an almost "perfect" box

[steffda]

Hallo togehter,
i just simulated a 2-way-box and with a very simple crossover i got a really good result.
That made me sceptical and i looked at different sites for crossover designs. What i found was always more complicated and in my simulation i got worse results than my one.

Here is the Box:

I just see, that quc (my simulator) cutted the paramter d=11 and QL=7 for vented box.

...and here the models i used:

What i'm doing wrong?

 

[diypass]

hi, what software is this ?

 

[chris661]

The simulation all looks fine, so long as you've got some drivers that are perfectly flat. As soon as you try to work with real drivers that have their own inherent rolloffs etc, it becomes a lot more difficult.

Chris

 

[steffda]

hi, what software is this ?

I'm using qucs. Therefore i created the driver/box models as library components.

The simulation all looks fine, so long as you've got some drivers that are perfectly flat. As soon as you try to work with real drivers that have their own inherent rolloffs etc, it becomes a lot more difficult.

I know, but other simulation software also has only the Thiele Small parameters. Thats why i'm wondering about the (a lot) different results.

 

[diypass]

I'm using qucs. Therefore i created the driver/box models as library components.

thank you it seems a very instresting software

 

[steffda]

Here is the laudspeaker library for qucs.
De-zip it and copy it to .qucs/user_lib
View attachment Loudspeaker.lib.zip

 

[diypass]

where you downloaded them ?

 

[steffda]

where you downloaded them ?

The Loudspeaker.lib i did by myself, the electrical models for speaker an boxes i found in internet.

 

[diypass]

oh,good!, should be very usefull also other enclosure loading and "special" loading
however, very usefull software

 

[steffda]

Thanks!

 

[cowanaudio]

800Hz X/O?

 

[steffda]

800Hz X/O?

Yes, why not?

Just to explain the "workflow":
I'm starting with 1kHz and then, depending on the result of simulation, i change the parameters as long until i'm enjoing the result.
It's a lot of experimenting, but not much science behind.

 

[cowanaudio]

That poor little SB21RDC sure won't enjoy the result. You'd be well served to aim for an 1800-2KHz X/O.

 

[steffda]

That poor little SB21RDC sure won't enjoy the result.

Why?

You'd be well served to aim for an 1800-2KHz X/O.

Ok, i will try that.

 

[cowanaudio]

Why?

You can't drive a tweeter that low without severe distortion or failure. Two times Fs is about as low as you can go.

Ok, i will try that.

Also be aware that your models are over simplified and will not reflect the final result.

 

[steffda]

Ok, where can i find bether models?

 

[cowanaudio]

I use CALSOD to simulate passive crossovers using real measurement files. It is really old DOS based software that has for the most part been superseded by newer programs. Have a look around here for some current suggestions.

 

[Pano]

XSim is a very good crossover simulator what will allow you to use real measurement files. You could also build and use those models you have in your schematic, but real measurements are better.
I does not have a box simulator, tho.