Another view of damping factor?

You can troll around and find more or less two views of damping factor:

A) A high DF doesn't matter because the [typical] amplifier's output impedance is swamped by the speaker wire and voice coil resistance.

B) Amps with high DF have "more control" over cone motion and sound better.

Sometimes there is a middle view
BA) High DF is good but doesn't matter once great than [insert magic number here].

I was talking with a friend who opined that high DF was not important for its own sake. But his thesis was that designing an amp* for good sound, particularly with subs, results in a high damping factor. That by shooting for wideband frequency response and low noise and especially indifference to back EMF, a high DF often results as a byproduct.

What do you think?


*The context of the discussion was mainstream solid-state designs, including more lately Class D designs which often don't have good damping due to the physical output filters.
 
You can troll around and find more or less two views of damping factor:

A) A high DF doesn't matter because the [typical] amplifier's output impedance is swamped by the speaker wire and voice coil resistance.

B) Amps with high DF have "more control" over cone motion and sound better.

Sometimes there is a middle view
BA) High DF is good but doesn't matter once great than [insert magic number here].

I was talking with a friend who opined that high DF was not important for its own sake. But his thesis was that designing an amp* for good sound, particularly with subs, results in a high damping factor. That by shooting for wideband frequency response and low noise and especially indifference to back EMF, a high DF often results as a byproduct.

What do you think?


*The context of the discussion was mainstream solid-state designs, including more lately Class D designs which often don't have good damping due to the physical output filters.

IMO it's hard to really 'use' the damping factor beyond 100. Your speaker wire and connectors would have to be less than 0.08 ohms for an 8 ohm system to exceed 100. I have heard high damping amplifiers that sounded 'constricted' (I hate 'descriptions' of sound). That same amp sounded dramatically better when the feedback was reduced. No, the gain wasn't changed but the open loop gain _was_ reduced so the feedback was less. The bass was tight, solid and well defined. I apologize for _more_ of these descriptive terms. The owner was as pleased as I with the result.

I wish I had measured the damping after the change but I would guess it was lower.

 
Damping factor -- or better to consider output impedance -- needs to be considered in the light of the speaker (and wires), it is a system with.

High damping factor can result in an overdamped system with some speakers. We are staring to see more & more speakers that are happy with amplifiers with high output impedance (low damping factor).

dave
 
Damping Factor, as well as THD are numbers that are just about useless as far as the quality of sound is concerned. SET amps have very little, if any, negative feedback and very low DF as a consequence. Yet as Nelson Pass, and others have demonstrated, these designs can actually have better bass response (within their power limitations) than SS amps with far higher DF.

I grew up with "the Iron Grip that a high DF provides", and like discovering the secret about Santa Claus, it was a great disappointment when I learned the truth.
:bawling:

Best Regards,
TerryO
 
In my experience, the damping factor of the amp is very important, but for a different reason. As the output impedance of the amp changes, so does the frequency response of the system, because most speakers have such a highly frequency dependent impedance. We are very sensitive to relative amplitudes at different frequencies within the audio band. I'm pretty sure I can detect 0.1-0.2 dB variances. I've had to modify crossover values in my speakers to account for damping factor differences after changing amps. I think these output impedance differences are the main reason for different amps sounding different (assuming well designed amps).
 
View BA is nearest to the truth, with the caveat that it depends on the loudspeaker. Obviously a very high DF is swamped by wiring. A very low DF will result in frequency response variations, as few speakers have a level impedance.

Given a high enough DF that frequency response variations are small, the issue then becomes bass damping. This can be a mix of mechanical and electrical, complicated by the usual audio issue of 'accuracy' versus 'preference'. Maybe the speaker designer should specify what DF he intended.

High DF is usually achieved by some mixture of local and global NFB. My preference would be for as much as reasonably possible to come from local NFB (e.g. follower output, with a low impedance driver) with global NFB then adding a bit more.
 
Interesting comments, but straying from my question which I must have put poorly.

So, let me re-phrase, since I'm really curious about the opinions of folks who design and build amps.

