A couple questions about cone breakup...

[shadyJ]

I have been following the Tekton thread with some interest, the conversation about cone breakup was interesting, and I learned some new things there. Instead of throwing that conversation even further off topic for the answers to a few minor questions, I will just pose them here.

I understand the larger a cone is, the worse it will be at playing back a higher frequency, but is there any guidelines on cone size one should generally look for when purchasing a speaker? For example, I have a two-way bookshelf speaker with 8.75" woofers, how low does the tweeter have to go to prevent the cone from breaking up? (The speaker in question is the Behringer 2031p) What are the frequencies a 6.5" driver should stay within? an 8", a 4", and so on. I realize there are a lot of other variables at play here, but is there any rule, perhaps within factors such as cone material composition, that one can use when estimating the frequency range of drivers? I did some Google searches, but I think my own ignorance about the subject is preventing me from finding the information I am after, so that wasn't very helpful. I'm looking forward to being more informed!

 

[TLS Guy]

bb

I have been following the Tekton thread with some interest, the conversation about cone breakup was interesting, and I learned some new things there. Instead of throwing that conversation even further off topic for the answers to a few minor questions, I will just pose them here.

I understand the larger a cone is, the worse it will be at playing back a higher frequency, but is there any guidelines on cone size one should generally look for when purchasing a speaker? For example, I have a two-way bookshelf speaker with 8.75" woofers, how low does the tweeter have to go to prevent the cone from breaking up? (The speaker in question is the Behringer 2031p) What are the frequencies a 6.5" driver should stay within? an 8", a 4", and so on. I realize there are a lot of other variables at play here, but is there any rule, perhaps within factors such as cone material composition, that one can use when estimating the frequency range of drivers? I did some Google searches, but I think my own ignorance about the subject is preventing me from finding the information I am after, so that wasn't very helpful. I'm looking forward to being more informed!

No, there is no rule.

In general pistonic motion, will give you the optimum polar pattern.

However, a very rigid cone will break hard with a big peak.

Here is a very rigid magnesium alloy cone woofer.

Note the huge break up mode between 3.5 and 6 kHz. Note however the very good off axis response in the operating range out to 2 kHz.

Most drivers however flex to a degree by intention.

Soft polypropylene is a good example.

Take a look at this 6.5" polycone woofer.

You can see it is in bend mode after 1.5 kHz, but there is no sudden break up mode, and it rolls off smoothly at the upper end of its range, and there is no eruption like the rigid cone show.

However, as it enters bend mode, the off axis mode deteriorates, especially above 2 kHz.

Obviously the larger the cone the earlier the break up will begin.

If you look at the 10" Magnesium Excel driver it erupts at 2 KHz.

Now look at the same driver, with a doped paper cone.

It starts to break up before 1 kHz, but not as violently as the rigid cone.

 

[shadyJ]

Thanks for the explanation, it is much appreciated! I see now, I suppose that's why my searches weren't coming up with anything. I guess I was looking for some kind of guideline that didn't exist. Now I have another question: I would guess from the differences between the stiffer cone and softer cone material that you would cross over the softer cone woofer to the tweeter at an earlier point but at a more gradual slope, whereas the stiffer cone would get a higher crossover but a much steeper slope. Is this correct, or would the crossover point and slope depend more on other factors? Also, how far down in db does that nasty break-up have to be to be effectively hidden, and does it need to be down as far on the softer cones even though the break-up isn't as severe? Again, thanks for the answers, I am lucky to have such an informed response!

 

[Swerd]

I understand the larger a cone is, the worse it will be at playing back a higher frequency, but is there any guidelines on cone size one should generally look for when purchasing a speaker?

To keep things simple, let's limit this to 2-way speaker designs.

You've already got the general concept, that smaller woofers can go higher than larger woofers, as long as all other things are equal.

If you look at a woofer's Frequency vs. SPL curve, its break-up noise sometimes is visible as a prominent peak or peaks, but other times is not. TLS Guy provided some nice examples of these. In general, softer cone materials start falling off in performance or enter break up at lower frequencies that stiffer cone materials. A very rough order of soft to stiff cone materials is:

Polypropylene
Paper & coated paper
Paper doped with natural or manmade fibers, or wood pulp
Glass, Kevlar, etc. fiber
Aluminum
Coated aluminum (anodized)
Magnesium
​

Also, a woofer's Impedance vs. frequency curve can also show irregularities that occur at or near break up in an otherwise smooth curve. I'll try and find an example.

TLS Guy's examples also each show three FR curves. One is measured on-axis (0° off-axis) and the others are probably 30°, and 45° or 60° off-axis. A woofer's off-axis response deteriorates as the frequency increases, and this usually happens before woofer goes into break up. This feature is actually more important to designing a good crossover than avoiding break up frequencies.

Imaging is generally best when the drivers are radiating in the widest possible manner. As the wavelength produced by a driver approaches the diameter of the driver, the driver starts "beaming". It radiates sound in a smaller and smaller horizontal window in front of the speaker. Stated another way, imaging is the ability of the speakers, when operated in stereo, to accurately place the instruments and voices horizontally within the sound field. You need wide dispersion to achieve this.

