Question on enclosure design programs

[Low Rider]

Hello everyone,

Here's a kind of a noob question about those programs like winisd and alike.

The frequency response they'll show for a particular driver/box - does it mean absolute difference from driver's SPL (so when it shows -3dB at 40 Hz and drivers mean SPL is 92 dB the box will have 89 dB@40Hz),
or is it just offset that one would apply to the published freq. response of the driver (so if the driver is allready at 90.5 dB@40 Hz, in the box it would be at 90.5-3=87.5 dB)?

Also the published freq. responses, do they usually mean free-air operation of the driver, or infinite baffle or what?

[Svante]

Originally posted by Low Rider

The frequency response they'll show for a particular driver/box - does it mean absolute difference from driver's SPL (so when it shows -3dB at 40 Hz and drivers mean SPL is 92 dB the box will have 89 dB@40Hz),
or is it just offset that one would apply to the published freq. response of the driver (so if the driver is allready at 90.5 dB@40 Hz, in the box it would be at 90.5-3=87.5 dB)?

The response curve that most softwares show, is a prediction on the response of the final system. However, the models used in simulation softwares do typically not include every little detail of the drivers. This means that eg the whiggles in the response caused by the cone breaking up at higher frequencies are not predicted by the softwares. So in that sense the curves will have to be merged.

Originally posted by Low Rider

Also the published freq. responses, do they usually mean free-air operation of the driver, or infinite baffle or what?

That should be stated somewhere in the data sheet, but not all manufacturers do that, at least not on every sheet.

[Low Rider]

Originally posted by Svante


The response curve that most softwares show, is a prediction on the response of the final system. However, the models used in simulation softwares do typically not include every little detail of the drivers. This means that eg the whiggles in the response caused by the cone breaking up at higher frequencies are not predicted by the softwares. So in that sense the curves will have to be merged.



That should be stated somewhere in the data sheet, but not all manufacturers do that, at least not on every sheet.

So, to recap -

To acurately predict how my box will behave (in ideal conditions) i take the response curve from the software for ~20-200 Hz and the rest of the curve from manufacturers frequency response graph since the box seldom influences response above that. Am I on track here?

[Svante]

Originally posted by Low Rider

So, to recap -

To acurately predict how my box will behave (in ideal conditions) i take the response curve from the software for ~20-200 Hz and the rest of the curve from manufacturers frequency response graph since the box seldom influences response above that. Am I on track here?

Sort of. There will be effects of the box in the ~200-4k region too, due to baffle diffraction effects. Some simulators can include diffraction effects, others don't. There is also a difference in how advanced/accurate the diffraction models are.

In my experience, good drivers can have a very flat response, and in that case, almost all little jiggles in the frequency response above ~200 Hz can be explained by diffraction effects. Less good drivers often show peaks and dips in the response due to cone break-up.

It is the task of the designer to get a grasp on these effects for the driver in question and design the box and filter accordingly. The way to do this is to do measurements of one's own.

Since response curves is nowadays part of the marketing, one should always be a bit sceptic to the curves. I have seen cases when the manufacturer has changed the cone material to something cheaper and more resonant, but left the curves of the old drivers on the web.

[Pit Hinder]

Svante,
to quote Sir Winston Churchhill: "Never trust statisistcs you haven't cooked yourself"

[MarkMcK]

Computerized design tools are wonderful aids to the design process. They are not, however, the entire design process.

Their output is a prediction or estimate of what should be. This estimate could also be called an informed guess.

I am not certain why you want to construct this frequency response graph from a low frequency performance guess and a seller produced frequency response. If you do construct your graph and you publish your graph construction, please note that it is a construction. Please do not claim that it shows actual performance.

Mark

[Low Rider]

The speaker in question here is Visaton's BG20 - I trust Visaton not to be dishonest about the freq. response of their speakers.
They even publish the method and test enclosure that the speaker was measurred in, along with phase-frequency data, impededance-phase-frequency...everything you could wish for.

I have just found out about the Visaton's proprietary software called BoxSim, and I must say it's a mile away from anything I've seen so far (in both the detail and acurracy comp. to real measuremets).

However the extreme detail of the simulation requires just as much quantity of driver's parameters it' s making it unsuitable to simulate anything other than visaton's own drivers for which they have database files to go along with the software.

Originally posted by MarkMcK

I am not certain why you want to construct this frequency response graph from a low frequency performance guess and a seller produced frequency response. If you do construct your graph and you publish your graph construction, please note that it is a construction. Please do not claim that it shows actual performance.

I am by no means interested in going commercial with this project or whatever did you mean by 'publishing the constructed graph' and claiming it shows actual performace.

[Pit Hinder]

Low Rider,

Visaton - yes. I wave their flag and bang their drum, anytime. Not only are they honest about their drivers' specs, they spend a lot of money on research too, and share the results with the DIY crowd.

Greetings to Croatia,
Pit

[Low Rider]

Gotta love the Visaton

Not that I wouldn't like to see some more PA or instrumental speakers from them

[Pit Hinder]

OK, allow me to wave the German flag again. Look at Monacor. Thorsten says they don't have dealers all over the world, but googling in your neck of the woods might bring a happy surprise.

Pit