Never mind if damping factor is important or not, and never mind about the speaker wires and voice coil impedance.

What do you think about the thesis that designing for good sound will generally result in a high damping factor/low output impedance?
(not for tube or some other exotic designs, of course, he meant typical solid state)
 

forr

Member
2004-12-01 6:46 pm
Next door
Whatever the damping factor of the amplifier, even infinite, the limiting factor in the damping of a driver resonance is the voice coil resistance which, for an 8 Ohm unit, may be something like 5 Ohm.

However there is a way to circumvent this limitation : using an associated circuit such that the amplifier presents a negative output resistance, the resistance of the circuit in then decreased (for example, the initial 5 Ohm value can be safely reduced to 2 Ohm).

This is a neglected but powerful technique.
 
Hi,


With solid state power amplifers its almost impossible not to end up with
a decent / high damping factor. The only exception I can think of is for
output inductorless designs that use a resistor instead, typically 0.22R,
this limits damping factor to around 40 (for 8 ohms) by default. Its also
possible to manipulate the output impedance for driving subwoofers.

rgds, sreten.
 
Last edited:
My 2 cents. High damping factor is nearly always a consequence of a large amount of NFB, whether the amplifier design is good or bad. A good design with a lot of NFB will have a high damping factor, a design with low (or no) feedback will generally have lower damping factor. Even class D amps can have high damping factor, if the feedback goes around the inductors. To me, the issue goes back to the speaker designer - if a speaker is designed to be flat with an ideal voltage source, then high DF is desirable. I am curently toying with larger boxes and lower amplifier output impedances, for increasing bass response with high Q woofers. There is some good reading in the Rod Eliot audio site for higher output impedance amplifiers, as well as here, but this is digressing from the orignal question. GOOGLE "Dick Pierce Damping" for a good read on damping factor.
 

Rodeodave

Member
Paid Member
2004-12-26 9:38 pm
in the alps
As far as I understand the topic, an amplifier on it's own doesn't have a damping factor, not without a load. It just has an internal resistance or output impedance. The damping factor is the ratio between load impedance and source impedance.
So shouldn't the question then rather be if a pairing of amp+speaker has to result in a high damping factor in order to work well?
 
It depends what you mean by high. Given the ohmic resistance already present there is no point in aiming for very high DF. It may occur as a natural by-product of negative feedback, which is not necessarily bad but depends on how it is used.

A negative output impedance will cause frequency response variations just like a positive output impedance, unless it is simply compensating for a long cable run. This assumes that the speaker was designed for a voltage source (i.e. zero impedance). If the speaker designer has unusual tastes in amplification then it might be a good idea to copy him, as he will have tuned the speaker for his idea of an amplifier.
 
What do you think about these crossover network patents designed to minimize distortion?

US 2841648 Sound producing device
cross coupled phase splitter?

US 4138594 Small dimension low frequency folded exponential horn loudspeaker with unitary sound path and loudspeaker system including same

US 4198540 Compensated crossover network

US 4237340 Crossover network for optimizing efficiency and improving response of loudspeaker system

US 4475233 Resistively damped loudspeaker system

US 4597100 Ultra high resolution loudspeaker system

US 5373563 Self damping speaker matching device

US 5568560 Audio crossover circuit

US 5615272 Single loud speaker drive system
bifilar wound

US 5754667 Audio filter with magnetic field cancellation

Google Patents
 
To speak of Dampening Factor misses the point. It is all about system Q. You can be critically damped over damped or under damped.
These days when folks are speaking of damping things it is mostly the driver plus the box with a target of Q = 0.71. This has the assumption that the amplifier is perfect and stuff up to the box is perfect and add no Q. The amplifier can have a Zo of several ohms plus the zip cord plus the not so passive crossover network.
This is my opinion, the amplifier + zip cord + crossover should be added up then the loudspeaker Q adjusted for the desired Q the user wants.
DT
All just for fun!
 
What do you think about the thesis that designing for good sound will generally result in a high damping factor/low output impedance?
(not for tube or some other exotic designs, of course, he meant typical solid state)
I totally agree. It's actually quite difficult to design a decent amp with a low damping factor.