(Before GrantedEV chimes in about wave guide speaker designs, I'll say that I'm talking about speakers with standard flat baffles, or 180° wave guides .)

So the more important question is at what frequency does a woofer still produce widely dispersed sound. Look at the FR curves, on-axis, and at 30° off-axis. Pick a crossover frequency that is as high as possible, but where the frequency responses on- and off-axis are within 3-4 dB of each other.

For most woofers, this will be well below the frequency at which break up becomes a problem.

 

[Swerd]

I would guess from the differences between the stiffer cone and softer cone material that you would cross over the softer cone woofer to the tweeter at an earlier point but at a more gradual slope, whereas the stiffer cone would get a higher crossover but a much steeper slope. Is this correct, or would the crossover point and slope depend more on other factors?

Yes you seem to have it right. But generally, the advantage of stiffer woofer cones comes from the more detailed sound they are capable of producing. But because of their often large break up noise peaks, they require a steep crossover slope.

Also, how far down in db does that nasty break-up have to be to be effectively hidden, and does it need to be down as far on the softer cones even though the break-up isn't as severe?

To effectively suppress break up noise, you would need to lower its SPL at least 25-30 dB. With some metal woofers, a 4th order crossover at a frequency well below the noise can get the job done. In some cases, an extra filter, called a notch or LCR filter, has to be added.

 

[shadyJ]

Thank you very much for the further explanations! Interesting stuff about beaming. I take it wave guides and horns make it so you can cross over a bit higher than you ordinarily would with flat baffles? Do wave-guides and horns make it so you can cross over tweeters lower than ordinary too? As an example, take the Klipsch RF-7 which has 10" woofers, I suppose you can't cross over those above 1,356 hz to meet the tweeter, but most tweeters don't go below 2 khz if I remember correctly. Or is compression drivers a different animal altogether? This also brings me back to the Behringer 2031p, with its 8.75" driver, it has a wave guide on the tweeter, does that wave guide let the tweeter meet the woofer before it breaks up or runs int bad off-axis response? Sorry for all the questions, but this is very interesting to me!

I made a chart for myself to get an idea where beaming frequencies would be, it is simply the sound wave frequency frequency lengths are at standard woofer sizes. I suppose you would have to cross over the drivers well below that point to avoid beaming. I'm guessing from the off-axis measurements start to fall away from on-axis from that first graph that TLS Guy posted that it is a 12" woofer? I'll post the chart here for the benefit of anyone else who might be following the conversation, these are approximations I made from this.

18,000 = .75"
13,500 = 1 inch
4,500 = 3"
3,875 = 3.5"
3,390 = 4"
3,010 = 4.5"
2,710 = 5"
2,582 = 5.25"
2,269 = 6"
2,085 = 6.5"
1,695 = 8"
1,550 = 8.75"
1,356 = 10"
1,130 = 12"
904 = 15"
753 = 18"

 

[Swerd]

Thank you very much for the further explanations! Interesting stuff about beaming. I take it wave guides and horns make it so you can cross over a bit higher than you ordinarily would with flat baffles?

Do wave-guides and horns make it so you can cross over tweeters lower than ordinary too? As an example, take the Klipsch RF-7 which has 10" woofers, I suppose you can't cross over those above 1,356 hz to meet the tweeter, but most tweeters don't go below 2 khz if I remember correctly. Or is compression drivers a different animal altogether?

I've largely ignored wave guides, so I'm not able to give good answers. But I can always guess .

Most wave guides I've seen are on tweeters, as in Klipsch speakers or on your Behringers. I don't know a direct answer to your question. I think compression tweeters are different animals than dome tweeters, and that accounts for why they can be combined with larger drivers crossed at lower frequencies.

Most 1" dome tweeters don't go below 2 kHz, but there are well known exceptions. There isn't something inherent in the design that prevents dome tweeters from working at frequencies under 2 kHz, it's a matter of money.

If a woofer is crossed over at frequency at which it beams, people can hear an audible "disconnect" between it and the widely dispersed sound from a tweeter. If I understand the purpose of wave guides on tweeters, they limit the wide dispersion of a tweeter so that there isn't such a noticeable "disconnect" between its sound and that of the beaming woofer. Overall the speaker's dispersion isn't as wide as it might be, but the dispersed sound is more uniform across the crossover range.

A wave guide, or horn, also has the ability to make a tweeter more sensitive, via the megaphone effect.

I made a chart for myself to get an idea where beaming frequencies would be, it is simply the sound wave frequency frequency lengths are at standard woofer sizes. I suppose you would have to cross over the drivers well below that point to avoid beaming. I'm guessing from the off-axis measurements start to fall away from on-axis from that first graph that TLS Guy posted that it is a 12" woofer? I'll post the chart here for the benefit of anyone else who might be following the conversation, these are approximations I made from this.

Consider that list of numbers only as a general guide. Remember that for many woofers (but not all) those diameter numbers are the outer diameter of the metal frames and not the diameter of the moving cone.

The real measure of a woofer's dispersion comes from frequency response graphs made at several different angles on and off axis.

