Circuit simulation using driver impedances.

[prash]

Hi all,

I have been trying to do some simulations for some xover circuits using SIMetrix (nice package). First of all, do I put the voice coil inductance and resistance of the driver in series to represent the driver in order to measure the dB (gain) across to get the driver graph giving a frequency sweep from 20Hz to 20kHz?

If this is correct, I have been trying to simulate the xover ckt of the Ariels with this package and I see a graph which starts rolling off at around 100-200Hz for the Vifa driver pair. Something is wrong here and I dont understand. Can someone please shed some light on this issue? ( I am basically plotting voltage versus freq)

The ckt is available on the Ariel website if anyone wants to try it out. http://www.aloha-audio.com/Arieltxt2.html#top

Please help.

thanks

[f4ier]

do I put the voice coil inductance and resistance of the driver in series to represent the driver

Yes, but speakers are more complicated than that. Here is a basic driver electrical model and here is a vented-box electro-mechano-acoustical model that you might find interesting.

Filter simulation is best done (and easier) with a set of SPL and impedance measurements of the drivers in their candidate enclosure. If your simulation package is flexible enough to accept measured data, then you're all set. Otherwise, you could use Speaker Workshop (which is free), SoundEasy, LspCAD, LEAP, CALSOD etc.

hope this helps

[prash]

Originally posted by f4ier


Yes, but speakers are more complicated than that. Here is a basic driver electrical model and here is a vented-box electro-mechano-acoustical model that you might find interesting.

Filter simulation is best done (and easier) with a set of SPL and impedance measurements of the drivers in their candidate enclosure. If your simulation package is flexible enough to accept measured data, then you're all set. Otherwise, you could use Speaker Workshop (which is free), SoundEasy, LspCAD, LEAP, CALSOD etc.

hope this helps

Thanks, that really helps. I will look at it in more detail. Does that mean that I can use the same model for all drivers except that he various values (mass of cone, compliance etc) are different? ie., should I plug in the values for the specific driver into the ckt.
That would also explain why I was getting the wrong plot.

[f4ier]

You're welcome, but I should note again that modeling filters for mid-high frequency drivers is best done with measured spl and impedance data. You could probably go as far as simulating the impedance to calculate the filtered transfer function, but you really must use measured spl data.

hope this helps and good luck