Driver Mechanical, Electrical and Total system Q

[btnhfan 5,000+ posts]

Man you need to pick up loudspeaker design cookbook 7.

 

[JonJT 10+ year member]

I am sure you are going to get the worst answer for people like me and you....you have to test because it's application specific. I am sure you can model it for chambers, but then they would be worthless as your application would change the results.
I tried to request a model be build about predicting impedance rise, as the factors involed grew, I knew I couldn't model it to predict.

Yeah empirical data rules supreme, but I'm also not looking to predict actual real world performance here, and neither are you. The theory behind it all may be well off, but I'm still interested in hearing it. I'm sure those in the know would love to provide it with a nice disclaimer! //content.invisioncic.com/y282845/emoticons/smile.gif.1ebc41e1811405b213edfc4622c41e27.gif

 

[JonJT 10+ year member]

Man you need to pick up loudspeaker design cookbook 7.

I guess I'll add it to the list of all the chassis design books I need to pick up!

 

[thch 10+ year member]

How so?How can you model it?

just keep adding to it. you can of course try to best-fit the T/S model to the system, but the T/S model won't show distortion.

as you've seen, BL is a part of Qes. As the cone moves from rest, BL drops. Cms also changes as the cone moves from rest. So now Qms, Vas, Fs, and Qts have also changed. the T/S model is most accurate in the region where such changes to Qes,Qms,Fs,Vas,Qts are small.

Also, in terms of modeling, EVERY mechanical element that can transfer energy through the cone will be part of the electrical circuit. thus ports will affect the speaker, as will reflected sound. ported and TL boxes work on these concepts respectively.

 

[JonJT 10+ year member]

O, its that simple then? Well I guess modeling woofers at full power or excursion really isn;t too accurate then.

 

[MiniVanMan 10+ year member]

O, its that simple then? Well I guess modeling woofers at full power or excursion really isn;t too accurate then.

I can be if you plan to utilize them at full power.

There are no easy answers here, and the only concrete information that can be given are the specs themselves. You can go to 10 different sites, all modeling the same drivers, and get 10 different sets of T/S parameters, and they'll all vary from the stated manufacturers specs. Manufacturers even have a hard time being consistent within the industry. If you tested Vifa drivers at the Seas facility there's a good chance the numbers would come out differently than Vifas original numbers.

In other words, take all of it with a grain of salt. There definitely should be a solid understanding of what it all means, but honestly, none of it is truly indicative of how well a driver will sound in a given application. Knowing this information at least lets you narrow down where to start when trying to implement a driver.

Remember, there is no spec that can overcome a poorly designed system.

 

[newusername 10+ year member]

O, its that simple then? Well I guess modeling woofers at full power or excursion really isn;t too accurate then.

To an extent, the capability is there. Aside from the Klippel machines limitations with measuring inductance, it is great for measuring BL and Cms over stroke. Take a gander at some of the measurements done by npdang in the Reviews section over on http://www.diymobileaudio.com (and I imagine you'll find your answers better addressed there or on SSA than you will here, speaking in generalities).

If we can measure a driver's linearity and prove that there is little variation in BL or Cms over the range of our intended use, the small signal thiele/small parameters will hold truer. However, this still does not consider any thermal power compression, nor does it necessarily represent what will sound pleasing to the ear. Though I am a big fan of designs that minimize odd order distortion, it is worth noting that even the most linear products are still occasionally poorly received. Drs. Earl Geddes and Lidia Lee have really been bringing this concept to light.

edit - Something I often forget: I am looking at this from an engineering standpoint. For most customers that a company would cater to, the ability to corroborate any high-level testing is slim-to-none and, even worse, the desire to do so is extremely minimal. Your average consumer wants to go to BestBuy and buy what they are told is the best.

 

[MiniVanMan 10+ year member]

Though I am a big fan of designs that minimize odd order distortion, it is worth noting that even the most linear products are still occasionally poorly received. Drs. Earl Geddes and Lidia Lee have really been bringing this concept to light.

Great point.

When modeling drivers and testing them, distortion figures don't really mean a lot either. A very nice sounding driver is the Vifa XG18, but has considerably higher distortion than other like priced drivers (Seas RNX series for example). However, it sounds very warm, and pleasing, and can be found in "audiophile" grade loudspeakers.

Low distortion drivers have a tendency to be sterile to many ears, and that's the "poorly received" part. Paper coned woofers with low distortion like the Scan Revs are very clean sounding, while retaining good warmth, but also quite expensive.

However, many of the same audiophiles that complain of a sterile driver also run tube amps with tons of distortion, and don't know the difference between "warmth" and distortion.

So, again, take all of it with a grain of salt. Subjective listening tests, in enclosures modeled as closely as possible to optimal is really the only way to tell if you like a driver or not.

 

[JonJT 10+ year member]

Well, I've been in the habit of separating theory with subjetive measurements like sound quality. Certain woofers, alignments, speakers or headphones are taken to sound good despite their T/S parameters or any other technical information one can provide about them. From there I'd listen to whatever I am/was interested in and pick what I liked.

The largest impetus for my inquiries is simply to increase my understand of how these machines work and perform from a strictly technical standpoint. I don't have plans to use the theory behind a resonating system as a justification for good sound quality.

