• WARNING: Tube/Valve amplifiers use potentially LETHAL HIGH VOLTAGES.
    Building, troubleshooting and testing of these amplifiers should only be
    performed by someone who is thoroughly familiar with
    the safety precautions around high voltages.

Can someone explain damping factor

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Yes I suppose so, although the resistance you use will probably be only a fraction of the speaker coil resistance, so I suspect the stability change won't be very much in most circumstances.
Yes, a fraction of the speaker impedance but don't forget the transmission line effects of the speaker cable. The latter is a concern for wide-bandwidth NFB amplifiers. Also, noise pick-up.
I reckon for most home systems an output resistance of <0.5 ohm is fine.
 
planet10 said:
The twiddles is the symbol for "proportional to”.
No it isn't. Twiddle means 'approximately equal to'. The symbol for 'proportional to' looks a bit like a Greek alpha. See Wikipedia.

DF is a measure of output impedance, scaled by nominal speaker impedance. It only has much relevance at low frequencies, so it is the low frequency output impedance which should be used. However, as output impedance is quite likely to be inductive it is really only the resistive part which counts. What DF you need depends on the details of the speaker design, and what sort of bass you like.

What DF you achieve depends on amplifier architecture and amount of feedback. You can't say '15dB of feedback gives DF=N'; you can say '15dB of feedback with this particular circuit gives DF=N'. DF beyond about 20 makes little difference.
 
Damping Factor (DF) on it's own is not important.

Source impedance/resistance seen by the speaker voice coil is what really matters.

DF = Rated speaker impedance divided by the output impedance of the amplifier at the specified frequency.

One has to add on all the other resistances between the amplifier output and the speaker driver terminals.
Starting from the amplifier output one sees that the speaker current travels through the sequence:
speaker leads1 , amplifier output inductor, speaker terminals1, speaker leads2, speaker terminals2, speaker leads3, crossover inductor, speaker leads4.
One should sum all these resistances and add them to the amplifier's output impedance at the resonant frequency of the speaker driver to end up with the resistance seen by the driver.
You will find that the "other resistances" swamp the amplifier output impedance.

Let's assign some numbers to that list.
speaker leads1 5mr
speaker leads2 10mr
speaker leads 3 20mr
speaker leads4 15mr
speaker terminals1 10mr
speaker terminals2 10mr
crossover inductor 25mr
amplifier output inductor 12mr

Amplifier output impedance @ ~50Hz. 30mr. Gives a DF = 8ohms/0.03ohms = ~267
total resistances seen by the speaker driver = 137mr (add on the other terminals in the route)
Damping effect of all those resistances relative to the 8ohms speaker driver = 8/0.137 = 58

The DF of 267 bears no resemblance to the effective damping of 58

Ignore DF and look instead at milli-ohms of effective resistance in the whole route from amplifier to driver and back to the amplifier.

Now as an exercise, see what effect on the effective damping there is by changing the amp's DF from 100 to 1000

I reckon that DF was invented by a copy writer to try to make his employer's amplifier look good. "see how good I am, I have made your numbers look bigger than all your competitors"
 
Last edited:
Damping Factor (DF) on it's own is not important.

Source impedance/resistance seen by the speaker voice coil is what really matters.

DF = Rated speaker impedance divided by the output impedance of the amplifier at the specified frequency.

One has to add on all the other resistances between the amplifier output and the speaker driver terminals.
...

...total resistances seen by the speaker driver = 137mr (add on the other terminals in the route)

I agree with everything you say so far. But you also need to add in the DC resistance of the speaker voice coil as well. This is needs to be added in to the denominator too. And subtracted from the numerator. It counts just as much as the other series resistances or impedances, as adding to the effective source impedance. And it completely dominates the total source impedance.

The effective equivalent circuit, simplified, is an ideal zero-resistance voice coil in series with a total resistance R(tot), and then driven from a zero impedance source. R(tot) is the sum of all the things you listed, plus the DC ohmic resistance of the voice coil. And this last item, being several ohms, is far bigger than all the other contributions to the effective source impedance that you listed. In the numerator should go just the effective impedance of the ideal zero-resistance voice coil in its dynamical setting as it oscillates in the magnetic field.
 
Last edited:
The back emf tries to drive the source impedance.
Short the drivers input terminals for a zero effective source impedance and you have a fully damped voice coil.
It's the same as regenerative braking of an electric motor. One shorts the motor for fast deceleration without any lock up. Slot cars were a great example of this.
 
{snip} ...
I reckon that DF was invented by a copy writer to try to make his employer's amplifier look good. "see how good I am, I have made your numbers look bigger than all your competitors"

You are close, Andrew. DF was first proposed as a measure of amplifier quality in the late 1950's/early to late 1960's when the first transformer-less transistor amplifiers entered the market, vs transformer-coupled (mostly vacuum state) amps in competition. Some of the advertising contained long treatises on why the new type of amplifier was superior to the old, damping factor being one, a factor that was essentially ignored in previous marketing literature.
 
Last edited:
So how do you explain the big difference in (controlled) sound between a high df amd and a low df amp on high powered subs then?

It is "explained" by the choice of LF driver, in closed systems (powered subs).

Properly designed sub systems do not necessarily have problems with control, regardless of the DF. I do not know of sub frequency systems that fail with "control" due to DF alone.

Also it is easier to design a good bass amplifier, so why not low measured distortion with feedback / high DF, as the sonic issues (high order distortions, odd order distortions) are out of the operating frequency bands, while the advantages are just sitting there waiting to be used.

Same reason why sub amps are inevitably solid state; it's the proper approach to high power, which is needed as frequency and driver motor requirements move toward 1 Hz.

