Are v1,v2,v3 digital inputs?  You don't specify what v4 is if (v1+v2+v3) is, say, 2.5 (nor, actually, if it's exactly == 1 or == 3), yet clearly v5 is a real-valued output.

Thanks for the reply.

Let me explain the problem again.

Just trying to realize spice statements like the following into veriloga

.FUNC Q(v1,v2,v3,v4) { (Cg1*v3+Cg2*v4+C1*v1+C2*v2)/E }  
.FUNC VN(v) { CSUM*v/E }                                    
.FUNC GAMMA(u) { IF(T==0,IF(u<0,-u,0),IF(u==0,T,u/(EXP(u/T)-1)))}


.FUNC ALPHA(v1,v2,v3,v4) { ROUND(-Q(v1,v2,v3,v4)+(CSUM/E)*(v1*RN2+v2*RN1)) }  
In the above statements the function N_OPT calls Function Q, CSUM.


.FUNC R1L(n,v1,v2,v3,v4) {GAMMA(0.5 - n - Q(v1,v2,v3,v4) + VN(v1))/RN1} In the above statement the function R1L calls GAMMA.

Now my query is regarding
> best modeling practice regarding function calling functions  the
> number of values that can be returned If  declarations in functions are real and these cant take values which are defined electrical?
For  example:-

`include "constants.vams"
`include "disciplines.vams"

module setmod2(v1, v2, v3, v4,v5);
inout v1, v2, v3, v4,v5;
electrical v1, v2 ,v3, v4,v5;
parameter real C1=1                     ;
parameter real C2=1                      ;
parameter real R1=1                       ;
parameter real R2=1                       ;
parameter real Cg1=1                     ;
parameter real Cg2=0.1                         ;
parameter real C0=0                          ;
parameter real Q0=0                          ;
parameter real TEMP=4.2                      ;

parameter real  PI=3.14       ;
parameter real  E=1.5            ;
parameter real  KB=1.3             ;
parameter real  CSUM={C1+ C2+ Cg1+ Cg2+ C0}     ;
parameter real  T={TEMP*CSUM*5.3785467E14}  ;
parameter real  RN1={R1/(R1+ R2)}            ;
parameter real  RN2={R2/(R1+ R2)}            ;


analog function real Q1;
input v_1, v_2, v_3, v_4;
real v_1,v_2, v_3,v_4;
Q1 =  (Cg1*V(v_1)+Cg2*V(v_2)+C1*V(v_3)+C2*V(v_4))/(E+Q0) ; endfunction

analog begin

V(v5) <= Q1(v1,v2,v3,v4);

endmodule

gandhi_gaurav wrote on Nov 4^th, 2008, 11:20pm:

There's no N_OPT in what you posted ...

Quote:

also no R1L ...