simulation question

[schiller]

Hi to everyone. Recently, i was simulating some 15'' prodrivers in winisd, till a friend suggested to use his JBL speakershop software he has. We sat together and i was pleased to see that the sims gave identical results/optimal boxes with winisd. They must use the same equations.

Nevertheless, to make it more precise, i used the "acoustic responce " option of the JBL software with the data from the published FR curve of the manufacturer- it was the Eminence deltapro 15''. These are taken on open baffle from 1m distance.

The difference was huge!! the -3db point moved 20-25 Hz upwards!! I repeated the procedure with other Louspeakers - Beyma15LW30, Fostex FW405-, it was always much worse (15-25Hz up) wenn i used the published infininite baffle responce. instead of the default setting.
Can someone tell me if i am doing something wrong??? I really don't know what to conclude from this.

Thanks in advance
Konstantinos

[nunayafb]

If I am undestanding what your results are they sound right. IB always causes a woofer to roll off sooner than an optimum ported/sealed.

[richie00boy]

No it doesn't, infinite baffle will always allow a speaker to go lower than sealed.

And for sake of the abuse of the term infinite baffle lately, I mean a massive baffle such that the speaker is effectively in free air.

I'm not quite sure what the 'acoustic response' option is exactly, but beware of mixing measure frequency response curves up with other models. What size baffle was the speaker measured on?

[schiller]

Thanks for the reply. I suppose i didn't describe it accurately enough.

In both cases it was the ported box simulated. Once with the default setting of the program (identical result with winisd) and the second run using the published FR of the speaker (taken on IB, according to the datasheet) as a basis to calculate the box-loudspeaker system output. This is an option provided by the program, if the coresponding table with the FR is filled (in 3Hz steps)


Regards
Konstantinos

[nunayafb]

No it doesn't, infinite baffle will always allow a speaker to go lower than sealed.

I thought it was clear that we were talking about when it rolls off but maybe not. In IB the Qtc of speaker in baffle is same as the Qts of the driver, however in a sealed box in a maximally flat alignment the Qtc is .707. So the sealed box rolls off lower and steeper than IB if the Qts is less than .707. Ported boxes roll of much later and much steeper.

[richie00boy]

You are right about the Q, but I think you have overlooked the fact that Fs will be raised by the sealed box more than the gain had by having Qts optimal, usually.

[nunayafb]

Out of curiosity, what does fs have to do with anything? The topic here is why does his IB fr from the manufacturer have such a high f3 compared to the in box prediction by winisd.
Regardless of what changes the end result is that with all speakers designed for "normal" sized boxes (as opposed to putting a high Qts driver in a 40ft^3 box) the f3 will raise up when on an infinite baffle, whether it is truely infinite or not.

FYI this is an adire extremis, winisd calc. optimum ported box, and the IB is a sealed box of a couple thousand cubic feet. IB requires about 10X Vas to behave as IB.

[nunayafb]

and an IB dayton 15" in a sealed and IB situation notice Qtc is raised by using an enclosure designed for .707 alignment. This woofer was designed for IB, and the difference is trivial but it's there.

[Salas]

If the responses are taken on an IEC baffle and not flush in a wall then its normal you see less lows.

[schiller]

Well, thanks to everyone for the contributions. I have looked the issue a little closer and this is what i think it happens:

First the relevant text from the help menu of the program:

"By adding the measured acoustic responce of the loudspeaker to the amplitude responce of the box desing, a more complete picture of the design can be achieved"

Second, i wanted to see, which transducer response is assumed, when only the 3 T/S parameters are used (Vas, Qts, Fs). That is, wich responce must have a loudspeaker, to produce identical results -wenn put in the vented enclosure- with the default sim, which in turn gives exactly the same results with winIsd- ?

The perfect response!!! flat from 5Hz upwards! But there is no such loudspeaker in the real world!

Conclusion: Boxes has to be tuned 10-20 Hz higher than proposed by winIsd (at least the pro 15'' i simed, there were more than 10) if an optimal anechoic response is to be achieved. The optimal tuning produces a much earlier (15-30Hz!!) - but nicht very steep - roloff.

This correlates very well with my subjective impresions with an Altec 604-8G and a Fostex FW405, which with the proposed enclosures had bad bass, adding only a kind of background low frequency noise. In the same rooms other loudspeakers performed normally. Of course the in-room response is a totally different issue, but it is difficult to add 5-10db all over from 20-100Hz .
These hearing impressions was the reason i got myself into the simulation trouble, since i'm a genuine newbie in loudspeaker questions.

Greetings
Konstantinos