Debugging a B3S Zaph clone

[gmarceau]

I'm trying to build an instance of Zaph's design with the B3S he described there:
http://www.zaphaudio.com/smalltest/compare.html

This is what they look like:


I would grateful if you could help me debug them.

I am getting a 12 decibel boost around 7000 hertz. According to Zaph's description, I should be getting a 4 decibel boost there as a result of the breakup mode of the aluminum cone, not 10. So I am wondering if I am seeing some other effect. My boost is also much wider than his.



I am measuring from an apartment room. It is a pretty bad environment to be measuring in. I am trying to compensate by taking a variety of near field measurements. In the graph below, the thin colored lines are the individual measurements. Yellow is at 10 cm, blue is at 20 cm, orange is at 40 cm. There are two blue lines and to orange lines because it took to measurements from slightly different positions. The thick white line is the average of all the measurements. The thick purple line is the response curve Zaph shows on his web site.

This is with the crossover Zaph proposes in his article. This is measuring with an unmodified Radio Shack dB meter. They are not stuffed, and not damped/

Do you have any idea what this could be? What would be the best way to try to fix it? Thanks.

[kristleifur]

Cool speakers. Very.