Parametric EQ Schem

Magnus Danielson magda at it.kht.se
Fri Mar 7 21:59:59 CET 1997


>>>>> "g" == gstopp  <gstopp at fibermux.com> writes:

 >> 
 >> Have someone noted that I kind of enjoy fiddling about with filters? :)
 >> 

 g> Oh yes we notice, happy to hear your stories! Certainly a worthy field of 
 g> study... many interesting places to go on the S-plane. Yet another example of 
 g> imaginary numbers giving real results.

As always I let the theory work for me :)

 g> Although it is interesting to speculate on the utilization of
 g> state-variable filters as parametric EQ elements, there comes a
 g> point where we must stop and ask ourselves exactly what kind of
 g> box we are trying to design. In other words, what are the user
 g> specifications? I just realized that I don't even have a
 g> paremetric EQ anywhere in the depths of my equipment pile, except
 g> maybe for an SPX-90 patch or a D-50 effect. All the EQ's on my
 g> Mackies are fixed.

Shame on you :)

Ok, I have had parametric filters in my rig for years, and I have diged
into a couple and have schematics of several more...

The minimum EQ is usually just the boost/cut one. This is cheap to
implement into multiple ones and therefore used in graphical
EQs. Really cheap ones uses only and LCR cursuits or later a few caps,
resistors and an op-amp in a Sallen & Key connection. Theses have
fixed frequency and fixed Q-value.
Now, for mixerboards you typhically will not be happy with fixed
frequencies, so the first second parameters you throw in is to sweep
the frequency. Typhically you let the frequency range be quite limited
but this has the benefit that the quality of the filter migth be
overall better (cheap optimation :). However, any good EQ should also
have a variable Q value as well... and if you are doing speaker system
EQing you *need* it (beleive me... I spent hours over EQs to fix a
system, but then I got it all within a +/- 0.5 db window :).

I would not call an EQ lacking one of these three controls an
parametric EQ. An EQ with only boost/cut and frequency control is an
sweepable EQ to me.

My parametric EQs is a pair of Trident Audio Developments Model CB9066
Parametric Equalisers. These are well built and uses transistors as
the active element (no op-amps!). These have three sections of full
EQs with +/- 16db boost/cut, bandwidth (in a way Q value), and
frequency sweep (low section 60-700 Hz, mid section 600-7k and high
section 3.5k-14k). There is also an lower shelf with 0-22db slope and
100-400 Hz frequency and an upper slope of 0-22db slope and 4k-15k
frequency. Each of the full EQ sections may be bypassed separately and
the full system may also be bypassed. In my opinion is the bypass
function really a valuble one sinse you can then make a correction,
swap it out and get it back in without loosing the precise any of the
values. Graphical EQs lacks the possibility of pinpointing the
problems, but usually is 3 parametric EQs not enougth...

BTW. These particular EQs used to belong to the maniac Michael
B. Tretow, the engineer and producer behind ABBA and the Caramba
record.

A good EQ should have sections that may overlap in frequency so that
one migth cover "problems" which happends to be close.

 g> What are the controls and capabilities of a standard parametric? Of a cheap one?
 g> Of a high-end one? Anybody got schematics for the Moog unit?

Kevin ougth to have schematics for that...

To add up to all this in it has lately become popular to add
compressor/expansion cursuits to each EQ step to be able to change the
dynamics of certain ranges... put let's take that as a separate post :)

Cheers,
Magnus




More information about the Synth-diy mailing list