Damping factor

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Does anyone know how much of damping factor could I expect from bridged TDA7293 amp ?
I am planning to build two of these amps for my new active subwoofer project and the speakers I have need big damping factor to sound right.
I need about 120W of undistorted output per channel on 8 ohm speaker and a lot of damping factor.
 
It is ability of amplifier to control the speaker cone. It is the opposite of output impedance (DF=1/Zo). When DF is very low speaker cone will not stop moving when polarity of input signal changes. Than bass loses strength.

You can calculate DF of your amplifier with this formula:
DF = U1 / (U2-U1)
U1 = output voltage of your amplifier with speaker connected
U2 = output voltage of amplifier with same input signal, but speaker disconnected.
It is best to use sinewave signal generator for this purpose.

DF is important when using big speakers in big boxes (like subwoofers)
 
b<ben>n said:
this is probly a stupid question but what is damping factor and what dose it do.

This is kinda one of my 'hobby horses' - if you know what I mean. Let me explain:

A value that is calculated by the ratio of the Load Impedance (suposedly the speaker) and the output Z of the amp. Say the load is 8 Ohm and the amp's output Z is 0.1 Ohm, ergo 8/0.1 = 80; the damping factor is 80. It is of rather inflated importance IMHO.

Solid State has high DP. Tubes, especiallly zero or low f/b has low DP. But is it related to sound quality? Not really.

I am highly suspicious about that DP measured under static conditions (which is the only way to measure it) has very little with DP under dynamic conditions. Rather than low static output impedance, what is more important is stable or predictable output Z ????

Besides, DP is often the end result of applying feedback, so more f/b and we get lower DP. Too good to be true? You bet!

These days few really mention DP. What is more important is perhaps how high output Z modulates the frequency response of loudspeakers. Here, strangely, even though Tubes are supposed to be pour in this respect, they usually shift the response around less than one might expect. Wonder why?

Joe R.
 
I've never understood the point of chasing damping factor.

The output resistance of an amplifier, the result of a finite damping factor, certainly does exist - but it's still incredibly small compared to that of the winding resistance of the speaker it's driving.

And amp resistance, speaker cable resistance and winding resistance produce the exact same result - they act as one big resistor in series with the reactive impedance* of your speaker which is the part of the speaker's overall impedance that actually produces music. All energy that goes into the winding resistance of a speaker only produces heat.

*eg, the L + L||R||C part of a Thiele-Small model.

so why isn't anyone trying to make 0.1 ohm speakers? hell, even going from an 8 ohm speaker to a 4 ohm speaker would be a gross improvement over trying to trim the output resistance of an amp from 0.02 to 0.01 ohms. And attempting to accomplish the latter by excessive feedback cound render an amp unstable and bad sounding...
 
Have you ever tried to knock on the bass speaker cone and listen the sound when speaker is disconnected and when speaker leads are shorted? You will hear a big difference in depth. The same is with amplifiers. Amplifier with low output impedance acts like a short circuit to speaker. In that case movement of speaker cone will follow input signal with no delay when changing directions.

The example of this situation are my sub and sub of my friend. This 2 subs use the same speaker in the same enclosure. One uses 120W amp with big damping factor, and other 120W amp with low damping factor. Sub which uses amp with lower DF makes much bigger amplitudes (about 50%), but with no audible benefit It sounds too soft and reaches it's bottom with only half available power applied. The other one wants to rip down my walls and is able to stand the full power from the amp without reaching the bottom.
 
diyAudio Member
Joined 2003
Cro maniac said:
It is ability of amplifier to control the speaker cone. It is the opposite of output impedance (DF=1/Zo). When DF is very low speaker cone will not stop moving when polarity of input signal changes. Than bass loses strength.

You can calculate DF of your amplifier with this formula:
DF = U1 / (U2-U1)
U1 = output voltage of your amplifier with speaker connected
U2 = output voltage of amplifier with same input signal, but speaker disconnected.
It is best to use sinewave signal generator for this purpose.

