TEST OF JBL 2445JPL Compression Drivers with horns

[♔Chello]

I have obviously been away from theoretical calculations of speakers for too long..?

Now I know why the JBLs sounded a bit aggressive on human voices:
The first setup was connected in parallel, which halved the resistance, which I assumed was 16 Ohm.
The second setup was clearly better, which was connected in series, which doubled the resistance, which based on the assumed impedance should have been 32 Ohm.

Along the way, I imagined that the crossover frequency had been somewhat higher than the theoretical one, and made a measurement:
This showed only 6.6 Ohm!

This means that the first setup gave a crossover at about 1400Hz, and the next at 700Hz.
When I calculated a new value for the capacitor based on the measured resistance and took into account that this was doubled as a result of the series connection, I ended up with a little more than doubling the value (40µF), and the crossover frequency was then about 300Hz at just over 13 Ohm load (series connection).

This sounds MUCH better - now all aggressive tendencies are gone. 

I also have some strange experiences that some of the sound sometimes comes from the side behind me, and that all recordings are now pleasant to listen to - even those that were bad before - all have a clear perspective, completely without negative stereo effects.

And although JBL recommends cut at a minimum of 500Hz with a 2nd order filter, it works great at 300Hz and a 1st order filter.


It is very strange that my measurement shows only 6.6 Ohm, while all found variants of the series are stated at 16 Ohm.

However, I am fairly certain that the actual impedance is somewhat lower than what I measured, because my ears perceive the crossover frequency as higher than 300; more around 500.

This is in any case far above the expected and hoped-for result, and I can finally put the finishing touches on the work by making the dipole membranes on the outside of the horns.