a short piece of wire and 'thump' the cone. Totally damped? yep. Remove the jumper and try again. Rings, if not quite 'like a bell' than you can certainly hear the resonance.
I work in the semiconductor fabrication industry. One of our BRAG number is what we call RDON....Resistance Device ON.....when it is conducting. Lower is 'mo betta and although a known parameter, is constantly measured and efforts made to improve it in new devices. There are of course physical limits when dealing with Silicon devices.
As an aside, Sunfire /Carver used our (company I work for) devices in the output sections of many of his amps / receivers.
Truthfuly, I have not heard a quality sound produced by an amplifier whose damping factor is rated more than 200. Specs don't reveal any secrets how musical and enjoyable an amp will sound, so believe your ears. Digital amps seem to be the exception which normally have a damping factor of up to 1000.
High damping factors are often achieved through the use of large amounts of global negative feedback, which has negative sonic consequences that aren't obvious from the distortion specs (which themselves are virtually useless for evaluating sound quality because they correlate very poorly with subjective perception).
I have designed speakers to work well with amplifiers having a damping factor of about 1, and I'm certainly not the only manufacturer to do so. What sort of damping factor is appropriate for your speakers depends mainly on the impedance curve (primarily the shape of the curve itself, rather than the rated "nominal impedance"). Speakers with a smooth, fairly even impedance curve above the bass impedance peaks can work well with a much wider range of damping factors than speakers with roller-coaster impedance curves.
Yes, that all sounds right. As you imply by referring to instantaneous ratio, the damping effect is frequency dependent (since speaker impedance varies with frequency). By convention, though, the numerical value specified for the damping factor of an amplifier is its output impedance divided into 8 ohms.
Putting the RDON of the output devices within a feedback loop will lower the effective output impedance, with consequent side-effects as Duke and I noted. By producing output devices with lower RDON, the amount of feedback which may otherwise be needed is reduced.
Kijanki: Inductor in-series with the woofer has resistance approx. 0.08 ohm limiting DF to 100.
Good point! Which brings to mind that the dc resistance of the woofer voice-coil itself will also (probably even more significantly) limit the damping which can be achieved, and make extremely high amplifier damping factors meaningless. Since the path of the current which flows as a result of back emf has the amplifier output impedance, the speaker cable, the crossover inductor, and the woofer voice-coil in series, resistance anywhere in that path will affect that current similarly.
"Digital amps seem to be the exception which normally have a damping factor of up to 1000".
My Icepower amp has DF=4000 at low frequencies. It is inherent in class D amps since speaker is always shorted by Mosfets to power and GND (to very low source impedance). Mosfets have fraction of an ohm resistance while even small gain before feedback (I suspect around 100) will lower it a lot. Conventional class AB SS amp has gain before feedback of many thousands (lowers THD and IMD but increases TIM). Class A SS amps are better, with gain before feedback of only few hundred.
As Al mentioned, deep feedback brings negative effects (but improves spects). I wouldn't buy class AB SS amp that has very good spects (something has to give) or at least it wouldn't be my first choice to audit.
Here is the link: http://www.stereophile.com/reference/70/index.html to an article about feedback. Stereophile tests Cary amp with adjustable feedback and the sound is the best when feedback is set to the lowest.
Most speakers do not need a great deal of 'damping' to achieve their design goals. A high damping factor is another way of saying 'low output impedance'.
However there are two schools of thought about what is meant by low output impedance! One school has it that you have to use a lot of loop feedback to achieve it. This is the path taken by most transistor amplifier manufacturers. Not all do though and they are noteworthy has having made some of the finest transistor amplifiers available: Nelson Pass, Ayre and Ridley Audio are the names that come to mind.
However, funny thing- there are a good number of tube amplifiers that IMO/IME play bass better than the transistor amps with high 'damping factor'. These amps do so with 'damping factors' between 2 and 20. Tubes are inherently lower distortion than transistors and so need far less feedback to make the resulting amplifier circuit linear- in fact a good number of tube amps do this with no feedback at all. As Kijanki points out, feedback has very audible detrimental effects.
Our ears use the 5th, 7th and 9th harmonics as a means of determining loudness. We are more sensitive to these harmonics than just about **anything else** in audio! So if these harmonics are emphasized by distortion at all- we will hear it. Audiophiles use the terms dry, hard, harsh, brittle, clinical, chalky, white and the like to describe very slight enhancement of these harmonics.