DF is important when using big speakers in big boxes (like subwoofers)


Yeah thats prety much what ive been told before too when i used to make diyamps about 8-10yrs ago. Basically the higher the DF the more responsive the speaker is, its rather important in subwoofers, especially when listening to high integral music like drum and bass. Running anything with high integrety and a low DF will destroy your speakers, especially over 100w continuos rms.

Trev
 
Cro maniac said:
It is ability of amplifier to control the speaker cone.

That's how it's commonly described. But it seems to me that it's not so much a case of the ability of the amplifier to control the speaker, but rather the ability of the amplifier to allow the speaker to control itself.

"Control" ultimately refers to the speaker's resonant behavior. If you have a speaker with a Qts of say 0.7, it'll have a Qts of 0.7 whether it's hooked up to an amplifier with a source impedance of zero Ohms or simply has its terminals shorted and no amplifier hooked up to it at all.

Similarly, if that same speaker has a Qts of say 1.2 when the source impedance is say 2 Ohms, it will have that same Qts if it simply has a 2 ohm resistor across its terminals.

Also, if you drove the speaker from an infinite source impedance (i.e. an ideal current source), its Qts would end up being its mechanical Q, Qms and would have that same Qts if its terminals were left an open circuit.

se
 
Steve Eddy said:


...not so much a case of the ability of the amplifier to control the speaker, but rather the ability of the amplifier to allow the speaker to control itself.

"Control" ultimately refers to the speaker's resonant behavior. If you have a speaker with a Qts of say 0.7, it'll have a Qts of 0.7 whether it's hooked up to an amplifier with a source impedance of zero Ohms or simply has its terminals shorted and no amplifier hooked up to it at all.

Similarly, if that same speaker has a Qts of say 1.2 when the source impedance is say 2 Ohms, it will have that same Qts if it simply has a 2 ohm resistor across its terminals.

se

Hi Steve

There are good reasons to have low Z when amp is driving a sub. But has it got anything to do with damping factor? I think you might on the right track here as I don't really think so either.

You mention a driver Qts of 0.7, but that is a component of Qe and Qm. Let's say the Qe component is 0.8, then a little playing with a calculator will tell us that Qm = 5.5

Now let's say that the source impedance is high at 1 Ohm. Measure the DCR (Re) of the driver, and let's say it's a 4 Ohm sub driver and Re = 3 Ohm.

The 1 Ohm is in series with the 3 Ohm and this erodes the Qe by the same ratio:

0.8 x (4/3) = 1.066

Now a bit more fingering on the calculator will tell us that the original Qts has eroded from 0.7 to 0.9 !!!

If the sub is sealed (good subs are) then the sealed box alignment has changed. If we put all the parameters into a suitable program, then analyse closely the cone amplitude/excursions, we see clearly that cone excursions at critical frequencies are now considerably larger.

What's this got to do with damping factor? Not much, and this is why: The output Z of the amp and the Re are in series, so in reality the best real case scenario is ONE - or unity damping factor.

So damping factor really is a myth. It's an arbitrary mathematical curiosity that has no real world import.

On the other hand, Low Z is important to keep bass alignments on track.

Oh, mea culpa, what is DP? Damping Phactor ? :cannotbe:

Joe R.
 
roddyama said:

Eh... Sorry Joe, all.:blush:

Speaking of Small. I remember many years ago having a conversation with him, I think he was at Sydney University in those days. I was disagreeing with him re his idea of of adding a series resistor to modify Qe (and hence Qts) and tuning its value to vary Qts as well as increasing Vb of a sealed box.

I was of the opinion that this would ruin the damping factor. Now I have a different opnion altogether.

But it does show that Small realised that adding resistance does no more than varying the Re of the driver and by using a larger Vb, one could get a targeted Qc and lower Fc at the same time. So the use of a larger box, corrected Qc (higher due to resistor), lower Fc, you could get much more extended bass. The sacrifice is lower sensitivity/efficiency. But this is then an alternative to mass loading of the cone, as used by manufacturers like Dunleavy.

But again, this example shows that DF is a myth and Small had already discarded it in those days.

Joe R.
 
phase_accurate said:
DF is not a myth, it is simply one way to express an amp's output resistance. Nothing more, nothing less.
The expression is somewhat misleading and furthermore the value shown is just valid under specific conditions (e.g. a static signal into a real load).