The manufacturer should supply these measurements. They are often made on very large baffles or "infinite" baffles. A good speaker designer always measures a woofer's FR and dispersion again with the woofer mounted in a cabinet with the intended front baffle width. The existence of a front baffle will change the shape of the FR curve from what the manufacturer provides.

 

[Swerd]

This seems like a good time to remind people of one of my favorite web pages.

It explains, in a very general way, some of the steps a crossover designer takes while doing his thing: Crossovers 101

 

[GranteedEV]

Thank you very much for the further explanations! Interesting stuff about beaming. I take it wave guides and horns make it so you can cross over a bit higher than you ordinarily would with flat baffles? Do wave-guides and horns make it so you can cross over tweeters lower than ordinary too?

It gets tricky.

The most sensible crossover region, is one where the midrange and tweeter have a similar directivity index.

If you have a tweeter in a waveguide, it's directivity index is raised (or in other words, its wide angle off axis dispersion is converted to forward dispersion). This lets you raise the midrange crossover frequency - pros being improved tweeter power handling, reduced early room reflections, and the fact that most speakers already cross their drivers over at these frequencies anyways - without the directivity match.

If you have a woofer near 90 degree beaming, it's however highly likely that it's near breakup. The 90 degree waveguide does allow you to cross higher from an off-axis response perspective, but that doesn't really "solve" the issue of breakup.

The only way to solve breakup is to have very well behaved drivers AND good crossovers. Something like a horizontal BG Neo10 Planar or something like Harman's Ceramic Metal Matrix Diaphragm, or maybe B&W's FST midrange.

The only way to have both supressed breakup and smooth power response is to use acoustically small drivers and keep the pattern around 120 degrees. Basically something like the Salk Soundscape. But then you can't expect high efficiency or output in the midrange from a 4" driver. Also, with wide dispersion patterns, you have a lot of side wall reflections if you don't treat them. Great in a wide room where the side wall reflections have been delayed a good 15ms (about 6 feet) but less useful in a narrow room.

Overall, since the waveguide does improve efficiency (by redirecting the side wave forward), you can lower the crossover frequency slightly below the "exact" directivity index matching point and still have a smooth behavior. When I say efficiency is improved, I mean that the tweeter itself has to move less to produce more output.

What to do? Same with a speaker on a flat baffle - pick whatever frequency just "works" as there is always a compromise. The waveguide just minimizes the compromise.

Most wave guides I've seen are on tweeters, as in Klipsch speakers or on your Behringers. I don't know a direct answer to your question. I think compression tweeters are different animals than dome tweeters, and that accounts for why they can be combined with larger drivers crossed at lower frequencies.

A compression tweeter is really just a larger dome or ring radiator firing into a smaller narrowing exit with a phase plug. A 1" exit compression tweeter is probably using a 1.75" or 2" dome. That's why it can play lower (that... and the GIGANTIC motor! )

That's why the best compression tweeters - IE Radian 951 - use 3" beryllium diaphragms firing into 1.4" or 1" exits. That 1.4" exit + Be diapgragm give you treble extension like a 1.4" inch flat piston tweeter, the waveguide will determine the off-axis response, and the 3" diaphragm, big voice coil, and huge magnet let you cross over right around 700hz or so. As far as wide bandwidth drivers that might be the best out there. Not cheap though - these things are 900 Euros each:

 

[ridikas]

Why would anyone delete my post because I said the OP's question was a very loaded one? It is!

 

[shadyJ]

Why would anyone delete my post because I said the OP's question was a very loaded one? It is!

Lol my question? I didn't see your original post but if my question is loaded, believe me I do not realize it. I am just trying to learn a few things here!

 

[ridikas]

Very much so Your Behringers measure right on the money:

http://www.zaphaudio.com/temp/B2031P-system-FR.gif
http://www.zaphaudio.com/temp/B2031P-system-horizontal-offaxis-0-15-30-45-60.gif Horizontal.
http://www.zaphaudio.com/temp/B2031P-system-vertical-0blue+15red-15green.gif Vertical.
http://www.zaphaudio.com/temp/B2031P-tweeter-offaxis-0-15-30-45-60.gif This is how important the waveguide is.

 

[shadyJ]

Lol I take it they are good speakers. To think I got four of them for $260! I don't understand what you mean by the importance of a wave guide with respect to that last graph. Can you explain?

 

[ridikas]

In baby terms, a proper waveguide can lower a tweeter's distortion and allow you to cross it lower.

 

[shadyJ]

It gets tricky.The most sensible crossover region, is one where the midrange and tweeter have a similar directivity index.

I had to look up directivity index (this site had a decent explanation). One question that brings to mind is why are so many woofers concave in shape, and so many tweeters are convex? I suppose the reason for a concave shape is the energy gets dispersed too widely from a dome-ish shape, so the speakers just won't get loud? But, if so, why are tweeters dome shaped? And here is the stupidest noob question of all time: why wouldn't you just want all of them to be flat drivers?

 

[shadyJ]

In baby terms, a proper waveguide can lower a tweeter's distortion and allow you to cross it lower.

ah, I see!