Spice speaker model help (Mal & Ral)

[Dag]

Hello!

I'm trying to model a speaker in a closed box using LTSpice and I'm almost there BUT I have a small error in the mechcanical<->Accoustic interface.

The "output trafo" (mech-accoustic transformer) is loaded by 2xMal and Ral (and the box if there is one) . But to get correct results I have to multiply Mal with 0.4 and not 2. With 0.4 I get the same results as Win ISD Pro Alpha and Basta!

I would be very, very happy is anyone can help me because I have not found any god accurate info on the webb!




** Trafo .subckt trafo 1 2 3 4
e1 1 2 3 4 {-Sd}
f1 3 4 e1 {Sd}
.param Sd=1
.ends Trafo

** Mal ("Two=2" => both the air on the inside and outside is included.)
.subckt Mal 1 2
g1 1 2 1 2 Laplace={1/Sd/p0/Cair/s*(1-1*(s*rs/cair)^2)/Two}
.param Two=2 cair=345 p0=1.180
.param Sd=1 rs=1
.ends Mal

** Ral
.subckt Ral 1 2
g1 1 2 1 2 Laplace={(2*pi*Cair/p0/s/s*(1-1*(s*rs/cair)^2))}
.param cair=345 p0=1.180
.param rs=1
.ends Ral

The box is modeled with a C={V/p0/Cair^2}
V=vol in m3

[MJK]

Is your Sd set to 1 as the script seems to indicate? I have not used PSpice in a few years so I may be totally off base. Typically when doing an acoustic to mechanical transformer the turns ratio is defined as 1:Sd^2. Is 0.4 equal to your Sd^2 value?

[Dag]

I'm using a Sd:1 trafo where Sd (Cone area) is a variable and not 1 (sorry for the confusion) It is an 8" speaker with Sd=235cm2 (23.5*10^-3 m2).

[Svante]

Hmm, I am not that good at reading Spice scripts.

What are the exact expressions you use for Mal and Ral? I suppose they are there in the code, but ...

It seems as if your model is a pulsating sphere, it is easy to get constants wrong there when converting Sd of the baffled piston to an equivalent radius to the sphere.

You might also consider usin the parallel model for the radiation impedance; if you do, the components (for the pulsating sphere) become an inductor in parallel with a resistor. Ordinary components, no fishy frequency dependencies.