Analysis of Neve High/Low Shelving Tone Control Circuit

As used in the 1066, 1073, 1084 and more

The Neve high/low shelving circuit is a variation on a Baxandall tone control. The vintage de-
sign features a custom, discrete inverting amplifier. For ease of analysis, an ideal opamp can be
substituted as an approximation.

To find the frequency response of the circuit, the ratio VVout

in
needs to be determined. Nodal analysis
is performed to yield a system of linear equations, which are then placed in matrix form and solved
using Cramer’s rule.

Note that input resistor Rin is not part of the tone control circuit proper, but represents the output
impedance of the circuit providing the input signal. To get a flat response with the potentiometers
set halfway, feedback resistor RF can be set to match Rin .

Nodal Analysis
Treble potentiometer RT is modeled as resistors RT 1 and RT 2 , connected at the wiper. Since R4 , C4 ,
and RT 1 are in series between nodes 1 and 5, their impedances may be combined as ZT 1 . Likewise,
R5 , C5 , and RT 2 may be represented as ZT 2 . Using Laplace variable s (complex frequency):
1
ZT 1 =R4 + RT 1 +
sC4
1
ZT 2 =R5 + RT 2 +
sC5

1
Node #1: Using Kirchhoff’s current law (KCL),

1 1 1 1 1 Vin
v1 ( + + ) + v2 (− ) + v5 (− )= .
Rin R1 ZT 1 R1 ZT 1 Rin

Node #2: Bass potentiometer RB is modeled as two resistors, RB1 and RB2 , connected at the
wiper.
1 1 1 1
v1 (− ) + v2 ( + + sC1 ) + v3 (− ) + v7 (−sC1 ) = 0
R1 R1 RB1 RB1

Node #3: The positive and negative inputs of an ideal opamp are modeled as having no voltage
difference between them. Since the positive input is grounded, the negative input (and therefore
the connection to R3 at node 8) is considered grounded as well.
1 1 1 1 1
v2 (− ) + v3 ( + + + sC6 ) + v4 (− ) + v7 (−sC6 ) = 0
RB1 R3 RB1 RB2 RB2

Node #4:
1 1 1 1
v3 (− ) + v4 ( + + sC2 ) + v6 (− ) + v7 (−sC2 ) = 0
RB2 R2 RB2 R2

Node #5: As with R3 above, the connection of C3 at node 8 is considered grounded since it is
connected to the negative opamp input.
1 1 1 1
v1 (− ) + v5 ( + + sC3 ) + v6 (− )=0
ZT 1 ZT 1 ZT 2 ZT 2

Node #6:
1 1 1 1 1 1
v4 (− ) + v5 (− ) + v6 ( + + ) + Vout (− )=0
R2 ZT 2 R2 ZT 2 RF RF

Node #7:
1
v2 (−sC1 ) + v3 (−sC6 ) + v4 (−sC2 ) + v7 ( + sC1 + sC2 + sC6 ) = 0
R6

Node #8: The inputs of an ideal opamp are modeled as conducting no current.
1
v3 (− ) + v5 (−sC3 ) = 0
R3

Matrix Form
There are 8 node equations with 8 node voltage variables. These can be stated in matrix form

Ax = b

2
where A is a matrix of the coefficients (admittances), x is a column vector of the variables (node volt-
ages), and b is a column vector of the right-hand sides of the equations (inputs and constants).

− R11 − Z1T 1
 
Y1 0 0 0 0 0 v1
  Vin 
Rin
− 1 Y2 − R1B1 0 0 0 −sC1 0 
 R1  v 2
  0 
− R1B1 − R1B2 −sC6
 0    
Y3 0 0 0  v3 
   0 
  
1 1

 0 0 − RB2 Y4 0 − R2 −sC2 0   v4    0 
   
 1 =
− ZT 1 0 0 0 Y5 − Z1T 2 0 0  v5   0 
  
1 1 1 
− R2 − ZT 2 − RF  v6    0 

 0 0 0 Y6 0   

0   v7   0 

 0 −sC1 −sC6 −sC2 0 0 Y7
0 0 −1 0 −sC3 0 0 Y8 Vout 0
R3

For readability, the admittances along the diagonal of A are given separately here.
1 1 1 1 1
Y1 = + + Y5 = + + sC3
Rin R1 ZT 1 ZT 1 ZT 2
1 1 1 1 1
Y2 = + + sC1 Y6 = + +
R1 RB1 R2 ZT 2 RF
1 1 1 1
Y3 = + + + sC6 Y7 = + sC1 + sC2 + sC6
R3 RB1 RB2 R6
1 1
Y4 = + + sC2 Y8 = −sCF
R2 RB2

The output voltage, Vout , can now be found using Cramer’s rule.
 
A8 
 
Vout =  
A
 

where matrix A8 is formed by replacing the 8th column of A with the contents of b. Since b
contains only one non-zero element, the determinant of A8 is equal to that element multiplied by
its cofactor.
  Vin Vin
A8  = C1,8 = (−1)1+8 (M1,8 )
Rin Rin
V
A8  = − in M1,8
 
Rin
where M1,8 is the determinant of A8 with row 1 and column 8 removed.
 1 
−
 R1 Y2 − R1B1 0 0 0 −sC1 
1
 0
 − RB1
Y3 − R1B2 0 0 −sC6 
 0
 0 − R1B2 Y4 0 − R12 −sC2 
 1
M1,8 = − ZT 1 0 0 0 Y5 − Z1T 2 0 

− R12 − Z1T 2

 0 0 0 Y6 0 

 
 0 −sC1 −sC6 −sC2 0 0 Y7 
− R13

 0 0 0 −sC3 0 0 

3
Substituting into the equation for Vout ,
Vin M1,8
Vout = −  .
Rin A

The transfer function of the circuit is then found to be

Vout 1 M1,8
=−  .
Vin Rin A