Drivers selection criteria.

[beppe61]

Dear Sirs,

first of all I declare my almost complete ignorance on speakers design.
Then I would like to ask you if when selecting a driver for a given purpose you base your choice only on Thiele-Small parameters analysis or also by judging them with real listening tests.
I ask this beacuse I remember a very interesting series of articles written by Lynn Olson who states that a real listening test of a driver mounted on a IEC standard open baffle drives his choice.

Opinions will be very welcome and appreciated.
Kind regards,

beppe

[beppe61]

Dear Sirs,


On the topic of " STRAIGHT DROP IN REPLACEMENT " I have a question.
In order to establish if two different model/brand drivers can be replaced without any required mods to the crossover, which electrical parameters should be the same?

Is it sufficient that just the following parameters :
1) impedance
2) sensitivity (I mean same dB/1W)
are the same ?

Physical dimensions constraints apart of course.
And of course I am talking of woofer with woofer, tweeter with tweeter, etc.
This technique seems to me extremely interesting indeed.

Thank you very much indeed.
Kind regards,

beppe

[pinkmouse]

Hi Beppe

As your two threads were asking more or less the same thing I've merged them.

As for your questions, in general, swapping drivers based on T/S parameters will get you a basic match, but does not account for any differences in frequency response, which your existing crossover may or may not be designed to compensate for, so to do it properly you will need to to compare these curves as well, and probably redo your crossover for best results.

As for listening, yes, it's good if you have the budget to buy samples of all the prospective drivers to compare, otherwise you have to rely on manufacturers published data and the comments of others that you trust, or well regarded complete projects.

[sreten]

Hi,

I'll answer the question here for a 3-way speaker bass unit :

For a typical 3 way speaker is reasonable to assume the bass
unit is operating over its pistonic region and therefore the
response is according to its parameters.

Even if it it is being used above is pistonic region if you have
reliable frequency response plots they can be compared.
If they do not differ appreciably below say 1kHz then all is well.
(edit : and there are no huge differences above 1Khz - see below)

If a driver has nearly the same Re, ~ the same Le and
the same sensitivity, then it could be used as a substitute.

If the crossover has a zobel for the bass unit, then the Le
does not need to be ~ the same, if it is not ~ the same
then the zobel capacitor value would need to be adjusted.

The Fs, Qts and Vas of the driver will tell you how it should
perform in the original box, but note that parameters don't
tell you about things like distortion, so the reputation of
the driver manafacturer should be considered.

Finally for a 3 way bass driver if it is a metal cone its possible a
"trap" circuit is built into the crossover for its major break mode.
This could safely be removed for a less stiff cone substitute.
Conversely using as metal cone bass driver as a substitute
for a less stiff cone is not a good idea if you do not also add
a "trap" circuit for its major break mode.

/sreten.

[beppe61]

Originally posted by pinkmouse
Hi Beppe
As your two threads were asking more or less the same thing I've merged them.
As for your questions, in general, swapping drivers based on T/S parameters will get you a basic match, but does not account for any differences in frequency response, which your existing crossover may or may not be designed to compensate for, so to do it properly you will need to to compare these curves as well, and probably redo your crossover for best results.
As for listening, yes, it's good if you have the budget to buy samples of all the prospective drivers to compare, otherwise you have to rely on manufacturers published data and the comments of others that you trust, or well regarded complete projects.

Dear Sir,

thank you very much indeed for your kind and very useful reply.
I asked about the selection of drivers (i.e. cone drivers) by hear after reading an article about the differences in tone of different materials (bextrene, kevlar, paper, plastic, metal, etc.).
On the other hand it seems to me that a lot of speakes designers relies, above all, on softwares and T-S parameters to design their products, maybe leaving the listening for the final tuning (is it called "voicing"?).

Thank you veyr much again.
Kind regards,

beppe

[beppe61]

Originally posted by sreten
Hi,
I'll answer the question here for a 3-way speaker bass unit :
...
/sreten.

Dear Mr. Sreten,

thank you sincerely for your kind and extremely valuable "accelerated" lessons on speakers design !
I am collecting your directions as reference.

Thank you very much again.
My best wishes,

beppe

[Curmudgeon]

We bought samples of everything that looked interesting to us, and then listened. For the woofers (and to a lesser extent the midrange) we ran simulations on LEAP. I had already had very good experiences with SS woofers, and the first unit we evaluated wound up in the final design.

For midranges, we tried units that had good reviews in commercial designs, or that interested us because of cone material, published curves, etc. LEAP was used to develop test cabinets to ensure bass response down to the woofer, but listening tests failed all of them. Partly our fault, as we did not then know how important associated system units, cables, and crossover components were. Several upgrades in system and associated components were required, and we finally settled on the Audio Technology 4" (somehow their inches are bigger than everyone else's ) which, as we continue to improve the associated components, just keeps getting better. Marvelous driver.

For tweeter, at that time the SS Revelator was used in almost all flagship designs, and after listening tests we never looked for another.

That's the long answer. The short answer is that we do both, at every stage of the design process.

[beppe61]

Originally posted by Curmudgeon

We bought samples of everything that looked interesting to us, and then listened.
For the woofers (and to a lesser extent the midrange) we ran simulations on LEAP.
I had already had very good experiences with SS woofers, and the first unit we evaluated wound up in the final design.

For midranges, we tried units that had good reviews in commercial designs, or that interested us because of cone material, published curves, etc.
LEAP was used to develop test cabinets to ensure bass response down to the woofer, but listening tests failed all of them. Partly our fault, as we did not then know how important associated system units, cables, and crossover components were.
Several upgrades in system and associated components were required, and we finally settled on the Audio Technology 4" (somehow their inches are bigger than everyone else's ) which, as we continue to improve the associated components, just keeps getting better.
Marvelous driver.

For tweeter, at that time the SS Revelator was used in almost all flagship designs, and after listening tests we never looked for another.

That's the long answer. The short answer is that we do both, at every stage of the design process.

Dear Mr. Curmudgeon,

thank you very much for the kind and valuable reply.
Very interesting indeed.
You have satisfied all my doubts and questions.
Kind regards,

beppe

[sreten]

Hi,

Details of the crossover components and circuit for the 3 way would help.

Note that only the components for the bass unit are relevant.

/sreten.