About the Alpha 15A frequency response and passive filtering.
I've recently done some justMLS measurements of impedance and response with this driver on a small baffle 60x60 cm. The response is similar to the latest measurement by Scorpion earlier in this thread, but the 2 kHz peak is about +10dB.
The surprise, however, was how the driver responded to passive crossovers. I ran simulations in lspCAD with different filter topologies and component values. It seems like the drivers don't respond well to large series inductors. A larger inductor does very little to reduce high frequency response, and Fc seems to be somewhat fixed around 100-200 Hz. The inductor mostly just reduces overall output and flattens out the 2kHz peak. Some larger values even introduced new peaks in the response. So I landed on 1,2 mH.
When adding a parallel capacitor there is a sharp rolloff at around 500Hz. Then a flat response up to 2 kHz. Increasing the size of the cap does nothing to the cutoff frequency, but only reduces the level above 500Hz relative to the level below. The optimal value here was found to be 220 uF (!).
More complex filters only introduced large peaks in the response.
So crossing these drivers passively may not be as straightforward as one would wish. Using simple crossover theory may produce different results than expected.
Optimal crossover was found to be 1,2 mH in series, 220uF in parallel, Fc=500Hz. Try it out!
