Hi,
I was playing around with creating a zener diode model.
Since a zener looks like two diodes in antiparallel, i placed two diodes in antiparallel with one of them having a zener voltage parameter Vz.
It seems to work and the two 'diodes' can be independently fitted to any zener curve.
I havent actually tried to fit this to any particular zener diode, but it looks like it should work. See if you agree or not.
The simple model is:
Id=Isz*(e^((Vd-Vz)/NVTz)-1)+Isd*(e^(-Vd/NVTd)-1);
where
Id is the current though the pair of diodes regardless of the polarity of Vd,
Vd is the voltage across both diodes (they are in antiparallel),
Vz is the required zener voltage (like say 7.2 volts),
Isz is the Is for the zener diode,
Isd is the Is for the reverse regular diode,
NVTz is the NVT for the zener,
NVTd is the NVT for the reverse regular diode,
either NVT is N*VT where N is the ideality factor and VT is the thermal voltage, although each diode has their own NVT as shown.
What happens is if we make both diodes have the same Is and NVT we get a curve that works the same at each end of the Vd range except for polarity. For example, if we had the zener voltage Vz set at 7.0v then if for Vd=+7.7v we get a current of 33ma then with Vd=-0.7v we get a current of -33ma.
The actual spice model seems to have too many conditionals while this equation is continuous so no conditionals are needed.
If you are interested play around with it a little and see what you think.
I was playing around with creating a zener diode model.
Since a zener looks like two diodes in antiparallel, i placed two diodes in antiparallel with one of them having a zener voltage parameter Vz.
It seems to work and the two 'diodes' can be independently fitted to any zener curve.
I havent actually tried to fit this to any particular zener diode, but it looks like it should work. See if you agree or not.
The simple model is:
Id=Isz*(e^((Vd-Vz)/NVTz)-1)+Isd*(e^(-Vd/NVTd)-1);
where
Id is the current though the pair of diodes regardless of the polarity of Vd,
Vd is the voltage across both diodes (they are in antiparallel),
Vz is the required zener voltage (like say 7.2 volts),
Isz is the Is for the zener diode,
Isd is the Is for the reverse regular diode,
NVTz is the NVT for the zener,
NVTd is the NVT for the reverse regular diode,
either NVT is N*VT where N is the ideality factor and VT is the thermal voltage, although each diode has their own NVT as shown.
What happens is if we make both diodes have the same Is and NVT we get a curve that works the same at each end of the Vd range except for polarity. For example, if we had the zener voltage Vz set at 7.0v then if for Vd=+7.7v we get a current of 33ma then with Vd=-0.7v we get a current of -33ma.
The actual spice model seems to have too many conditionals while this equation is continuous so no conditionals are needed.
If you are interested play around with it a little and see what you think.
