It’s almost a year since I posted the magnet system and the simulation results of the EPOS woofer. So what happened within the year…well, a lot? First of all, we made a tooling for the cone and I did not like the sound of it. It measured as expected, but had a somewhat quacky character I did not like. So back to the drawing board and the result was a new cone with variable thickness and a different shape.
Also, the original magnet did not survive and we moved back to Ferrite material plus a compensation magnet on the back and a small Neodymium magnet on the pole. The pole piece got a special shape and we also added Aluminum rings inside the magnet and on top. The top of the pole is shaped as well (the so-called Marvin pole, originally designed by my friend Joachim Hammers).
So why did we make the magnet that shape? One reason of course was to make the motor as linear as possible, following all the rules we learned with Klippel measurements.
The second reason was to make the impedance that the crossover “sees” as linear as possible. Not many people know that the impedance of a loudspeaker is different depending on the position of the voice coil. So the poor crossover never sees the same load and the response of such a speaker gets modulated. My mate Lars Goller did a lot of work on this during his days at VIFA and Scan-Speak which ended up in the SD-2 magnet as far as I remember. Below is the measurement of a 4-layer voice coil at three different positions, showing the change in impedance. Four-layer coils are often used to roll off the response curve of a driver without a crossover.
Below is the result of the EPOS woofer with the same 3 positions of the voice coil.
Now the impedance curve is stable over excursion and that makes it possible to add a crossover without negative effects. By the way, the modulation is not only a theoretical idea, it can be simulated.
Below you see the difference in the response curve on three different positions. The filter used was a standard 1.2mH with a 5.6uF plus 1 Ohm to ground. Scary…….
When you do the same to the EPOS woofer, it looks like the measurement below. Please note this is a 2-layer coil, so the roll-off is different.
Essen, 12. April 2022 KHF