ClassD/T damping factor.

[oon_the_kid]

Hi,

I am using a tripath sure electronics to drive the bass part of my speaker system. However I find the sound to be lacking in control. I suspect that it could be caused by low damping factor? Does anybody have any idea what the damping factors of Class D amps are?

I am presently using a TK2050 from surelectronics part for the job.

Oon

[kristleifur]

Damping factor should be high with these amps.

The output stage transistors' RdsON is about 0.2Ω; Paralleled they ought to be 0.1Ω. (AFAIK this is the biggest contributor to damping factor, and pretty low in this case ... ?)

But I think that's not related to the issue — The TK2050 has very good control of the lower registers. (Incredible control actually.) In 41Hz Amp4 and Amp11 the bass is dynamic, fast, taut, deep, clear, and warm. Switching to a Tripath usually sounds like you've added an octave down below or put it into focus ... IMO bass control is one of the best features of a good Tripath.

I don't want to be a party pooper but it's probably a problem with the Sure design :/ ... rather than the chipset. I'd guess it's a current handling problem due to board design or component selection. Might be the input capacitors or output filter too?

[Saturnus]

Originally Posted by oon_the_kid

Hi,

I am using a tripath sure electronics to drive the bass part of my speaker system. However I find the sound to be lacking in control. I suspect that it could be caused by low damping factor? Does anybody have any idea what the damping factors of Class D amps are?

To answer the last part of your question dampening factor could be anything just like any other type of amp. It would depend on the specific amp used. However, dampening factor would normally be lower than a comparative class A/B amp but the difference would be marginal in nominal dampening factor and would not matter in real life conditions at all.

However, the amps dampening factor rating means very little as it depends on the speaker cable resistance and thereby also filter coils if there are any. The higher the total series resistance, the worse dampening factor will be. And it also depends on the impedance of your speakers. The lower the impedance, the worse the dampening factor will be.

Lack of control in the bass also depends on many other factors that are more important than dampening factor. Speaker placement for example will greatly affect bass control. Placing the speakers close to one or more walls will make bass control disappear completely. Also incorrect cabinet tuning for the listening room could be the cause. A speaker designed for a very large room, Qtc ~1, will lack bass control in a very small room for which a Qtc ~0.5 would be optimal.

[oon_the_kid]

Hi Saturnus,

Thanks for the tip. Forgot to mention, that I am using it to drive subwoofers. And on a different matter, different speakers require different amount of damping factor, fullrangers require very little, but unfortunately for this case, subwoofers need a lot more.


Had a look on speaker cables, I suspect mine might not be thick enough. On a different matter the reason I brought up this thread is the damping factor for most of the class D used here is unknown I believe. I looked up the datasheet of Tripath chipsets, its not written there either, meaning it could be anywhere between 10 and 1000...

Oon

[sharpi31]

Have you bypassed or replaced the Sure TK2050 board input coupling caps? If not, this is where to start. The stock units roll off deep bass leading to a lack of audible bass weight. The situation is even worse if you also have output coupling caps in your preamp.

Amplifier:Sure Electronics TK2050 2x100W - Modification - diyAudio

[MOER]

Damping Factor - This amplifier specification has been blown out of all proportion. What it means is the ability of the amplifier to resist a change in it’s output voltage. The formula is DF= Speaker Z / Amplifier output Z (where Z is impedance). So many manufacturers have claimed ridiculous, and often false damping factors. A damping factor of 1000 implies that the output impedance of the amplifier is 0.004 ohms (4 ohm load). The only way to attain this figure is to apply masses of negative feedback (or use positive feedback) and too much feedback makes amplifiers sound harsh and clinical. Also damping factor changes with frequency. The lower the frequency the higher the DF number. Typically the DF can be ten times larger at higher frequencies.

Let us take this amplifier whose output impedance is 0.004 ohms (Zout). The speaker circuit is a series circuit and the following impedances(resistances) are in series with this 0.004 ohms. Let us assume that this DF measurement was made at the amplifier’s speaker terminal. The first extra contact resistance is the speaker wire to the speaker terminal (WT ohms). Then there is that of the wire itself for two conductors (W). Next is the contact resistance of the wire to the speaker terminal (WS). Next there is the contact resistance of the wire from the speaker terminal to the voice coil (WV) and lastly there is the DC resistance of the voice coil itself (DCR). So what we have is a series circuit with the following resistances in series and adding up. WT+W+WS+WV+DCR+Zout. WT,W,WS,WV and Zout are very small indeed. Certainly less than 0.1 ohms. Whoa, look what has happened the EFFECTIVE DAMPING FACTOR has been reduced from 1000 to 40 by just taking into account those pesky unavoidable contact resistances. Now for the cruncher, remember that the DCR is also in series and is typically 3.2 ohms for a nominal 4 ohm speaker. So we must add 0.1+3.2 = 3.3 ohms and now EFFECTIVE DAMPING FACTOR is now a magnificent 1.212! (4 divided by 3.3)This is the real world. We see that the DCR of the speaker swamps all other resistances in the speaker circuit and the 0.004 ohms amplifier output impedance is almost meaningless. It has been found that a DF of about 20 is quite sufficient to dampen the voltage spikes from the speaker. An eye opener this one is it not? Good tube amps sound marvelous - low damping factors!!

