[sdiy] Steiner filter mods

[fmg]

Fredrik,

Q4 and Q5 forms a non-inverting amplifier: voltage at base(Q4) increase,
collector(Q4) current increase, voltage at base(Q5) decrease, collector(Q5)
current (and voltage) increase.
What you see is an amplifier applying positive feedback to itself through
an RC network. In short: An RC oscillator. The diodes D1-D6 are the 'R' part
of the RC network (made so to let the frequency be voltage controlled).

The resonance pot is a basically negative feedback in the amplifier to damp
and stop the oscillations: the smaller the resistance value the higher the
negative feedback or the other way, increasing the resistance (less negative
feedback) it will start to oscillate again. In other words you can see this
filter as an oscillator being constanted spanked to be quiet.

Sorry for my (definitively) non academic explanation.


Fredrik Carlqvist wrote:
> 
> 
> 
> I don't understand how the Q4/Q5 connection works. Does it have a name?
> 
> Is Q4 used as a emitter-follower and Q5 as an inverting amplifier? Q4
> should have "constant" voltage on the collector (15V - Vbe ~ 14.3V?).
> 
> If so, what is Q5 amplifying? And why doesn't the Q5 emitter go directly
> to the +15V supply?
> 
> Q5's collector has very low impedance to the emitter of Q4 or vice
> versa, if the resonace pot is very low. I suppose this would be the
> highest resonance setting?
> 
> Also, why are the diodes D1-D6 placed so assymetrically? Does it matter
> at all if D1 and D6 are left out? Shouldn't the number of diodes be
> equal between each access point, i. e. shouldn't D3 and D4 be duplicated
> or D1 and D6 removed? (I remember seeing a version with eight diodes
> inside the diffamp.)
> 
> Can someone help me out?
> 
> Thanks,
> Fredrik C
> 
> -----Original Message-----
> From: owner-synth-diy at dropmix.xs4all.nl
> [mailto:owner-synth-diy at dropmix.xs4all.nl] On Behalf Of HOSHUYAMA
> Sent: den 6 december 2005 16:52
> To: fmg
> Cc: synth-diy at dropmix.xs4all.nl
> Subject: Re: [sdiy] Steiner filter mods
> 
> Hi Fabio,
> 
> Important thing is just to keep the current of the differential pair.
> Any type of current source should be OK.  Of course, you have to adjust
> V/Oct scale.
> 
> Actually, R13 is not needed.  I used it just because the PCB from
> Motohiko has the holes and I thought it will save the circuit in case of
> accident.
> 
> If you used an JFET with large Idss for Q2, maybe you should connect the
> gate of Q2 to -15v directly to get appropriate current.
> 
> Sam HOSHUYAMA
> 
> fmg wrote:
> > HOSHUYAMA wrote:
> >
> >>Hello List,
> >>
> >>I also built a Steiner-style filter with simple modifications.
> >>
> >>Schematic:
> >>
> >>  http://houshu.at.webry.info/200511/img_1.gif
> >>
> >>Constant current source Q2 reduces common-mode gain of the
> >>differential amplifier to almost zero, resulting in far less CV
> >>feedthrough.
> >>
> >>Another modification is the output point to the buffer U1.
> >>Response of the resonance pot is more like Korg MS20.
> >>
> >>I also used a diode limiter as Yves Usson.
> >>
> >>Sam HOSHUYAMA
> >>Saitama, JAPAN
> >
> >
> > Sam,
> >
> > Shouldn't gate of Q2 be directly to -15V?
> > I don't understand how Q2 keeps current constant as is. Was a drawing
> > mistake? I'm confused.
> >
> > Fabio
> > {I want an anatomical zoom
> > /Soda Stereo}
> > .
> >
> >