Updated : nov. 27, 2007 Steiner VCF
En français 
back to summary


Description

The Steiner VCF, also referred to as the Synthacon VCF, has recently become well-famed thanks to Ken Stone. In 1979, I came across the original article by Nyle Steiner (Electronic Design, 1974). I was seduced by the simplicity and low cost for such a versatile filter. I gave it a try and eventually built three of them. One was integrated in my very first modular synth (built in 1979-1980), and the two others were never incorporated in a synth...

I started a new modular project, and it appeared to me that I needed to include a few Steiner VCFs in it. But instead of using the original circuit I thought it would be interesting to design some variants of this fascinating filter. Eventually, I chose to build one around a CA3096 transistor array and a second one around a 2SC1583 dual transistor and a regular OPA.
Below, is a picture of one out of the three Steiner VCF I built in 1979. The PCB was hand drawn with an etch resistant pen. Those three guys are still alive and kicking ! By the way, the transistors are BC547 (NPNs) and BC557 (PNP).
old

How does this simple basic filter sound ? Well, great I'd say ! Following are two sound clips (mp3) excerpted from improvisations I played back in 1980 with my US1 modular.


Clip 1 : "Sunset sentier" , in this short excerpt the VCF is fed with two VCOs, the VCF frequency is modulated with a short sequence.

Clip 2 : "A la Tangerine" ,  in this excerpt you can hear the bass sequence with the cutoff being modulated, the strings are played on the Crumar Multiman S.

Next are audio examples for the prototype based on a 2SC1583 and an OPA. The CA3096 version sounds exactly the same.

Clip 3 : Square signal with PWM, low Q

Clip 4 : Square signal with PWM, medium Q

Clip 5 : Square signal with PWM, high Q

Clip 6 : Sawtooth signal, tweaking with the frequency knob


Schematics
CA3096 based
This version works nicely and sounds exactly the same as the traditional circuit based on discrete components. For this version, I use a dual OPA (TL072) one as an input buffer, the second to level up the output signal. By the way, both are inverting amplifier, thus the output signal is not inverted with respect to the input signal.
The core of the filter is based on a low cost transistor array, the CA3096. Why did I choose such an array ? Well, simply because I have something like twenty of these sleeping in a drawer of mine. If you cannot find it (I think it's obsolete now) you can replace it by an Intersil HFA3096 that is pin to pin compatible and costs around US$2.20.
When T1 is properly adjusted, the filter tracks in V/octave over a four octave range, above and below this range the tracking is no longer linear.
2SC1583 + TL071 based
 
This version works and sounds also very nice. In my first attempt, it did not work properly due to an error of mine : I used a 22k for R12. As a matter of fact, 22k was too small and lowered to much the input impedance of the OPA. Osamu Hoshuyama found the bug and suggested that the value of R12 must be significantly increased (e.g. 2M2). I did it and it fixed everything (a big thank to Osamu !). 
The 2SC1583 dual transitor can be purchased for  a reasonable price at  AMS ( http://store.americanmicrosemiconductor.com ). or in Europe at Reichelt ( http://www.reichelt.de ) or at BMM ( http://www.bmm-electronics.com ).
By the way, if you cannot find 2SC1583 you can use either matched BC547 or matched 2N3904 trannies. If you use the PCB design shown below, you must drill a second hole for the emitter of the second transistor (the emitters in the 2SC1583 use a common pin [3]). I tried with BC547s and it works perfectly.

When T1 is properly adjusted, the filter tracks in V/octave over a four octave range, above and below this range the tracking is no longer linear.
Besides, I have included a diode limiter (D9/D10) in the feedback loop of the TL071. This makes it possible to limit the amplitude of oscillation of the filter at maximum resonance. This improvement was introduced by Yasunori HIRAKAWA .
 
