Original Large Advent - Reverse Engineering - Some Measurements

[Pete B]

I've been working on a pair of original Large Advents from November 1972. These are Rev1, 16uF/16uF crossover type.

[Pete B]

Measured the woofer crossover transfer function which at a glance is first order being simply an inductor (1.5 mH .28 ohm).

However, since the speaker load is highly inductive it is actually more gradual than first order from an electrical standpoint. Another thing to note is that it actually combines to increase the voice coil inductance which forms roughly a second order low pass with the cone mass. It should be noted that the cone mass can be frequency dependent due to cone break up and decoupling, so it is rather complex.

Woofer crossover electrical transfer function in system:

-.18 dB -10.4 deg at 100 Hz

-3.3 dB -19.9 deg at 500 Hz

-4.7 dB -21.3 deg at 1 kHz

-8.3 dB -22.4 deg at 5 kHz

-9.9 dB -23.1 deg at 10 kHz

The amplitude slope is nearly a straight line from 100 Hz to 10 kHz on a log scale, and is clearly not first order as would be expected. The phase also does not reach 45 deg which is expected with true first order networks.

Looking at 1 kHz relative to 100 Hz it seems that there is about 4.5 dB of baffle step. There are many possible reasons why more is required, the woofer could be acoustically peaking in this range, and/or the tweeter level is clearly on the high side.

I would prefer a second order network here, but this was a good economical solution.

[Pete B]

The Large Advent Rev1 crossover schematic is provided in the link below. It does not render well in IE, but if you save it and open it in a viewer it should look fine. Rev1, as I call it, is the 16uf/16uf capacitor version, with the "button" style white and red inductors:

http://baselaudiolabs.googlepages.com/LrgAdventR1.GIF

[dynaco_dan]

Hi Pete;

Another excellent job with your drawing, thank you.

The crossover certainly is a complex one, a lot more parts than I would have thought.

[Pete B]

Thanks Vern!