NetNews Usenet Archive 1992 #20/NN_1992_20.iso/spool/rec/audio/12148

Internet Message Format | 1992-09-07 | 2.8 KB

Path: sparky!uunet!zephyr.ens.tek.com!gvgpsa!gold.gvg.tek.com!chrisc
From: [e-mail] (Chris Christensen)
Newsgroups: rec.audio
Subject: Re: Bi-Amp vs. Bridged Bi-Wire
Message-ID: <[e-mail]>
Date: 5 Sep 92 00:26:31 GMT
References: <[e-mail]> <[e-mail]> <[e-mail]>
Organization: Grass Valley Group, Grass Valley, CA
Lines: 48

In article <[e-mail]> [e-mail]
(Alex Mitaru) writes:
>In article <[e-mail]> [e-mail] (Chris Christensen) writes:
>>>What is a "Zobel" or response compensation network?  I'd guess its
>>>an E.Q. type device to compensate for speaker non-linearities??

>>A Zobel is a simple network that helps to compensate for the imprdance
>>bump of a driver at resonance.  It's selection is somewhat iterative
>>based roughly on the frequency of the bump and it's Q.

>  I'm sorry Chris, but that is wrong. A Zobel is a RC network to compensate
> for the increase in impedance at higher frequencies. A uncompensated woofer
> can present an impedance of 20 to 40 ohms to the low pass filter at 2 - 3 KHz;
> hardly what you have computed your filter. Using a RC network in paralel with
> the woofer you can maintain a stable impedance and therefore your low pass
> will behave as expected.

I have seen L's used in conjunction with an RC to control the Zobels
characteristics.  This is a nit to pic but Zobels are more than a
simple RC,  that's why I refer to them as networks.  It also serves to
make me sound smarter than I am  :-)

As I have viewed speaker system design I have seen  designers use a
Zobel for LF impedance compensation and cross over tha driver before
the rising response becomes an important term in the equation.

I am looking at a swept impedance chart of an 18" LF driver that I am using.
The impedance peak at resonance (50 Hz) is at about 12 ohms. The chart runs
out to 20 kHz.  The impedance is flat from 100 Hz to just above 1100Hz.
It then rises to about 15 ohms at 20 kHz.

I will put in a Zobel to flatten out the bump at resonance and cross
it over about 200 Hz and I think it will be just fine thank you.  What
do you think?

I still contend that the Zobel is used, most of the time, to
compensate for the higher Q bump at resonance and not the rising
impedance that occurs above resonance.

I may be wrong here and I will be the first one to admit it after I
have checked my reference books.

Oh and I also believe that whatever works for you and sounds good so
rules and ideas are mmeant to be broken and changed "-)
--
Chris Christensen                  The opinions I express are my own,
[e-mail]            and sometimes they are wrong!
916-478-3419 FAX [phone]      After all, I *AM* only human.