speaker response

[trem]

I have been reading about speakers, but most of the info deals with physical design of the cab, or cross-over design.
My question deals, primarily, with a single 12 inch, guitar speaker.
What accounts for the difference in frequency response.?

Just for example.....
Eminence:
Governor
and
Wizard

All else equal, why do they reproduce the same frequencies at different amplitudes.?
Is it the type of paper, voice coil, ribs in the cone.?
Thank You

[Jazz P Bass]

As to those 'ribs' on the cone, I have learned that they help in radiating the treble frequencies.

[guitician]

I have found that the "Spider" that holds the top of the voice coil centered has as much to do with the response of the speaker as the cone itself. In fact, it could be considered a second cone that is in a push-pull harmony with the cone.

[Enzo]

Everything in the speaker has a contribution to its sound. The cone is made of some material that has a stiffness/flexibility about it, that's a factor, the cone may be striaght sided, or it may have a curve to it - so cone shaped or horn shaped so to speak. The cone is smooth or ribbed. The dust dome is large or small, and made of various materials. The spider could be stiff or pliable. The surround could be stiff or pliable. Those things both can allow the cone excursion to be tightly constrained or allowed to run, and for that matter the cone might be relatively free to move within part of its movements, but be stiffer the farther it goes, that sort of thing. The voice coil sits in the magnet gap, so we have the whole magnet thing, how large is the gap, how deep is the magnetic field. How strong is the magnet. The voice coil itself, is it single layer, multi layer? How far along the form does it extend, does it leave the gap at excursions or does it always stay in the gap?

Those are just a few things I could think of off the top of my head, and I don't pretend to know much about speakers. Someone like Juan who makes the things for a living will know a hell of a lot more about it than I do.

[trem]

Per usual.....Thanks Enzo (and Jazz and guit).
I guess my real question, in simple "layman's" terms (if that is possible) .....what are the Main Considerations of speaker design that effect tone.?
If you had a speaker, and you liked it, but thought it had too much Bass response.....what would you change.?
Do some things effect some frequencies more than others.?
That is to say.....does the Cone Material have more to do with Bass, than does the Spider, or do the different parts have an equal affect on what we hear.?
Uhhg.....you know what I am trying to ask.....Thank You

[Enzo]

I'll wait for Juan or someone to chime in.

But let me return your question in another way. Lets say we were talking about amps, and the question was what makes the tone of an amp? Of course there is no one thing. Maybe an OT is more conducive to bass in an amp, but the preamp is set up for treble response. You could put in that OT or reconfigure the preamp, either one might improve your bass. Or other factors like powr supply filtering might be more important in the particular amp. There is no real this-does-that lookup list.

All those speaker things do not operate individually. The speaker itself is a system, an electromechanical system. Look at the cone. If i double the thickness of the material, it makes it stiffer, but it also increases its mass. And those things when considered alone might have different effects on overall tone, but taken together, it may well change things. I am sure there are some general tendencies that go into making a bass speaker over a guitar speaker or a PA speaker. Hifi speakers tend to be a lot less efficient than instrument speakers. They sacrifice efficiency in return for a smoother frequency response. Or at least I believe that.

You ask what we would change if we thought a speaker was too bassy. Can I assume you are asking acedemically as a speaker designer would? As opposed to something we users could do to our own speaker?

[J M Fahey]

Regular paper cone speakers have dozens of resonant peaks all over the place , from the lowest to highest frequencies, which usually have different amplitude; those resonances also create dozens of frequency "dips" .
Reason is that usable cones are very thin and flex all over the place, like it or not.
Similar to what happens with new strings, where besides the note you are playing, there's lots of harmonics to begin with which are related to the main note, plus lots of much higher ones which are not "musical" by themselves but give the string character or voice.

The published frequency responses that you posted in fact are "averaged" or "smoothed" ... in part to avoid scaring customers .... but are in fact much rougher :

on the left, real "as measured" frequency response, so much detail that in practice it's useless to the designer itself, so the measurement system, or the software used, "averages/smooths" that unreadable mess into something like what you see on the right :


in fact it's a miracle (not kidding) that speakers put out recognizable sound at all, in such a wide range.

