problem: Spice Simulation Model

I want to simulate the following model in proteus.
But this model is made in Pspice, and Proteus only supports spice 3f5


*****************************************************
* GENERATOR_BEH.LIB *
* *
****************************************************
*BEHAVIOURAL MODEL OF A PV GENERATOR
* INPUT PARAMETERS:,AM1.5 JSCMR,AM1.5 VOCMR,AM1.5 PMAXMR
* AM1.5 VMMR,AM1.5 IMMR, CURRENT TEMP COEFF.,VOLTAGE TEMP.COEFF,
* NOCT, REFERENCE TEMPERATURE


* NODES
* (400) REFERENCE
* (401) INTERNAL NODE
* (402) INPUT, IRRADIANCE
* (403) INPUT, AMBIENT TEMPERATURE
* (404) OUTPUT
* (405) OUTPUT, (VOLTAGE) VALUE=SHORT CIRCUIT CURRENT(A) AT
* IRRADIANCE AND TEMPERATURE
* (406) OUTPUT, OPEN CIRCUIT VOLTAGE AT IRRADIANCE ANDTEMPERATURE
* (407) OUTPUT, (VOLTAGE) VALUE=CELL OPERATING TEMPERATURE(ºC)
* (408) OUTPUT, MPP CURRENT
* (409) OUTPUT, MPP VOLTAGE

.subckt generator_beh 400 402 403 404 405 406 407 408 409 params:
+ iscmr=1, coef_iscm=1, vocmr=1, coef_vocm=1,pmaxmr=1,
+ noct=1,immr=1 , vmmr=1, tr=1, ns=1, nsg=1 npg=1

girrad 400 401 value={v(402)/1000*(npg*iscmr+npg*coef_iscm*(v(407)-25))}

eiscm 405 400 value={v(402)/1000*(iscmr+coef_iscm*(v(407)-25))}
evocm 406 400 value={if (v(405)>1e-11, (vocmr+coef_vocm*(v(407)-25)+8.66e-5*
+ (v(407)+273)*log(v(405)/(iscmr))),0)}

etcell 407 400 value={v(403)+(noct-20)/800*v(402)}

gidiode 401 400 value={npg*v(405)/(exp(v(406)/(ns*8.66e-5*(v(407)+273)))-1)*
+ (exp(v(401)/ (ns*nsg*8.66e-5*(v(407)+273)))-1)}
rsg 401 404 {nsg/npg*((vocmr/(iscmr)-pmaxmr/(iscmr**2*(vocmr/(ns*0.0258)-log
+ ((vocmr/(ns*0.0258))+0.72))/(1+vocmr/(ns*0.0258)))))}
.func frsg() {nsg/npg*((vocmr/(iscmr)-pmaxmr/(iscmr**2*(vocmr/(ns*0.0258)-log
+ ((vocmr/(ns*0.0258))+0.72))/(1+vocmr/(ns*0.0258)))))}

gimg 400 408 value={npg*(immr*v(402)/1000+coef_iscm*(v(403)-25))}
rimg 408 400 1
evmg 409 400 value={if (v(402)>0.001, nsg*(ns*8.66e-5*(v(407)+273)*log(1+(v(405)-v(408)/npg)/v(405)*(exp(v(406)/
+ (ns*8.66e-5*(v(407)+273)))-1))-v(408)*frsg/nsg),0)}
.ends generator_beh
*_____________________________________________________

As I make the model compatible with proteus

please help

*****************************************************
* GENERATOR_BEH.LIB
****************************************************
.subckt generator_beh 400 401 402 403 404 405 406 407

girrad 400 401 value={v(402)/1000*(5*0.37+5*0.13e-3*(v(407)-25))}

eiscm 405 400 value={v(402)/1000*(0.37+0.13e-3*(v(407)-25))}
evocm 406 400 value={(21-0.1*(v(407)-25)+8.66e-5*(v(407)+273)*ln(v(405)/(0.37)))}

etcell 407 400 value={v(403)+(47-20)/800*v(402)}

gidiode 401 400 value={5*v(405)/(exp(v(406)/(2.86e-3*(v(407)+273)))-1)*(exp(v(401)/(8e-3*(v(407)+273)))-1)}

rsg 401 404 {2/5*((21/(0.37)-5/(pwr(0.37,2)*(21/(33*0.0258)-log((21/(33*0.0258))+0.72))/(1+21/(33*0.0258)))))}

.ends generator_beh


I have made ​​some changes, and the simulation seems to already be compatible, however when measuring voltage on the output pins 401 and 404, the result is NaN.

please help to fix this

I added a resitencia of 1000ohm to node 401, and I have placed inicial condition = 0 on the same node, now allows me to run the simulation, but does the dc sweep correctly



- - - Updated - - -