While loop feedback will reduce overall distortion (when listening to sine waves, the jury is out on whether this is true of a constantly changing waveform), these harmonics in particular are actually **enhanced** by loop feedback. The result is that lots of feedback in an amp will result in an amplifier that will never sound natural.
for more information see
Atmasphere - also speaker design can make for better quality of bass. As far as I know typical, used in 90% of cases, overhung motor speakers have a lot of distortions at lower frequencies. More expensive underhung speakers,used by some manufacturers (like Acoustic Zen) have much lower distortions (but are more expensive).
Deep loop feedback is equivalent to high gain before feedback. With such high gain (in order of thousands)any delay in the signal path (limited bandwidth) results in improper (late) summing of the input and the feedback signal causing TIM - tendency to overshoot, altering shape of the signal (exaggerate odd harmonics).
Sane designer would design amp as linear as possible without feedback and then would introduce just enough feedback to bring distortion below 1%. After that it would be necessary to reduce bandwidth at the input to one that amp had before feedback (to prevent TIM). At the end we would get nice sounding amp that has horrible spects - it wouldn't sell. There is probably much more to it but I wouldn't buy class AB amp with extremely low THD or extremely high Damping Factor (deep feedback).
Another aspect of loop feedback is that all amplifiers have a delay time- the time that it takes for the signal to propogate from input to output.
What this means is that the feedback signal will always be a little late getting back to the input of the amp. As frequency goes up, the problem gets worse as the propagation delay of the amplifier remains constant (in effect the feedback signal is progressively later). At very high frequencies this can cause the amplifier to oscillate if not treated properly in the design.
The result is a sort of ringing effect in the amp, which plays a role in odd-ordered harmonic enhancement. Keep in mind we are not talking about very much distortion; hundredths of a percent is all it takes to be audible.
An amp doesn't have a damping factor. Damping factor is the ratio of the rated impedance of the loudspeaker to the source impedance. In other words, you can only calculate a damping factor when you have a speaker amplifier combination.
What you are looking for is an amp with a low output impedance which then results in a good damping factor number.
"An amp doesn't have a damping factor"
Pauly - You'll often find DF in amps' specifications. "Having" DF simply means amp has that many times lower output impedance (at given frequency) than assumed 8 ohm speaker. DF is a measure of output impedance - hence amp "has" DF.
"What you are looking for is an amp with a low output impedance" - not likely to find in specifications, look for higher Damping Factor instead.
Pauly - I agree. Specifications are very flexible and often misleading. Amplifier (and any other) design is a series of compromises and when you see an amp with exceptionally high DF and extremely low THD something else has to give - the sound.
Joekapahulu - I wouldn't pay too much attention to DF. For instance Atmasphere MA-1 amps have DF<1 (tube amp without output transformer) but sound great according to 6Moons review ("gigantic imaging, excellent dynamics, frequency extension and potent music-making").
An amp doesn't have a damping factor. Damping factor is the ratio of the rated impedance of the loudspeaker to the source impedance.
The convention is to report Damping Factor into an 8 ohm load - so for pratical purposes one might say that power amps do indeed have a Damping Factor.
It can be over rated as the woofer drive motor and mechanical as well as acoustic suspension can play a much bigger role. A high damping factor amp is not going to fix a badly designed speaker with low cost drivers with cheap (small) motor - that will be putting lipstick on a pig.
1st, DF is just a spec, If all manu. measure the same, you should be able to compare. Not that this would make it mean any more or less!
Also, propagation delay is as good a reason as any why certain bi-amp combinations don't work out. 1milisecond is roughly 14" at the speed of sound and will result in phase/time smear effects. Even I can hear that! Pair a fast amp with a slow amp and you could quite possibly get to that level.
Sorry for the late post, but I have a question about DF. I own an ARC tube amp. I just happened to notice that most of the tube amps in the ARC line have pretty low DFs, e.g, less than 20. Nevertheless, most of the ARC tube amps sound great. So what's all the hubbabaloo about DF?? Does it matter that most if not all ARC tube amps use output trannys?
Magfan, I read the posts. I am not an electronics techie, so most of the discussion went over my head. Perhaps Kijanki will pick my post up and answer in layman's terms my question:
"I just happened to notice that most of the tube amps in the ARC line have pretty low DFs, e.g, less than 20. Nevertheless, most of the ARC tube amps sound great. So what's all the hubbabaloo about DF?? Does it matter that most if not all ARC tube amps use output trannys?"