 

[JonJT 10+ year member]

To an extent, the capability is there. Aside from the Klippel machines limitations with measuring inductance, it is great for measuring BL and Cms over stroke. Take a gander at some of the measurements done by npdang in the Reviews section over on http://www.diymobileaudio.com (and I imagine you'll find your answers better addressed there or on SSA than you will here, speaking in generalities).
If we can measure a driver's linearity and prove that there is little variation in BL or Cms over the range of our intended use, the small signal thiele/small parameters will hold truer. However, this still does not consider any thermal power compression, nor does it necessarily represent what will sound pleasing to the ear. Though I am a big fan of designs that minimize odd order distortion, it is worth noting that even the most linear products are still occasionally poorly received. Drs. Earl Geddes and Lidia Lee have really been bringing this concept to light.

edit - Something I often forget: I am looking at this from an engineering standpoint. For most customers that a company would cater to, the ability to corroborate any high-level testing is slim-to-none and, even worse, the desire to do so is extremely minimal. Your average consumer wants to go to BestBuy and buy what they are told is the best.

Alright capacity is there, but its not really an exact science (but then again, what science is exact when in practice).

The caveat about the average consumer is something I know all to well. Its quite amazing how many people don't care about how the product they are buying works. They just want it to work even if its far from the best they can get.

 

[JonJT 10+ year member]

I'm having trouble understanding how Vas impacts woofer parameters and how enclosure size effects alignment Q.

Earlier in this thread, it was made clear that a lower Q system is a lossy one. There is a significant restoring force that dampens what would be (theoretically) periodic motion of the cone.

Here is Eminence's definition of Vas quoted verbatim:

Vas represents the volume of air that when compressed to one cubic meter exerts the same force as the compliance (Cms) of the suspension in a particular speaker.

Through reading other threads on other forums, I've seen many people state explicitly that a lower Vas value implies a lesser Cms value. I cannot understand how this can be. If a smaller (comparatively speaking) number of moles of air is contained within 1 cubic meter, Ideal Gas Law states that pressure MUST be less than a comparatively larger number of moles of air (assuming constant humidity and temperature). I do not understand how the lower number of moles of air, with its lower pressure can posses a higher "spring constant" (Cms, with inverted units) than a larger number of moles of air. I can just look at a bunch of woofers t/s parameters and see that I'm missing something. Take Soundsplinters RL-S line for example. Cms value half that of Soundsplinters other offerings of the same woofer diameter, yet a TINY Vas value. What gives? What am I missing/not understanding here? Maybe some level of ambiguity in the definitions I'm using?

My second question is very much related to the first because it has to do with the effective "air spring" sealed enclosures place on woofers.

Further reading on other boards as well as use of modeling programs has brought me to the conclusion that smaller enclosures increase system Q; they decrease effective dampening and make the cone a better resonator. Why is this the case? Comparing the effect of cone movements in two different enclosures of wildly different net volume, you will see that the smaller enclosure will have a much greater variance in internal pressures. This equates to greater restoring force on the woofer cone, especially during high excursion, high volume (high cone acceleration) playback. Why in the world would this create a system that stores energy better (once again, comparatively)?

Thanks

 

[JonJT 10+ year member]

..

 

[JonJT 10+ year member]

Anyone?

 

[Papermaker85 5,000+ posts]

VAS represents the Amount of air compressed in a specific space(one meter cubed) that is takes to equal the same stiffness that the diaphragm/soft parts comply at rest. larger spiders and cones have more surface area and present a higher VAS for the same CMS(greater efficiency). this is true because your larger drivers have more surface area, hence they can displace more air and have less stiffness per inch^2 against the moving parts at rest for equal CMS over smaller drivers...

simple explanation... take a tire. both same sized tires. one tire has 25PSI and the inner volume measured one cubic foot... the other has 30PSI and it also measures 1 cubic foot in volume... the tire with the lower VAS is simply the tire that has the higher PSI... why? because the amount of volume is equal(SD), but the compression of air is greater for that given surface area( lower Vas) there for the tire with the higher PSI lends to have a higher Quality or is HARDER to distort(lower CMS) from its original shape. now in woofers.....

when the woofer resonates its trying to have the correct amount of damping to have the most efficient movement possible with a given input and a given amount of airspace.. having too much stiffness(low vas) causes the driver to want to deceleration before optimal excursion limits are reached where as too soft or too high a VAS it doesn't have the ability to properly stop the mass to change direction..

did that help a bit?

 

[thch 10+ year member]

Further reading on other boards as well as use of modeling programs has brought me to the conclusion that smaller enclosures increase system Q; they decrease effective dampening and make the cone a better resonator. Why is this the case? Comparing the effect of cone movements in two different enclosures of wildly different net volume, you will see that the smaller enclosure will have a much greater variance in internal pressures. This equates to greater restoring force on the woofer cone, especially during high excursion, high volume (high cone acceleration) playback. Why in the world would this create a system that stores energy better (once again, comparatively)?

Thanks

The addition of the box amplifies the pressure air exurts on the cone for a given amount of excursion. its like adding more spring to the cone. this lowers compliance.

Looking at the equations for Qms and Qes and Qts, this has the effect of causing changing all three. Qms increase, Qes increases, and Qts increases.

Does this make sense? yes. The system has added more energy storage elements, but has not changed lossy elements. keep in mind that at resonance, where Q is defined, energy is efficiently transfered from compliance to inertia -- the mass of the cone and the compilance of the box/spider are in harmony so to say.

and as compliance drops due to the small box, the resonant frequency increases as well.

as for another example, lets say you pulled the cone outward to some arbitrary position. the cone will now have MORE potential energy due to the increased spring constnat. further, it requires more force to hold the cone outward, thus when released the spring will exurt more force on the same amount of mass, causing it to move to rest faster. because the spring has no potential energy when the cone is at rest, and because the the mass has a higher acceleration, it is clear that the cone will reach its rest point faster and with a higher velocity. Thus higher Fs is expected. without increased losses, the cone will tend to wobble for more cycles.