Where do all those peaks and dips come from?

As Enzo said, every speaker part contributes its own sound:

* Cone/diaphragm: paper/aluminum/polypropylene/titanium ; paper mix "recipe" (different proportions and kinds of: cellulose paste + cotton/hemp/synthetic fiber + glue + carbon black (soot) + additives) , thickness , drying method (natural/pressed/heated/steamed/vacuum) , straight/curved , smooth/ribbed , also rib variations: amount/depth , etc.

* Voice coil wire: copper/aluminum , round/flat.

* Voice coil former: paper/Nomex/Kapton/aluminum/fiberglass.

* Voice coil winding: short/long , single/2/4 layer.

* spider diameter, stiffness.

* dust cap: cloth/paper/plastic/aluminum/felt/whizzer.

* magnetic circuit: larger/smaller magnet , thin/thick plate , wide/tight gap , how well was it magnetized or how much it lost since originally made (typical of 50/60 y.o. Alnico speakers) , etc.

As you see there are tons of interacting variables, which nobody (yet) could condense in some piece of design software, so all they can attempt to do is, as you noticed, some generic box tuning or designing crossovers, which at least are more predictable.

Speakers in practice, beyond the magnetic circuit which can be designed and simulated in FEMM Finite Element Method Magnetics: HomePage , all other parts are empirically designed which requires lots of experience.

I knew a former Jensen (USA) Engineer who told me: "I could go to the spare parts room and get dozens of cones / voice coils / spiders/ you-name-it and build lots of different samples for customers to approve ; the main problem was not Technical but big customers (car makers) wanted to pay no more than 5 or 6$ a speaker ... we had to oblige because typical orders run a million speakers each" .

Once made, there's not much you can do to modify its sound, except building a better cabinet to improve deep bass response, or what some do: add "beam blockers" to pad some annoying highs.

Hi Fi speakers get flat frequency response by using thick cones (that kills many resonances), and to boot use relatively soft/"spongy" paper which contributes to the same, also long voice coils which can move without leaving the magnetic gap, and soft edges (cloth/sponge/rubber) .

Big problem is that with all that damping speakers become very inefficient, no big deal in an apartment but a deal breaker if you have to match a drummer , so typical guitar speakers have:

* thin, light/hard, pressed/steamed cones ... which resonate like crazy, undamped, all over the place.
* short voice coils so they are always "pushing" ... but which will distort a lot if you feed deep bass to them
* strong magnets
* light adhesives (to minimize moving deadweight)
* stiff spiders (to keep voice coil tightly in the gap)
* narrow gaps to minimize magnetic loss

in sum, are fully geared for loudness, efficiency, nothing else matters, so frequency response is very rough.

PA speakers sit halfway, need reasonably flat response (although not to Hi Fi standards) and high efficiency.
Those contradicting requirements are met by making very expensive speakers, with *huge* magnets (think JBL, EV, Black Widow, etc.) , huge voice coils (up to 4" diameter, compare to typical 1" to 1 3/4" popular with guitars) , very expensive flat ribbon aluminum voice coils wound on space age formers, glued with space age adhesives, on precision machined cast frames: $$$$$$$$$$$$$$$

look at the very unflat guitar response:

first a typical Eminence one:


then a typical Jensen one:


Low frequencies: as-is most (all?) drop below 100Hz ; there is where better box design can improve things a little.

Mid frequencies: guitar speakers usually have a characteristic peak around 1 kHz which makes them "forward sounding", and most 12" speakers a strong dip around 1500Hz, no matter what brand they are.
That's because cones flex, and when voice coil moves forward, the rest of the cone moves later, and at a certain frequency cone edge is so much delayed that it's actually out of phase, and it "kills" that frequency, which happens to be around 1500Hz.
One way to reduce that is to dope (cover with a thick sticky paste) the edge.