If I manufactured an amp with low output impedance, I would certainly tout its high damping factor as a positive attribute. But, for anything other than really high output impedance amps, like some SET and triode pushpull tube amps, playing into speakers that present a particularly difficult load (low impedance at critical frequencies), the numbers are usually meaningless. Personally, I would not be that concerned unless the output impedance of the amp exceeds 2 ohms (nominal damping factor of 4) for any speaker that is reasonably rated at 6-8 ohms.
Bifwynne, most tube amps have much lower damping factors than most solid state amps. Some very highly regarded tube amps have damping factors in the vicinity of 2, although most tube designs are somewhat higher than that.
What is important is that damping factor and output impedance (which unless otherwise specified is equal to damping factor divided into 8 ohms) should be chosen to be suitable matches for the particular speaker. Mismatches can occur in two ways:
1)Tonal imbalances may result if the speaker design, particularly its variations of impedance as a function of frequency, reflects the expectation that it will be used with a solid state amp and it is used with a tube amp, or vice versa. See this Atma-Sphere white paper. What he refers to as a "power paradigm" amplifier would be one with high output impedance, and what he refers to as a "voltage paradigm" amplifier would be one with low output impedance.
2)Speakers rely to varying degrees on the amplifier for damping of "back emf" produced by the woofer. Inadequate damping will result in "loose" bass, because the woofer cone will tend to continue moving after the signal has stopped or changed. However, as indicated earlier in this thread, once a damping factor is provided that is adequate for the particular speaker, which is rarely if ever higher than a two-digit number, further increases in damping factor will not be helpful, and in some cases may be reflective of excessive feedback that would be sonically harmful.
The key thing is the matchup of speaker and amp. Some speakers are good matches for both tube and solid state amps, but some are not. A simple way to help narrow that down is to research what amps others have used successfully with the particular speaker.
Thanks Kijanki, Larryi and Almarg. The Atma-Sphere white paper was particularly interesting and helpful. Al's advice concerning matching amps with speakers now makes sense. In particular, I can see why DF may be a bit of a red herring. But I think there may be a practical problem with the other very good advice posted above. That is knowing how to research which amps and speakers match well.
For example, when I spoke with ARC a while back, the folks were not familiar with Paradigm Signature speakers. At the same time, when I spoke with the Paradigm tech folks some time ago, they were not very familiar with ARC tube gear.
Now to my ear, my Paradigm/ARC combo sounds ok, but is it really a good match? Am I really getting the best sound possible out of the combo? How can I really know? Let's face it, it's not practical to schlep a 62lb tube amp around with me when I audition speakers. At the same time, a dealer may not be driving his portfolio of speakers with a tube amp, but rather a SS amp -- which is often the case.
Can one say that if a speaker has a nominal impedance of 8 ohms, there should be a good speaker/tube-amp match? Is speaker efficiency a relevant factor? Any other good rules of thumb that might provide helpful guidance? Just thinking out loud, maybe the best thing to do is to call the speaker manufacturer and ask if a particular speaker matches well to a "power paradign" amp and/or a "voltage paradigm" amp.
Thanks again guys!
Can one say that if a speaker has a nominal impedance of 8 ohms, there should be a good speaker/tube-amp match? Is speaker efficiency a relevant factor? Any other good rules of thumb that might provide helpful guidance?Bifwynne, it's probably not possible to give a reasonably simple all-purpose answer to your question, because there are many variables involved. It is probably fair to say that a LOOSE correlation exists between higher nominal impedance + higher efficiency and suitability for use with a tube amp. However, a more telling factor is often the impedance vs. frequency curve of the speaker.
Keep in mind that an amp having low output impedance, such as most solid state amps, will supply current into a given load impedance that to a close approximation (and assuming the amp is operated within its limits) increases in direct proportion to a reduction in load impedance, while the voltage it outputs will be essentially unaffected by load impedance. Therefore if speaker impedance goes down at some frequencies, power delivery at those frequencies will increase (Power = Voltage x Current, oversimplifying slightly). An amp having high output impedance, such as most tube amps, will not behave that way, and it's power delivery will vary to a lesser degree as load impedance varies.
In general, the flatter the curve (i.e., the less variation of impedance with frequency), the more likely it is to be a good match for a tube amp. There are some exceptions to that, such as electrostatic speakers, which commonly have impedances that decline to very low values in the upper treble, and which are often much better matches to tube amps than to solid state amps. The low output impedance of a solid state amp would cause it to deliver more power into that lower impedance than a tube amp with higher output impedance would deliver, causing the upper treble to be over-emphasized with that kind of speaker.
