Woofer rated SPL vs graph

[johngalt47]

I was looking at different woofers and noticed something I don't understand. On more than one of them, the SPL was rated at a certain value (let's say 88db) but the graphs don't show them reaching 88 db until more than 200 or 300 hz.

Shouldn't they be rated at the range where they will be used (say, 30 to 200 hz)?

[AMC]

Boy are you opening a can of worms. Everyone would have a different opinion on this, so here is mine. One caveat going in, when you talking about woofers, they are so box dependent below 200Hz it's hard to standardize in that region, still they should use an optimal box design and go from there.

IMO, the rating should be the LOWEST response in the intended usable range. Reasoning like this, when you go to design your crossover, you can't add to the response, only take away. So when you are done smoothing out the response, the overall response would be at the low point.

[bangleiii]

Thruth in advertizing went out in the 80's!

[GranteedEV]

I was looking at different woofers and noticed something I don't understand. On more than one of them, the SPL was rated at a certain value (let's say 88db) but the graphs don't show them reaching 88 db until more than 200 or 300 hz.

Shouldn't they be rated at the range where they will be used (say, 30 to 200 hz)?

A lot of those graphs are taken on an infinite baffle. the qtc = qts and fc = fs with a 12db/octave rolloff.

Typically a vent or transmission line or passive radiators will be used to extend usable response down to near fs. so a 7" driver that begins to roll off near 200hz on an IB will instead be flat down to near 60hz in a vented box with 24db/oct rolloff below tuning.

Further, sensitivity will take another hit in a baffle due to the baffle step.

[billfitzmaurice]

Shouldn't they be rated at the range where they will be used (say, 30 to 200 hz)?The graphs are in a standard baffle and don't show what response is in a cab. Below roughly 150Hz the cab determines response and sensitivity.
There is a mathematical formula to determine sensitivity based on T/S specs. It only applies within the region of pure pistonic function, just as T/S specs do. Some box programs calculate it for you, WinISD Alpha Pro, for instance. In any event you should be using a box program to model response, and then any manufacturer hi-jinx will be revealed.

[johngalt47]

Ok, here is the deal:

1) It is hard to measure a woofer's freq response below 200hz

2) Even if you did, it would be expensive to buy every woofer

3) We can't rely on a manufacturer's specifications to accurately determine a woofer's performance

With those factors in mind, how does one go about choosing a woofer to mate with the other drivers to design a speaker?

[DanP]

You can get a good idea of a speaker's reference efficiency using the formula Bill is talking about: 112 + 10*log((Vas*Fs^3*2.7*10^-8)/Qes) with Vas in cubic feet. The value you get will be in dB/1W/1M. This will give you a more accurate number than the manufacturer's spec sheet as long as the T/S parameters they provide are accurate. Keep in mind this is an efficiency number and not sensitivity. You will need to account for its impedence when figuring out the actual sensitivity to match up to a mid or tweeter.

The FR at 2.83V at 1m tells the real tale, however. If the graph is accurate, you can make a very good assessment of how to use the driver. You have to trust the publisher of the graph at that point though. What I do is use the formula above to make sure that my woofer is the least sensitive driver in the system. This way I never have to pad my woofer (a big waste) and I can pad my mid/tweeter easily to match.

Dan

[billfitzmaurice]

how does one go about choosing a woofer to mate with the other drivers to design a speaker?
You model it in a box program.

[fastbike1]

That's lame and unfounded for this topic. When you model the woofer in my experience it virtually always shows an efficiency very close to the spec. This is one of the few specs that you can probably consistently believe.

However as Bill said, you really don't have any excuse not to model the driver in one of the box programs. Takes no more than a minute to input parameters for a quick look.



Thruth in advertizing went out in the 80's!

[johngalt47]

My original premise included the fact that manufacturer's specs are often not correct and I can't buy an example of every driver to test myself.

Therefore, putting the specs into a program to do the calculations may not deliver a solution that would work with the actual driver I purchase based on incorrect specs.

[ROTECH]

my Original Premise Included The Fact That Manufacturer's Specs Are Often Not Correct And I Can't Buy An Example Of Every Driver To Test Myself.

Therefore, Putting The Specs Into A Program To Do The Calculations May Not Deliver A Solution That Would Work With The Actual Driver I Purchase Based On Incorrect Specs.

Amen

[Deward Hastings]

When you model the woofer in my experience it virtually always shows an efficiency very close to the spec.Likewise when you build it. The models are not complex, and boxes, in general (and "closed" especially), are pretty forgiving of "manufacturing variation" in the drivers. The paramaters from PE and Mad and most of the manufacturers are, with only the rarest exception, good enough to accurately predict behavior (and easy to test, if you feel the need).

Two things are not so simple: it is hard to get harmonic distortion (linearity) numbers, which matter a lot because of where the sub's output lies on the Fletcher Munsen curve, and the room the sub is used in. Even allowing for the different "sound" of "closed" and "resonant" boxes those two things account for almost all the "differences" in the sound of subwoofers (with matching WinISD curves), and with woofers other differences don't begin to appear until you get up to first cone breakup.

Regarding the "rated SPL" phrase in the thread title, that is ambiguous. If you mean "maximum SPL" you also have to specify frequency (drivers are excursion limited (which is also box dependent) at the low end, and thermally limited at the high end), acceptable distortion, and driver power handling. If you mean "sensitivity" (dB/Watt) then the number will almost always represent the high end of the drivers range, because it only matters at crossover (and for passive crossovers at that), and becomes box (resonance) dependent at the low end, so a "bare driver" spec would be meaningless.

I agree that manufacturers specs generally don't tell you everything you need to know (meaningful distortion figures are uniformly missing) . . . but in general they do accurately provide sufficient information to predict SPL and sensitivity in any particular application.