Which driver will have the faster Transient Response?

[Oz_Audio]

I have been going over Unibox to try and find a driver that will give me a fast transient response.

In all cases, there are 2 x 10 inch drivers in a 122 lt closed box.

Driver 1:
2 x 4 ohm drivers wired in series.
Voice Coil Inductance 0.25 mH
QTC 0.542
fb 35.96

Driver 2
2 x 8 ohm drivers wired in parallel
Voice Coil Inductance 0.40 mH
QTC 0.483
fb 34.97

Driver 3
2 x 8 ohm drivers wired in parallel
Voice Coil Inducctance 0.35 mH
QTC 0.559
fb 38.59

As max spl, fb, f3, impedence and frequency curves are almost identical, which will have the better transient response?

Mark

[Svante]

What is your definition of "transient response"?

Is it perceptual or technical?

What are the criteria that makes it "better".

Without having answers to the above, the question is impossible to answer, sorry.

[Oz_Audio]

Transient response as I want is the least time delay of the driver becoming stable (not moving)after the electrical impulse has finished. Just listen to any Bass Reflex design regardless of who makes it!

I don't want a wobbling driver making sound long after the electrical impulse has stopped.

Two things are indicators of this, Voice Coil Inductance (as low as possible) and in box Q (0.5 to 0.55 max).

For this application, Transient Response is paramount. 2 of the drivers are good at 1 of the parameters, but not both (in regards to each other only). The third is in the middle for inductance, but the worst Q.

Mark

[Svante]

Ok, so with that definition a system with a 20 Hz lower cutoff is half as good as a system with a 40 Hz lower cutoff, all other parameters being identical?

For these cases, the 20 Hz system has a twice as long impulse response.

I ask this because I find it terribly hard to extract any useful information from measured impulse responses, and that the perception of transient capabilities in many respects are different from simple measures extracted from the impulse response.

I beleive that what makes comparison between impulse response and perception of transient capabilities so hard is that whenever a change in the impulse response is accomplished, there is also a change in the frequency response. So, when we do the comparison, how do we know that the impulse response is The measure to use?

...but of course, this can never be anything more than my personal opinion. I don't mean to kill this thread, how about the others, do you have similar or other thoughts?

[Oz_Audio]

Svante, I think you have missed the point.

Its not about the frequency used, but the time taken after the event, or the number of oscillations that represent Transient Responce.

Each of the 3 drivers are all expectional in this regard, but as they are all $400 AUD units, I do not want to waist money in experimentation when I am sure, someone on ths forum has far greater knowledge then I.

My primary goal to distinguish the best from this selection in regards to the speed of the driver, nothing else.

Mark

[Svante]

Originally posted by Oz_Audio
Transient response as I want is the least time delay of the driver becoming stable (not moving)after the electrical impulse has finished.

Originally posted by Oz_Audio
Svante, I think you have missed the point.

Its not about the frequency used, but the time taken after the event, or the number of oscillations that represent Transient Responce.

Ok, that was not what you said the first time... I think. Time is ususally measured in seconds, and for the 20 Hz system, the oscillations will go on for twice the time, for sure...

But, ok, if time is measured in number of oscillations, this number is the same for both systems and largely determined by the Qtc value in case of a closed box. Bass-reflex boxes are more complicated. The voice coil inductance, however would have very little impact here, unless for cases with an extremely high Le.

Le will, however have an impact on the frequency response, and this impact may very well be perceived as affecting the transient capabilities of the system.

[Ed LaFontaine]

I can't offer you the explanation you need to answer your question. A path you might try is to read Linkwitz about his low frequency transform. One of his observations was that using the Linkwitz transform to improve lower frequency response also improved transient response.

[AKN]

Hi Oz_Audio,

Transient response as I want is the least time delay of the driver becoming stable (not moving)after the electrical impulse has finished. Just listen to any Bass Reflex design regardless of who makes it!

In trying to narrow this a bit.
Is this a sub that is going to be LP filtered above lets say 100Hz?
If so the Le differences you are showing will not be of any significance for drivers ability to stop.

[DSP_Geek]

The systems are fairly close in Fb and Qtc, the predominant factors determining transient response. Including normal manufacturing tolerances, it's likely differences between the three choices are within the margin of error.

In any event, a lower cutoff frequency implies a longer impulse response. Why this seemingly contradictory result? Ideal single impulses contain all possible frequencies in the spectrum; real-world impulses used to measure speakers are constrained by frequency limitations in the electronics. Nonetheless, if you present a wide-spectrum impulse to a loudspeaker, the transducer will reproduce what it can, and the lower it goes, the longer the time response will be.

http://en.wikipedia.org/wiki/Dirac_delta_function


Qtc also affects transient response. It turns out Butterworth tuning (Qtc = 0.707) shows the fastest response. Lower Q shows a slower return to zero after the undershoot, and higher Q shows ringing, with both effects more noticeable the further you get from 0.707.


There really won't be a lot of difference between all three systems, anyway. If you have access to response plotting tools, you'll find the differences are quite mininal, perhaps a dB or two at most around the corner frequency. Plotting the transient response will show the same characteristic, with little difference in time-domain behaviour.

[kelticwizard]

Your Qtc's range from 0.48 to 0.56. Below is a link to a graph of "overhang" in the step response for various values of Qtc, (and Qts in free air).

As you can see, the difference between 0.5 and 0.7 is not overly large. The difference between 0.48 and 0.56 will be but a third of that.

Overall, I don't think you'll hear very much difference.

There is a whole discussion on Qtc and "tightness" here.

And here is the image. It is from Richard Small, (of Thiele-Small), in his article on Closed Boxes: