Comparing drivers

[ziocalepino]

Hi everybody. It's a long time that I'm making myself this question. Reading Thiele-Small parametrs of two or more drivers, which of these values are more significative if I have to decide if two drivers are "equal"?

I think Vas and Qts are very important if one (as me) has to build cabinets.

In other words, if a cabinet is good for a driver, and I wnat to use it for another one, which TS-parameters is better to evaluate?

Many thanks for help

Ziocalepino

 

[Squeak]

I'd say Qts, Vas and Fs.
SPL/surface area is also important, but has much more fudge-factor.

 

[planet10]

Fe, Qt, Vas yes.

SPL/surface area is also important

Not really. SPL is important in determining how much power you need. Sd only in the sense that the driver needs to fit into the box, and the bezel size is a more direct measurement of that.

dave

 

[Squeak]

But it's important if you are doing any kind of line/horn as 90 percent of the projects in this board is. Putting a driver with a huge output in too small a box is not good either if you expect to use the driver to its fullest.

 

[chrisb]

But it's important if you are doing any kind of line/horn as 90 percent of the projects in this board is. Putting a driver with a huge output in too small a box is not good either if you expect to use the driver to its fullest.

Sure, but why not first let the OP set the guidelines? There's a long list of variables in the calculus of design of a loudspeaker system, and the preconceptions or even years of experience of anyone offering advice are of little service if they ignore the end user's application and acceptable compromises.

90%, really?

 

[Squeak]

Sure, but why not first let the OP set the guidelines? There's a long list of variables in the calculus of design of a loudspeaker system, and the preconceptions or even years of experience of anyone offering advice are of little service if they ignore the end user's application and acceptable compromises.

90%, really?

I don't know about 90 percent, but a lot more than half, right?
The OP didn't mention anything about specific guidelines, so it must be taken as query for general advice.

 

[chrisb]

I don't know about 90 percent, but a lot more than half, right?

I was curious myself, you're probably right about at least 50%

The OP didn't mention anything about specific guidelines, so it must be taken as query for general advice.

fair enough

 

[planet10]

But it's important if you are doing any kind of line/horn as 90 percent of the projects in this board is. Putting a driver with a huge output in too small a box is not good either if you expect to use the driver to its fullest.

No they are not. Many people think Sd is important in a TL -- it is immaterial. The fact that Martin uss it as an area unit further confuses the issue. What is important is the Vas.

Same with a horn.

dave

 

[Squeak]

Wow interesting. Any links to further reading?
Guess it would behoove me to put a question mark after everything I state as though it is fact. ;-)

 

[planet10]

Augspurger for TLs.

Any modeling software including MJK (you can insert anything (within reason) Sd, it is used as a unit of area)

dave

 

[AEIOU]

Hi everybody. It's a long time that I'm making myself this question. Reading Thiele-Small parametrs of two or more drivers, which of these values are more significative if I have to decide if two drivers are "equal"?

I think Vas and Qts are very important if one (as me) has to build cabinets.

In other words, if a cabinet is good for a driver, and I wnat to use it for another one, which TS-parameters is better to evaluate?

Many thanks for help

Ziocalepino

Be aware that manufacturers listed specifications and the actual production specifications may vary considerably. If you can, you should break in the driver(s) and then measure them yourself.

 

[planet10]

Do be wary that the data you measure may be quite different than factory and yet both may be valid, just measured at different points on the curve.

Also note that driver to driver differences can deviate from the mean by 20% or more.

dave

 

[chrisb]

Do be wary that the data you measure may be quite different than factory and yet both may be valid, just measured at different points on the curve.

Also note that driver to driver differences can deviate from the mean by 20% or more.

dave

based on recent measures of a randomly selected pair - in some cases the differences are enough that meticulous modeling would suggest boxes that aren't exactly the same for "the pair"

 

[planet10]

based on recent measures of a randomly selected pair - in some cases the differences are enough that meticulous modeling would suggest boxes that aren't exactly the same for "the pair"

Provided by a client. This one is dramatic but i've seen batches of drivers where the 2 that are furthest apart approach this -- one of the reasons i provide pair matching.

dave