Well, I guess it's always like that, you think you made an original design and then you discover one year later that somebody else already made it several years before ! While surfing the internet I came accross a site (sorry I didn't bookmark it, I know that's dumb) where I found this schematic of a cheap guitar synthesizer by Terry Tsutsumi which uses the very same filter design : a 2SC1583 dual NPN and a generic OPA ! Since it was designed to operate with a single 9V battery and that some people asked me which modifications had to be applied on my circuit to achieve this, I have decided to make it downloadable.
Here it is :
gtr-synth


Printed circuit boards and component layouts
CA3096 PCB

Component layout






Get the schematic as a PDF file
Get the PCB as a PDF file
2SC1583 PCB
Component layout





Get the schematic as PDF file
Get the PCB as a PDF file




Building details
CA3096 version
reference
 value
 quantity
U1
 CA3096 ou HFA3096
 1
U2
 TL072
 1
D1,D2,D3,D4,D5,D6
D7,D8
 1N4148
 8
R25
 82
 1
R11
 390
 1
R16
 470
 1
R7,R8,R9
 1K
 3
R23
 1.2K
 1
R6,R10,R13,R15
 2.2K
 4
R24
 3.9K
 1
R12 8.2K
 1
R1,R2,R17
 10K (ajust to obtain the desired input and output levels) 3
R3,R4,R5 47K
 3
R21
 68K
 1
R19,R20,R22
 220K
 3
R18
 470K
 1
R14
 2.2M
 1
C2,C3,C4
 2.2nF
 3
C9,C10
 100nF
 2
C5
 680nF
 1
C1,C8
 10µF 25V
 2
C6,C7
 47µF 25V
 2
T1
 trimmer 2.5K
 1
P3
 2.2K lin or 2.5K lin
 1
P1,P2,P4
 47K lin or 50K lin
 3
SW1
 rotary switch 3 circuits 4 positions (make before break)
 1
2SC1583 version
reference
 value
 quantity
U1
 TL072
 1
U2
 TL071
 1
Q1, Q2
 2SC1583
 3
D1,D2,D3,D4,D5,D6
D7,D8,D9,D10
 1N4148
 10
R18
 82
 1
R11
 390
 1
R7,R8,R9,R23
 1K
 4
R6,R10
 2.2K
 2
R17
 3.9K
 1
R14
 4.7K
 1
R1,R2,R15 10K (adjust to obtain the desired input and output levels) 3
R13
 22K (change to 27K if using a 25K resonance pot)
 1
R3,R4,R5,R21
 47K
 4
R22
 150K
 1
R19,R20
 220K
 2
R16
 470K
 1
R12
 2.2M
 1
C2,C3,C4
 2.2nF
 3
C9,C10
 100nF
 2
C5
 680nF
 1
C1,C8
 10µF 25V
 2
C6,C7
 47µF 25V
 2
T1
 trimmer 2.5K
 1
P3
 22K lin or 25K lin (if using a 25K pot change R13 to 27K)
 1
P1,P2,P4
 47K lin or 50K lin
 1
SW1
 rotary switch 3 circuits 4 positions (make before break) 1
Wiring



Front panel
Panel design


 silkscreening and drilling
   
               
Download the silkscreen mask as a PDF file

Download the silkscreen mask as a JPEG file



Trimming
Trimming is very easy. Adjust the 2k trimmer in order to obtain a V/octave tracking.


References
Here are interesting links where to find schematics and infos about the Steiner VCF :
Original Nyle Steiner's paper http://yusynth.net/archives/
Ken Stone's site http://www.cgs.synth.net/modules/cgs35_syntha_vcf.html
Motohiko Takeda's site http://www.aleph.co.jp/~takeda/radio/steinerVcfE.html
Marc Bareille's site http://m.bareille.free.fr/modular1/vcf_synthacon/vcf_synthacon.htm
RadioJunkBox site http://homepage2.nifty.com/rjb/labo.htm

You can also purchase a ready made Steiner VCF at Cyndustries !
Doepfer is proposing a Vactrol based Steiner VCF.

By the way, pay a visit to Nyle Steiner's homepage !

The DIY builders' gallery
Here are the photographs of the yusynth MiniMoog VCF modules built by other synth geeks around the world.
Thank you  guys for sending me these nice photos.

Name : Meic
Modular project : Scaff beat
Location : Hamburg, Germany
Website : http://www.myspace.com/scaffbeat 		Name : Olivier Abplanalp
Modular project : Cave Studio
Location : Schaffhausen, Switzerland
Website : http://www.cavestudio.ch
Name  : Patrick
Pseudo : Baronrouge
Modular project: JHC live lab
Location :Toulon, France
Web site : http://myspace.com/patjhc


Name  : Julien
Pseudo :
Modular project:
Location: France
Web site : 		Name  :
Pseudo : Sebo
Modular project:
Location: France
Web site :http://www.cosaquitosenglobo.com.ar 		Name  : Frédéric Monti
Pseudo : Zarko
Modular project:
Location: Gradanne, France
Web site :
back to summary