Above that guitar speakers have a *strong* mid/high peak , usually between 2500 and 3500Hz which gives them "bite" and above 3500/4500Hz they drop catastrophically.

If you add a tweeter to reproduce higher frequency sounds the cabinet becomes unbearably buzzy, unusable.

For comparison, here's PA speaker response, think EV or "Greek" Eminence (Beta/Gamma/Kappa/etc.) .
Much smoother but not usually considered a Rock/Blues speaker,preferred by some Jazz/Fusion players though (think Larry Carlton, etc.):


see the monster magnet and cast frame it uses, plus the (hidden) expensive voice coil and tight machining:

[trem]

Enzo -
Yes, you are correct, I was asking a Theoretical/Academic question.
A....."what if"...kind of question.

Juan -
Great info up there...Thank You
Speaker design, as far as tailoring to a certain frequency, is a lot more complicated than I realized.
I kind of thought the cone was responsible for 95% of the freq respns of the speaker.
I imagine computers are a big assist in designing a speaker today. Must have been more difficult to do in (circa) 1970.?
Thanks Again

[guitician]

What has always perplexed me about guitar speakers is that the manufacturers of the "Best" ones like JBL D120, or Altec 417 went out of production. That's like GM saying back in the late 60's that they were no longer going to make the 350 V8 engine. Or Gibson saying that they will never make another Les Paul guitar!

Another thing is that paper coned speakers are still the very best for audio reproduction of the very highest quality. Some that come to mind are WE 755, Jensen 610 and Altec 600-B

[Enzo]

The best guitar speaker in a line may not be the best seller, but even if so, JBL probably made a lot more speakers for other applications than just guitar amps. SO perhaps instead of Gibson stopping the Les Paul, maybe more like Gibson no longer making the best guitar for polka bands.

[guitician]

OK, profit aside then, but Eminence did their 15" clone of the D130 for Lap steel players, and Great Plains Audio is doing old Altec stuff, so I guess it's just a matter of the well running dry for the big guys to keep up with.

[trem]

OK, profit aside then, but Eminence did their 15" clone of the D130 for Lap steel players, and Great Plains Audio is doing old Altec stuff, so I guess it's just a matter of the well running dry for the big guys to keep up with.

Well.....as with amps, how many "new" speakers are really available anyway.?
Not sue of the Exact Frequs...but what comes out of a 12 inch speaker .....200Hz to 6kHz maybe.?
How many meaningful manipulations of those frequs make a "real" difference.?

JMHO, but....... Seems like 5-10 speakers would cover just about most possibilities for a guitar amp.
My only experience with JBL for guitar is in a Fender BF Twin Verb. I realize they would handle the power of a Twin, better than most anything else of the day, but I always thought they were just average as a speaker for a guitar amp.
Anyway.....after reading the reply's in this post, I doubt I will be mastering the art of Speaker Design any time soon.....or anytime at all for that matter.

[Enzo]

How many meaningful manipulations of those frequs make a "real" difference.?

Well, look at the current market of 12" guitar speakers. You will find fans and haters of each one. Eminence makes several, they don't all sound alike. And of course Celestion makes several, again each with fans. Jensen. Etc. They all sound like guitar speakers, but then again they all sound different.

Real beer is made with only four ingredients. Yeast, malt, hops, water. How many combinations could they make of those? Thousands, check your local import/craft beer store.

[trem]

Well, look at the current market of 12" guitar speakers. You will find fans and haters of each one.

Funny you should say that.
Seems like No Speaker gets as much "hate" as the Celestion Vintage 30. I do not know for sure, but I get the feeling it outsells any other speaker by a Big Margin.

[Enzo]

yeah well, McDonalds outsells all other burgers, but I don't consider it very good food. They sell more Velveeta than any real cheese brand, I'd wager. Yellow goo.

yep, a lot of people don't like the V30, but a lot of them do. I have a Fender 2x12 cab clone with a pair of them in it.