moog hp

[Osamu Hoshuyama]

Hello,

I have an idea to reduce CV feedthrough in the Moog HPF.

> Date:    Tue, 30 Mar 1999 14:30:08 +0300
> From:    Joachim Verghese <jocke at netcontrol.fi>
> 
> Hi,
> 
> > Can you offer any suggestions for a transistor matching procedure that
> > will match NPN and PNP pairs?  Will the Moog matching technique work?
> > (as published in the Minimoog service manual)
> 
> I've also built a phaser using the Moog HPF circuit modified for
> all-pass response. This was a six-stage circuit with op-amp
> buffers.
> 
> The transistors do indeed need to be selected to avoid large
> CV feedthrough, but it's not very easy to match NPNs to PNPs.
> The PNPs I had (ZTX214C, if memory serves) seemed to require
> a significantly higher Vbe for a given Ic, compared to the
> NPNs (ZTX109C).
> 
> What I did instead was that I took six matched PNPs and six
> matched NPNs, and added an offset trim for the CV that drives
> the PNPs. This way, the mismatch between the NPNs and PNPs could
> be compensated for using a single trimmer.
> 
> I didn't end up with a feedthrough-free circuit, though, because
> as has been pointed out before, transistors really need to be
> matched at more than one point on the Ic/Vbe transfer curve.

If you have seventh matched NPN and PNP, this circuit can be used. 

           +15V
            |
            R: Current limitation
            |
            | C
             \|  NPN                            +CV Line
              |---------------+-----------------------
             /|  B  |\        |
          E |       |  \      |
            +-------|-   \    |
            |       |     >---+
            |     |-|+   /    |
            |     | |  /      |
            |   GND |/        |
            |                 |
            |         | |     |
            +---------| |-----+
            |         | | C: Phase compensation
            | E
       B  |/  PNP                                -CV Line
 -CV -----|------------------------------------------
  INPUT   |\
            | C
            |
            R: Current limitation
            |
           -15V

Resistors of the CV inverter are replaced by emitter followers.
Unbalance of NPNs and PNPs will be compensated by this inverter.
I've never built this circuit, but it should work if we select
appropriate capacitance for phase compensation.

> Fortunately, phasers are typically modulated by relatively slow
> signals, so a little CV feedthrough is quite harmless.
> 
> Also, like Osamu, I had to limit the CV range and capacitively
> decouple the CV lines for stability.
> 
> -joachim

- -- ---- ----------- --- --- - ----- -- --- --- - - ------ - - - --
 Osamu HOSHUYAMA          Digital Signal Processing Technology Group
    Dr. (Eng.)      C&C Media Research Laboratories, NEC Corporation
               Email: houshu at ccm.CL.nec.co.jp,  Fax: +81-44-856-2232
--------------------------------------------------------------------