Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Posting-Version: version B 2.10 5/3/83 based; site hound.UUCP
Path: utzoo!watmath!clyde!cbosgd!ihnp4!houxm!hound!rfg
From: rfg@hound.UUCP (R.GRANTGES)
Newsgroups: net.audio
Subject: Re: crossover design: for $0.02 plain
Message-ID: <1415@hound.UUCP>
Date: Wed, 16-Oct-85 17:35:33 EDT
Article-I.D.: hound.1415
Posted: Wed Oct 16 17:35:33 1985
Date-Received: Fri, 18-Oct-85 00:23:02 EDT
Organization: AT&T Bell Labs, Holmdel NJ
Lines: 49

OK, already.  I think there is a very simple solution to this problem.
One I used many years ago with great success. It simply amounts to using
a resistive mixing network across the existing stereo amp outputs, n-
stages of RC low-pass filter where n depends on how fast you want the
low pass filter to drop off (12 db/ octave is what I used), practically
any old (mono) power amplifier, and the subwoofer.  I used an old
"Williamson" tube amp to drive the woofer part of a Klipschorn.

The Khorn was actually located in the dining room off the living room where
the rest of the system was, but it all sounded good. The design included a pot
so the level of the subwoofer could be set as desired. Sort of a super
bass tone control (boost only).

I will rty to sketch the circuit:



   o--------/\/\/--o--/\/\/--o--/\/\/\/\--o----o
   Ain        R1   |   R3    |     R4     |    |
                   |         |            |    \         --------           /
   o--------/\/\/--o       -----        -----  /<--------|      |-------- /
   Bin       R1    |    C2 -----     C3 -----  \ R5      |      |         \
                   \         |            |    \         | power|--------- \
               R2  /         |            |    /   ------| amp  |        sub
                   \         |            |    |   |      ------        woofer
   o---------------o---------o------------o----o----          
   Common  
The idea is for R1 to be large enuf to not load the speakers. 100 ohms should
be fine. R1,R1 and R2 form a mixer. R2 should be small, say 10 ohms. The RC
section, R3C2 should have a high impedance level compared to R2. Again the 
usual rule of thumb is a factor of ten is sufficient, so R3= (say) 100 ohms.
The impedance of C2 should be 100 ohms at the cutoff (turnover,3 db) frequency
let's say around a 100 hz cutoff. My trusty (bottom drawer file) reactance
frequency chart says C2 should be about 15 microfarads. R4 should be ten times
R3 or about 1000 ohms. C3 can then be one tenth of C2 for the same turnover
frequency, so C3 = 1.5 microfarads. The pot, R5 should be again 10 times R4
approx. so R5 = 10K ohms. 
Thus you end up with a summed center channel with a 12 db per octave Low
Pass Filter turning over at 100 hz. For lower or higher turnovers, you
can vary the values of C2 and C3. This mixing filter has a voltage loss'of about a factor of 0.1. Most power amps will be able to make that up easily. If not,
you might have to resort to an amp stage after the pot using an ic.
R2 could be increased a bit without much harm. It could be 20 ohms for a loss
of 0.2x (one fifth).
Depending on your power amp, you can reverse the connections to the woofer to
get the woofer in pkase with the main speakers.

-- 

"It's the thought, if any, that counts!"  Dick Grantges  hound!rfg