Zed Audio

[ChocoHolic]

Originally Posted by MOER

Now for the cruncher, remember that the DCR is also in series and is typically 3.2 ohms for a nominal 4 ohm speaker. So we must add 0.1+3.2 = 3.3 ohms and now EFFECTIVE DAMPING FACTOR is now a magnificent 1.212! (4 divided by 3.3)This is the real world. We see that the DCR of the speaker swamps all other resistances in the speaker circuit and the 0.004 ohms amplifier output impedance is almost meaningless. It has been found that a DF of about 20 is quite sufficient to dampen the voltage spikes from the speaker. An eye opener this one is it not? Good tube amps sound marvelous - low damping factors!!
Zed Audio

...well, I do not agree on your definition of the damping factor, that you define to be the ratio of nominal Z by Rdc..
But I full agree that the sum of Rdc in the entire loop is the key for system damping and no doubt the Rdc of the voice coil itself is dominant.
In passive systems the next huge portion is the choke of the cross over.
But let's stay without passive filters, otherwise the discussion about high damping factors turns ridiculous anyway
Looking to Thiele Small parameters and speaker physics:
The mechanic Qm of a speaker is typically between 5 and 15.
But the electrical Qe is somewhere in the range 0.2 and 0.8.
And total Q results as Qts=(Qe x Qm)/(Qe+Qm).
The electrical damping is achieved by the speaker motor, say the BL product and the Rdc of the electrical loop (valid for woofer resonance frequencies, for midrange and tweeters also the complex portions might turn important)
As long as the output impedance of the amp + wiring is small vs. the Rdc, it will just have minor impact on Qe and consequently have just minor impact on the overall system damping.
Means, the output impedance of most transistor amps + wiring ( let's stay below 100mOhms) will not become visible when measuring the damping of the system.
That's the technical view.

Unfortunately this does not mean, that different amps could not sound different in bass region.
Sonic impression of a system does not necessarily follow simple school book theory. (Might be worth to extend the theory in a scientific way!)
I am still wondering about a phenomenon of my Rookie amps.
When you drive them with a low impedance source (below 100 Ohms) they give a pretty controlled, deep bass. Rookie can show unexpected authority for such a smallish amp!
When you drive the Rookies from a higher impedance source (2kOhms) they turn extremely shy, or even weak and at the same time less relaxed.
Especially looking at the Mosfet version of the Rookie I do not find anything in the measurements for bass reproduction!
Lower roll off frequency unchanged, damping factor unchanged - but a giant difference in sonic impression.
Very strange. If it wouldn't be a system that I measured on my own work bench - I would not believe it.

[Saturnus]

Originally Posted by MOER

Now for the cruncher, remember that the DCR is also in series and is typically 3.2 ohms for a nominal 4 ohm speaker.

There's a number of incorrect points in you summation.

Dampening Factor = Load resistance / Output resistances

That is the actual formula. If you then want to specify dampening factor at a certain frequency you use the impedances at that frequency.

So nominal dampening factor = Rvc (voice coil resistance) / ( Rout + Rser)

The interesting part with dampening factor is the difference at impedance peaks and lows at cabinet tuning frequencies. For example in a ported cabinet dampening factor should in principle be infinite at the the tuning frequency as the speaker driver in principle should not move at all at this frequency.

Anything over effective dampening factor 10 is quite acceptable.

[oon_the_kid]

Hi Moer,

I am not sure if the DCR of the voice coil should be added in to the whole equation, under that definition, damping factor of all system would be 1. Since it will be completely dominitated by the coil itself.

I believe the definition of damping factor is the output voltage of the amplifier without load versus the amplifier with load (difference). The ratio of output resistance is actually the derivation of that part. Althoguh I do agree that the ouput of the amp resistance is probably much smaller than the speaker cable... I suppose that is why people tend to build power subs, to keep the cable short...

Oon

[AP2]

Hi,
As always, many do take advantage of its publicity in the damping factor or other parameters. yes, I agree that a tube amp has a very low DF (40).
in fact, a df of 100 is enough for good speed.
In class D, a high DF is obtained by a forced use of the FB. (which I do not like).
Truth is, the MOSFET has a 0.02 rds-on, but you have to consider the output filter.