Thursday, December 20, 2012

Bozak Rebuild Project - Part 2

Bozak Rebuild Project - Part 2
In Part 1, We saw the starting point from which I planned to improve the basic sound of a vintage pair of speakers. Many folks are familiar with the Bozak name but unfamiliar with their sound so hopefully in this series we can all climb up the better-informed ladder and upgrade a good beginning into a truly contemporary pair of high-end speakers.
I started with quite a mess and a plan of attack to try to fix the sound of these old speakers hoping that from the other side of the grave Rudy Bozak himself would be proud.  Rudy did not like the sound of second-order crossover networks so he chose simple first-order designs.  While there is nothing wrong with this (and it keeps costs down), my ears are more sensitive to the overlap region on either side of the crossover point.  When drivers are not properly limited, the resulting coloration introduced by drivers asked to reproduce parts of the audio spectrum to which they are not intended to do is more objectionable to me than the effects of higher-order crossover networks.  So my first thought was to roll off the midrange more steeply with a second order network.
To confirm that the midrange was indeed the culprit and not the tweeter, I disconnected the woofer and tweeter, and ran another RTA measurement.  The results are shown below.
 Pink Noise with Tweeter Array Disconnected
This graph clearly shows that the midrange driver is still contributing acoustic energy up to 7KHz despite the fact that the crossover network is supposed to roll off these effects at 2.5KHz.  I had found the source of the problem and this was half of its solution. 
Checking the math and the value of the midrange inductor (1.12mH), I was satisfied that the 2.5KHz crossover point was indeed correct and suspected that the sonic contribution above the crossover frequency was from the rising impedance of the driver.  I tried many Zobel networks across the midrange terminals all with little success and then decided to move to a second-order network.  The results of this change are shown in the next RTA image.
 Second-order Midrange Network
I then added a Zobel network to the woofer and things really started getting to sound real.  The timbre of the drums came to life where you could hear the skins stretching as the sticks struck.  The mellowness of each drum became clear and distinct making one wonder exactly what kind of wood these drums were made of.  The over-exaggerated upper midrange was tamed and an edginess that was once predominant was now under control.  The depth of the sound stage improved and I knew I was on the right track.
Things were better but still not great.  The new network improved clarity in the lower midrange but introduced more sibilance, an issue to which my ears are very sensitive.  Life with a smoother midrange was more tolerable than the sibilance so we moved on to the missing top octave.  We’ll see what was involved with that next time in Part 3.

See also Part 1
Yours for higher fidelity,
Philip Rastocny

I do not use ads in this blog to help support my efforts. If you like what you are reading, please tell your friends and if you are so inclined, buy some of my work at  My titles include:
Copyright © 2015 by Philip Rastocny. All rights reserved.

No comments:

Post a Comment

To comment on this blog, you must first be a member. All comments are moderated.

Note: Only a member of this blog may post a comment.