0
\$\begingroup\$

I am trying to model a voltage-dependent current source. The current has the following dependence:

I = A + B*exp(voltage/C),

where A, B and C are constants and exp is the exponential function.

The voltage itself is between the two locations in the circuit.

enter image description here

Is such a kind of formulation possible in LTspice?

\$\endgroup\$
3
  • 2
    \$\begingroup\$ You probably have to use a B source for this. \$\endgroup\$
    – Hearth
    Mar 13 at 23:19
  • \$\begingroup\$ Why not try it and see? \$\endgroup\$
    – Andy aka
    Mar 13 at 23:37
  • \$\begingroup\$ I’m a bit confused. Is the intended current source you want to build independent of that circuit you drew or is it a part of it? Put another way, is G1 your attempt at creating the intended source or is it its own thing? \$\endgroup\$
    – Ste Kulov
    Mar 14 at 2:14

2 Answers 2

1
\$\begingroup\$

you may need use B. Arbitrary Behavioral Voltage or Current Sources in the LTspice, you can find detailed introduction from LTspice help.

here's an example for you. A=0.5 B=1 C=2 x and y are node name of termial. enter image description here

\$\endgroup\$
1
  • \$\begingroup\$ I tried with the arbitrary behavioral current source and it works. Thanks a lot! I now have to introduce a time-varying capacitor in parallel to the current source. \$\endgroup\$
    – RDS
    Mar 15 at 21:39
1
\$\begingroup\$

Note that the configuration in your schematic, where the current flows between the same nodes at which the control voltage is measured, means that your VCCS is actually just a nonlinear resistor.

I = A + B*exp(voltage/C)

A nonlinear resistor with this relationship (assuming you've got the direction of the current source in your diagram the wrong way around in your schematic and you don't actually intend to model a negative resistance device) can be modeled by a constant current source in parallel with a PN-junction diode, although the IS (reverse saturation current), N (emission coefficient), and EG (activation energy) parameters might have to take values that aren't realizable in a real diode, depending on the values of B and C that you need.

\$\endgroup\$
2
  • \$\begingroup\$ Yes. I agree. Usually in LTspice if you wanna tie the control nodes to the terminals you would use the g2 symbol instead of the g symbol. \$\endgroup\$
    – Ste Kulov
    Mar 14 at 14:48
  • \$\begingroup\$ Thanks a lot, all, for your comments. I'll try out the suggestions and report back on how it worked. The trio representation of a current source, diode and capacitor is to mimic a sheath in a plasma and hence this strange formulation. \$\endgroup\$
    – RDS
    Mar 14 at 22:46

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.