Charles

I would say that the amp's output Z is not a myth, it's real alright. And it is irrespective of the load, whereas 'damping factor' is the load divided by the output Z. Why should that supposedly be of any importance?

And what significance is it when the output Z and the Re of the driver becomes one and the same thing? The fact is the Re is many times that of output Z, so Z is not to be seen in isolation. Add to that the cable's resistance, the choke (as used in passive LPF), all are in series, all are really added to Re.

So damping factor is a myth.

Just to prove it, try this example:

We have an 8 Ohm bass driver. Let's assume it has Re = 6 Ohm. We now short the terminals together, to simulate an amp having zero output Z. Do we now have infinite damping factor? No we don't. Here is why:

If we were to push the cone in and out, we would generate current into a dead short. Here we are simulating back EMF which should be shortened out by the zero output Z, hence creating damping. But this doesn't actually happen, since the current has to travel through 6 Ohm of resistance.

Joe R.
 
So damping factor is a myth.

No it isn't ! Only what some people (e.g. scanting salespeople) make out of it is a myth. It is a simple measurement value like THD, output power etc..... It is howerver questionable if it should be mentioned at all.

A DF value of 10,000 does for sure look much better on paper than one of 200 even though the increase in damping is marginal (less than 0.5 % for a given Re and Qtc !).

DF = Rload / Ri

Regards

Charles
 
phase_accurate said:


No it isn't ! Only what some people (e.g. scanting salespeople) make out of it is a myth. It is a simple measurement value like THD, output power etc..... It is howerver questionable if it should be mentioned at all.

A DF value of 10,000 does for sure look much better on paper than one of 200 even though the increase in damping is marginal (less than 0.5 % for a given Re and Qtc !).

DF = Rload / Ri

Regards

Charles

Hi Charles

I think you might have missed my last edit. Can you explain it?

Joe R.
 
I haven't missed your statements. But you seem to have missed that my opinion regarding DAMPING is basically the same as yours.

Damping FACTOR isn't defined as damping of the driver, it is the value given by Rload/Ri of an amp (It is DEFINED like that, this has nothing to do with personal taste or opinion, it is just a definition).
So an amp representing zero Ri would in fact give an infinite DF. It doesn't mean in turn that the damping of the driver as such is infinite. It is mainly magazine writers and salespeople that claim things like that.

One could easily be misled by the fact that very often amps with high DF have indeed very good control in the LF range. Since a high DF is not that easily achievable, the better control might as well be caused by the generally higher effort put into design and construction of some of these amps. One parameter might be better ability to deliver large currents into rective loads or a more signal - independant behaviour of Ri (who said that it is the same for all load conditions ?).

Regards

Charles
 
phase_accurate said:
I haven't missed your statements. But you seem to have missed that my opinion regarding DAMPING is basically the same as yours.


Hi Charles

OK then, if our opnions coincide, then if you don't agree that DP is a myth, then perhaps we can agree that any claimed benefits are at least overstated.

Would you also agree with Lynn Olson's 'hard amp' concept.

It goes like this:

The output impedance, instead of approaching zero, is intentionally set to the impedance of the driver, - a damping factor of one or unity coupling. This requires realigning the bass enclosure... The impedance seen by the speaker driver is constant and equivalent to a fixed resistor. This is the essence of the 'hard amp' concept: the amplifier mimics a passive component under all operating conditions... This is very different from conventional [high feedback] transistor or PP pentode amplifiers, which enter undefined regions when the amp clips and feedback loses its grip on the amplifier... a 'hard amp' avoids gain transitions and treats back-EMF like a fixed low value resistor.

The output Z does not need to be quite that of the speaker's, but a high one of 2-4 Ohm is not a huge problem. Incidentally, when designing vented boxes I aim for one with Bessel like characteristics, as they are much more immune to high source Z than, say, 4th order Butterworth. Also low Q Sealed box with Fb below 50Hz, likewise copes quite well.

May I also refer to the point Richard Small (of Thiele-Small fame) made to me, that adding series resistance might actually improve bass. Are we in agreement here?

Joe R.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.