The wishful/sinful problem with the Bozaks is like any older speaker where it arrived at the time of the transition between tubes and transistors. Tubes in this era did not have a lot of power and if you could afford a MC60 or MC75, you were one of the fortunate ones. But typically a serious listener of that time would have a whopping 2 Watts class A or maybe 5-18 Watts class B at his/her command. Needless to say, since amplifier power was at a premium, home speakers were VERY efficient and bass response was limited to about 50Hz. The HF limit was also a challenge since cone tweeters were all that was available.
Rudy Bozak took the design approach of high efficiency and high transient response and decided that low-mass pistons (i.e., all moving mass) was as low as feasible while still retaining stiffness and low cone undulations. His choice for tweeters was an aluminum 2.5" cone with a 3/4" voice coil and an aluminum dust cap. He called it the B-200 (X, Y, and YC models) and it evolved from fabric to metal cone materials and Alnico to Ceramic magnet materials. The theory was much like that of a full-range driver today where different parts of the piston mechanically decoupled themsemves from the other parts of the piston and provided extended frequency response. Here, the metal dust cap extended this driver's ability to produce frequencies above 6KHz to about 8.5KHz after which they pretty much fell flat on their faces. The tweeters are really the weak link in this old technology design although the range they cover they do so very well.
Bozak B4000s have an estimated 1-Watt sensitivity of about 99dB/W/m but the B302As I have with only one 12" woofer in its compliment this is more realistically 96dB/W/m. Dome tweeters are few and far between at anything close to this sensitivity and it is why I settled on the Audax TW025A28. It is rated at 96dB/W at 0.5 meters (duh!) so this tweeter is a bit down from the SPL of the rest of the system. Finding tweeters or super tweeters today with a matching high 1-Watt sensitivity is a real challenge until you embrace the realm of horns. Bullets horn drivers are abundant in car stereo systems but most sound lke fingernails on a chalk board. They scream and screetch but they go really loud with 1-Watt sensitivities as high as 110dB/W/m, rather outrageous when compared to the typical 86dB/W/m high-end speaker today.
JBL introduced a hi-fi home tweeter called the JBL 075 using a non-metallic diaphragm. It went through several iterations (075, 2402, 2404, and 2405) in an attempt to improve HF dropouts and directionality (beaming) and ended up in the Pro lilne under different model names including the highly coveted "baby cheeks" 2404H (105dB/W/m). Although this model uses an aluminum diaphragm, musicians and audiophiles alike still covet these units and as a result they are fetching easily over $200/pair for those still in good shape. Those with the original diaphragms and matched pairs are even more desired.
Not having access to such a fine driver, I settled on getting what I could. Querrying Parts Express about current models, I ordered a pair of ST200 and waited for their arrival. Hoping for this driver to replace the current tweeters and super tweeter, I disconnected them both and added an L-pad to the existing network. Once the levels were matched (practically nothing left to turn down this driver, almost all the way to minimum), you could easily hear the range at which the Bozak tweeters fell flat. However, the compromise was that the quality of the ST200 driver was disappointing and after two days worth of listening it went back to Parts Express.
BTW, there was talk about a modification to the bullet on this generation of drivers that improved HF response so I tried it out to see if it did. Below are the two RTA near-field measurements showing the difference between the unmodified and modified bullet.
The microphone was positioned 1/4" from the tip of the bullet and the volume level was exactly the same from the amp in both measurements. The second order Bessel crossover frequency for this super tweetter is 12KHz. Notice the improvements above 8KHz as a result of the modification. BTW, this mod is a broccoli rubber band placed at the throat of the bullet (pushed down evenly with a chopstick. However, look at the compromises in the midrange as a result of the modification (even though it is an off-axis measurement).
While still disappointed at the results, it helped me to understand the limitations of the existing drivers. Returning everything back to normal, I appreciated the Bozak "sound" compared to the cheap compression horn (as opposed to a well-designed horn). I am still searching for a high sensitivity planar tweeter to do the job for both the tweeter and the super tweeter but finding one that is high-end quality and has good dispersion characteristics is proving to be a challenge.
For now, I will settle for what I have and work on other planned crossover network changes (second order Bessel on both the LPF and HPF ends of the tweeter). I am hoping this will correct some of the interaction issues between the midrange driver and the tweeter near the crossover point and possibly smooth our the overall midrange response. I am encouraged by the progress I have made in balancing the sound from these speakers. Here is an RTA graph of where the system currently sits as measured from my reference spot.
· Althea: A Story of Love