Conventional SS amps naturally produce lots of distortion. The normal way to reduce that is to apply lots of voltage feedback, which also reduces the output impedance. So high damping factor isn't a major (or even minor) design issue, it's just something that happens.

One minor exception: Most amps have a coil at the output to help with high-frequency stability. This causes the output impedance to rise at high frequencies, so "damping factor" probably drops below 100 as you go up through the treble.
 
Not that simple.
If someone ever think to use passive damper for his loudspeaker, there will be more choice to damp it, not only resistive. Capacitive, inductive, nonlinear or R+diodes(sounded like old solidstate with bootstrap) and its combinations.
My last known best is using both passive damper (parallel with loudspeaker, not in series) and active damper (negative feedback).
 
This is my opinion, the amplifier + zip cord + crossover should be added up then the loudspeaker Q adjusted for the desired Q the user wants.
not just your opinion.
Getting this total resistance correct to match the speaker designer's original Q target is absolutely crucial to getting the best out of the speaker.

If the speaker Designer chose to have 0r1 in the supply leads and amp output impedance at the LF end of the audio then that is what one must try to replicate if one wants what the designer set as his goal.

I have added short lengths of resistance wire in the bass driver supply leads to adjust the Q of the two drivers to match my target and to bring the two driver to match each other. I had already tested the T/S parameters and designed for the worse driver and adjusted the good driver to match. I recall adding 80milliohms (~95mm of nichrome wire) to one speaker lead. And, yes, it is audible.
 
Damping resistive, reactive, or absorbtive measurement?

Some background- most amps are DF speced by placing a resistive load on the amplifier and then measuring the output voltage change. This does not measure damping factor exactly as damping concerns the amplifiers ability to absorb back EMF from the speakers motion. Loading measures output impedance and not input impedance at the output. As an example there is an amplifier here which shows 0.010 ohm output resistance using a load test. A different test is to use the amplifier under test as a virtual ground. In this case a different amplifier drives into a load and that load is then connected to the amplifier under test with no signal applied. The "residual" signal appearing at the output of the amplifier under test is then used to determine damping. The same amplifier above shows almost 1 ohm when tested this way. This situation becomes far worse when the resistive load is exchanged for a reactive load such as a real loud speaker.

With this in mind it is not nearly so clear just what is happening with damping in any given amplifier. As someone else has pointed out damping is also frequency dependent further adding to the confusion. The method of measurement becomes extremely important to the final value derived.

As for the sound, it is my belief real damping is closely related to dynamic range (not signal range as the common error assumes SNR). Dynamic range is the ability to resolve a small signal when a very large signal is present. Lots of testing here has shown amplifiers which have real higher damping have measurably better dynamic range. If the sound one seeks includes lots of dynamic range then the sound is clearly clearer and better. If the sound one seeks is smoothed and undefined being very sweet and forgiving then no feedback at all in power amplifiers is the choice.

Rolland made his mark with low and no feedback amps that are are very sweet and pretty much completely lack definition as do many tube amplifiers. No feedback seems to always give a sweet sound and very nice to listen to but likewise tends toward lacking intimacy and impact found in musical closeness. On the other hand no feedback works perfectly fine in phono stage, line sections, and D/A converters where impedance is very well defined and can be planned for with a completely engineered design. This is not the case for the power amplifier.

Further, amps with high damping tend to run much hotter than amps with low damping as we have found here fully 1/3 of the energy transmitted to the speaker is returned to the amplifier as back EMF. For an amplifier to control this energy requires more energy creating more waste heat so my amps become hotter than other amplifiers. For all you amp guys, no it is not bias change or oscillation, it is back EMF absorption which causes the heating.

So to answer your question, real damping increases dynamic range and clarity as measured and heard. Low damping (per almost all amplifiers) increases artificial smoothness and is more forgiving provided that the amplifier is not oscillating or otherwise misbehaving which includes about 95% of the amplifiers out there.

Hope this does not confuse and somehow answers your posed question.:)