* 8FB2xxx LTspice model based on the generic reverb tank model from Adrian Immler, version i1, Febr. 2020
*A version log is at the end of this file
*Params fitted to measured data from a TAD sample by Adrian Immler, Febr. 2020
*This model considers also inverse mode (input as output and vice versa) but not magnetic core saturation
*
*                 input signal
*                  |  input reference
*                  |   |  output signal
*                  |   |   |   output reference
*                  |   |   |    |
.subckt 8FB2xxx.i1 IN+ IN- OUT+ OUT-
* input coil model
Rin  IN+ 1   200R  ;see http://www.accutronicsreverb.com
Lin  1   14 305m  ;this value leeds to the nom. input impedance of 1925 Ohms @ 1kHz
Rinp 1   14 150k  ;found by Q-factor investigations on a sample
Cinp 1   14 20p5  ;found by Q-factor investigations on a sample
*B4  14 IN- V=V(13a,0) + V(13b,0) + V(13c,0) ;needed for backward direction
Rbypass 14 IN- 1m
*springs model (forward direction)
B1  2   0   V=8.27*I(Lin) ;Input voltage source for the tlines. Gain found by investigations on a sample.
Chp 2   3   800u          ;Because a reverbtank blocks dc signals, a high pass is necessary.
Rhp 3   0   1             ;According to mesaurents, a first order highpass with fg = approx. 200Hz is suitable.
R1  3   4a  620           ;The insertion resistor values in relation to the tline impedances determines the tlines' reflection damping (decay time).
R2  3   4b  560           ;The adequate insertion resistor values were found by impulse response measurements.
R3  3   4c  515           ;To achieve identical decay times for all 3 springs (with their own different delay times), the insertion resistor values have to be different.
Ta1 4a 0  5a 0  Zo=50 Td=13m05 ;Impedances are set arbitrary. Sum of the 2 delaytimes found by measurements.
Ta2 5a 0  6a 0  Zo=60 Td=10m1  ;As measurements shows, there are more reflections then those at both end of the spring. Therefore, each spring is modelled by TWO tlines in series with slightly different Impedances.
Tb1 4b 0  5b 0  Zo=50 Td= 8m4  ;Impedance is set arbitrary. Total delaytime found by measurements
Tb2 5b 0  6b 0  Zo=60 Td=16m98
Tc1 4c 0  5c 0  Zo=50 Td=15m3  ;Impedance is set arbitrary. Total delaytime found by measurements
Tc2 5c 0  6c 0  Zo=60 Td=12m48
*springs model (backward direction)
*B3   9   0    V=9.58*I(Lout);Input voltage source for the tlines. Gain found by investigations on a sample.
*Chpb 9   10   800u          ;Because a reverbtank blocks dc signals, a high pass is necessary.
*Rhpb 10  0    1             ;According to mesaurents, a first order highpass with fg = approx. 200Hz is suitable.
*R4   10  11a  620           ;The insertion resistor values in relation to the tline impedances determines the tlines' reflection damping (decay time).
*R5   10  11b  560           ;The adequate insertion resistor values were found by impulse response measurements.
*R6   10  11c  515           ;To achieve identical decay times for all 3 springs (with their own different delay times), the insertion resistor values have to be different.
*Ta3 13a 0 12a 0  Zo=50 Td=13m05 ;Impedances are set arbitrary. Sum of the 2 delaytimes found by measurements.
*Ta4 12a 0 11a 0  Zo=60 Td=10m1  ;As measurements shows, there are more reflections then those at both end of the spring. Therefore, each spring is modelled by TWO tlines in series with slightly different Impedances.
*Tb3 13b 0 12b 0  Zo=50 Td= 8m4  ;Impedance is set arbitrary. Total delaytime found by measurements
*Tb4 12b 0 11b 0  Zo=60 Td=16m98
*Tc3 13c 0 12c 0  Zo=50 Td=15m3  ;Impedance is set arbitrary. Total delaytime found by measurements
*Tc4 12c 0 11c 0  Zo=60 Td=12m48
*
* output coil model
B2  7 OUT- V=V(6a,0) + V(6b,0) + V(6c,0)  ;The output voltage source gathers the 3 spring signals
Lout  7  8    409m  ;This value leeds to the nom. output impedance of 2575 Ohms @ 1kHz
Cpout 7  8    17p8  ;found by Q-factor investigations on a sample
Rpout 7  8    153k  ;found by Q-factor investigations on a sample
Rout  8  OUT+ 200R  ;see http://www.accutronicsreverb.com
.ends
*Version log
*i1 :Initial version
* 330mH inductor "saturates" at 2A (L*Isat = 0.66 volt-seconds)
*L1 0 1 flux=0.33*2*tanh(x/2)
*flux={L*Is}*tanh(x/{Is}); L= inductance and Is= "saturation" current
*flux={L*Is}*tanh((x/{Is})**{a})**{1/a}; a= saturation "sharpness" (1=normal, >1=sharper)
*.param L=330mH Is=2A a=3; fairly sharp saturation at 2A (the "a" factor is exponential)

