First Attempt



This page outlines some of the things I learned when doing my first crossover. It used the Vifa 5" full range driver and the Radio Shack (don't laugh) "Rocket" tweeter.  The tweeter has two pieces of Mylar film that launch the sound. 
This first graph is a plot from the Vifa 5" full range driver purchased from Parts Express.  I don't recommend replication of this project because half of the 8 drivers (i.e. 4) I received from P.E. had scratchy voice coils.  The bottom blue line is the 3rd order distortion.  It sounded terrible and the distortion plot indicates thus..  I believe the nasty nasally sound is related to uncontrolled upper frequency resonance, but I am not 100% sure.  I heard this nasally sound with the SS8545 using a 2nd order crossover at 2.2khz a few years ago.  The nasally sound is quite obvious to the ear after exposure.  It has about the same pungency as multi-vis gear oil.

This is a plot of the SPCA.  Notice the lack of 3rd order distortion.

This is what I did to fix the problem with the Vifa 5" full range driver.  This was my very first real crossover attempt.  I spent about 6 hours with it.  It measures pretty flat, but the null does not perfectly mirror the acoustic crossover point.  This is the reason for the slightly imperfect null and the slightly slanted phase line.

 learned the following bits of information when designing this crossover.  

1.  First order crossovers are fine if the tweeter will handle the low end juice and the woofer doesn't resonate.

2.  When using MLS and a gated time window the tweeter will make a "splat" sound if it cannot handle the juice.  The graph might show distortion too, but this wasn't always consistent.

3.  When using MLS and a gated time window the woofer will sound nasally if it is resonating.  It will also sound this way when used with music.  The graph will also display the distortion.

4.  Getting the phase perfectly correct is quite tricky.  I got it pretty darn close, but not quite perfect.  The null would probably be perfect at about 5khz.  I gave up because the guy wanting this project isn't picky about the sound and he has waited a looooong time for it.  Also, I don't think the minor phase improvement would make much difference.  I tried some different sized inductors and capacitors in an attempt to shift the phase and found that I could shift the phase desirably, but was killing my nice flat curve when accomplishing this.  I summarily decided that the above result was the best compromise.

5.  I chose the high crossover because I didn't want the tweeter blown when abused and also because the when the "right" size inductor was placed in  series with the woofer it looked like 5khz was about the right place to cross over.  I determined the "right" size inductor because it controlled what little baffle step existed.

It sounds pretty decent, and has a somewhat dark sound that is likely due to a dip in the off axis response between 3 and 5k.  This is fine, and sounds good with John Lee Hooker, "Boom, Boom, Boom".  The speaker sounds better than my old Polk Audio 10's, but the midrange certainly lacks when compared to the 1801's.  It is also slightly inferior to the SPCA's, but does a few things better.  There is some low level detail on my Bocelli CD that the little Vifa/Rocket combo revealed that I haven't heard before??

Overall, I am pretty pleased with this first effort. 

Woofer: 5mh, 7uf

Tweeter: 3uf, 7mh, 9ohm (parallel)

The woofer nominal impedance is 4ohms, and the tweeter is 6 ohms.