It is like a tractor vs a moto GP motorcycle. There is no point in designing either with the technology of the other.

In other words it is done because they can, without repercussions.
 
Last edited:
The back emf tries to drive the source impedance.
Short the drivers input terminals for a zero effective source impedance and you have a fully damped voice coil.

It depends what you mean by "fully damped." It is certainly as damped as you are going to be able to get, given that the voice coil has its own ohmic resistance.

Consider an extreme example where we replace the copper wire of the voice coil by an equal number of turns of high resistance wire, so that now the measured resistance of the voice coil is, lets say, 100 ohms. To say that this speaker would be "fully damped" if one shorted its terminals together would be misleading, to say the least!

The point I am making is an elementary one, which is easy to see if you think about the equivalent circuit. All the items you listed, PLUS the DC ohmic resistance of the voice coil, should be added in the denominator.
 

TNT

Member
Joined 2003
Paid Member

Attachments

  • approx.png
    approx.png
    89.1 KB · Views: 245
It depends what you mean by "fully damped." It is certainly as damped as you are going to be able to get, given that the voice coil has its own ohmic resistance.

Consider an extreme example where we replace the copper wire of the voice coil by an equal number of turns of high resistance wire, so that now the measured resistance of the voice coil is, lets say, 100 ohms. To say that this speaker would be "fully damped" if one shorted its terminals together would be misleading, to say the least!

The point I am making is an elementary one, which is easy to see if you think about the equivalent circuit. All the items you listed, PLUS the DC ohmic resistance of the voice coil, should be added in the denominator.

Just be be clear: There are two distinct measures of damping that one could discuss:

1) The quantity customarily called "the damping factor" is, by definition, the impedance of the speaker load divided by the source impedance.

DF = Z(load)/Z(source)

(The source impedance should include speaker leads from the amplifier, etc., etc.) With a modern solid state amplifier the source impedance is tiny, the load impedance is, say, 8 ohms, and "the damping factor" is huge. And that huge factor is essentially meaningless.

2) A proper measure of damping looks at the total impedance that the emf from the voice coil is driving into. This gives a true measure of how damped the speaker will actually be. One looks at the equivalent circuit to figure this out. Roughly speaking, neglecting details of adding reactive impedances and treating everything as if it were purely resistive, this "true damping factor" will be something like

TDF = Z0/Z(tot),

where Z(tot) = Z(source) + R(load), and Z0 is roughly Z(load)-R(load), where R(load) is the DC resistance of the speaker coil.

In other words, Z0 is essentially what the speaker impedance would have been if there were no DC resistance in the coil.

Of course this "true damping factor" is typically very much smaller than the usually defined damping factor, because the DC resistance of the speaker, R(load), is typically much bigger than R(source), and is a major component of Z(load).

That Radio Electronics article that was linked earlier in this thread essentially says these things in Part 1 (e.g. see the middle of column 1 on page 2.) Then in Part 2 it talks about the usually defined damping factor, and essentially says it is of limited value.
 
Does anyone have the formulas in a handy text?

Good topic Guys,

Seems these days the assumption is that the amplifier output impedance is low enough not to affect the total Q of the amplifier plus the Q of the speaker/enclosure combination.

The speaker enclosure calculators available will calculate the Q of the speaker/enclosure system, nothing is included for the influence of the amplifier. Q = 0.707 is a quasi-standard.

I like merlinb’s thought of adding a series resistor or the idea of stuffing the enclosure tuning port with a little fiberglass insulation and tuning by ear.

It gets more complicated trying to tune tube amplifier, speaker and enclosure.

Do any of the calculators like WinISD include a variable for the nonzero output impedance of a tube amplifier?

Does anyone have the formulas in a handy text?

DT
 
Glad you guys like the discussion

However since I am a novice (and so most of this has gone over my head), all I was looking for is a recommendation on the amount of GNFB. Again I'm using approx, 15db. I only mentioned my vintage JBL's in case they have an impact, damping factor wise.

I will definitely try to improve the wording of any future questions.
 
Can someone explain damping factor
I understand that it increases as GNFB increases. But how would one determine how much is needed. I have very efficient older JBL's with the 15" D130 LF drivers in smallish vented cabinets. Would these need more or less DF? I haven't done the calcs, but I believe that I am running about 15db of GNFB.


The JBL D130 are over damped speakers you do not need a very high damping factor. A damping factor of 10 to 12 will be ok; a very high damping factor would make very little difference. What tubes are you using? 15 db GNFB would be plenty with Triodes; KT88, KT66, 300B.and lots more.
 
planet10 said:
4 years of honours math at Uni. We used ~ a lot… and it/was is on a keyboard.
People sometimes use 'shorthand'. This can vary from place to place. There is a difference between 'twiddle means proportional' and 'some people use twiddle to mean proportional because it is quick and convenient while the correct symbol is hard to find'.

Like many, I often use 'u' for mu when on this forum as typing 'uF' is much quicker than having to look up how to insert Greek letters.
 
Can someone explain damping factor
I understand that it increases as GNFB increases. But how would one determine how much is needed. I have very efficient older JBL's with the 15" D130 LF drivers in smallish vented cabinets. Would these need more or less DF? I haven't done the calcs, but I believe that I am running about 15db of GNFB.

The JBL D130 are over damped speakers you do not need a very high damping factor. A damping factor of 10 to 12 will be ok; a very high damping factor would make very little difference. What tubes are you using? 15 db GNFB would be plenty with Triodes; KT88, KT66, 300B.and lots more.

My amp is a modded variant of an ST-70, using EL34 outputs. Also, how does GNFB numerically relate to DF. I assume they are not equivalent, therefore there is probably a formula which relates them, I just can't seem to find it.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.