At the other extreme, many dynamic speakers have low impedances in the bass region, and higher impedances in the mid-range and treble. That can sometimes be an indication that the speaker is intended for use with a solid state amp, where the designer is counting on increased power delivery into the low impedance to reinforce the bass.
I couldn't find an impedance curve for the Paradigm Signature S8 v2 which you indicated in another thread that you are using, but I'll assume it is similar to the curve for the earlier version of the speaker, shown near the bottom of this page.
As you'll see, the impedance curve is anything but flat. It ranges from about 3 ohms in a lot of the bass region, to a 21 ohm peak in the mid-range, and is in the general vicinity of 8 ohms in most of the treble region. That would indicate that in comparison to a solid state amp, a tube amp would de-emphasize the bass, emphasize the mid-range significantly, and provide some de-emphasis of the treble, but to a lesser degree than the bass.
Another thing to watch out for, in the plot of impedance phase angle, is highly capacitive phase angles (i.e., angles that are significantly less than 0, approaching say 40 degrees or so), that occur at frequencies where the impedance is low. The S8 has that combination in the area of 60Hz. That results in the amplifier having to supply relatively large amounts of current at that frequency. I suspect that would not be a problem for your VS115, but it might be for some less powerful tube amps.
Your VS115, btw, has a damping factor of 8, corresponding to an output impedance of 1 ohm on its 8 ohm tap. So the effects I have described would be smaller in degree than in the case of many other tube amps that have higher output impedances. A way of reducing those effects further, if you find it to be sonically preferable, would be to use the 4 ohm taps. That would provide an output impedance of 0.5 ohms, although maximum power capability would be reduced significantly.
Hope that helps,
Bifwynne, I did some quick calculations to provide a quantitative perspective on the effects I described above. In comparison to an amp having negligibly small output impedance, the 1 ohm output impedance of the 8 ohm taps on your amp will result in an emphasis of the mid-range frequencies for which the speaker has an impedance approaching 21 ohms, relative to bass frequencies where speaker impedance is around 3 ohms, of about 2db.
Using the 4 ohm taps would reduce the 2db figure to about 1.1db, although it would significantly reduce maximum power capability, as I indicated.
One point NOT addressed so far is that of phase angle.
When voltage leads or lags current, power to the speaker is compromised.
Tube amps do not like capacitive loads, either.
Sensitivity, while important, may in some cases be trumped by a phase.
A high sensitivity speaker with a huge phase angle will still not be proper for a tube amp. OTOH, a low sensitivity speaker which is a moderate load, can and does sound fine with tubes. Harbeth comes to mind.
To the extent that compatibility can be decided without listening to a given combination, I'd avoid big impedance swings and low impedance coupled with high phase angles at particular frequencies when using tube amps.
OK guys, here's the verdict. But first, let me wax a little philosophical. As a lawyer, but not an electronics techie, I am reminded of the axiom that some legal discussions are "multi-dexterous." In other words, on the one hand -- this, but on the other hand -- that, but then again, on another hand -- something else. When you add up the hands, you can have three or more hands -- a clear physical impossibility.
Ok, here's the bottom-line verdict, I spoke with Paradigm about the wisdom of trying the 4 Ohm taps on my ARC VS 115 tube amp. The tech said using the 4 Ohm taps "might" improve the sound, but then again, maybe not.
Well, I tried the 4 Ohm taps. I did not detect any sound improvement -- if anything sonic degredation. I tried the 8 Ohm taps again, and to my untrained ears, the sound improved. Well that's it. Fascinating technical discussion. I wish I knew more about electronics, but I don't. But I do know what sounds good, at least to my untrained ears.
I'll stick to the 8 Ohm taps, but continue to read on to learn more. In summary, I love the sound of my ARC VS 115 amp and Paradigm S8 v2 speakers, using the 8 Ohm taps. Maybe I shouldn't, but ignorance can be bliss. Thanks for the posts. BIF
In summary, I love the sound of my ARC VS 115 amp and Paradigm S8 v2 speakers, using the 8 Ohm taps. Maybe I shouldn't, but ignorance can be bliss.BIF, thanks for the update. I don't think there is any reason that you "shouldn't" prefer the configuration which sounds best to you. The impedance curve of your speakers seemed to suggest that trying the 4 ohm taps would be a worthwhile experiment, but clearly there are an enormous number of other variables in the system, the room, and your own listening preferences that can overshadow what MIGHT be best in terms of that single variable.