Optocoupler Igain and CTR relationship.

Hi everybody,

There are two optocoupler parameters that I need clarification on. Igain against CTR.

In manufacturer's datasheet, specifications are made for CTR but in Spice models are the Igain parameter. Both of these values are entirely different. Is there a relationship between them? How is Igain used in spice simulation? Does Igain have any effect on the choice of components around the opto?

Thanks in advance.

Current transfer ratio is transistor collector current divided
by LED current. I have not seen Igain but you'd think it's
the same thing?

I'd set up with an application-realistic testbench and mess
the params around until results get reasonable. It could be
(often is) that you are better off with a macromodel from
primitives you understand and can debug, than trying to
deal with the mysteries of (especially special-dialect) some
more complex but less inspectable primitive (for example,
never have gotten a winding/core xfmr to work nor any
transmission line except ideal, myself).

Like LED (model Vf & Is), cccs (model current to photo
to current conversion efficiency), BJT (model Tr, fT,
Vce(sat) and so on). In fact if the opto primitive doesn't
have at least most of these as params, I'd quit it now.

Thank freebird for responding.

The parameters are present but sometimes I wonder what can make a whole lot of difference. In all honesty, the difference might be staring you in the face unattended to.

Of course I know CTR but I don't know Igain but it is there in the netlist with values that differ for different optos that come with LTspice and the value is nowhere to be found in the manufacturer's datasheet. It's in a curly bracket meaning that it's subject to change. I think there's a reason why it is there and I am so curious.

It's okay modeling with all the available parameters but just that little touch might yield tremendous improvement. Please take a look at it and just tell me what you think it might be.

"It" = ???

I expect the "Igain" is just a fitting param that goes to CTR,
but is applied to some intermediate element of the macromodel
(or code / behavioral model - I have not seen optoisolators as
a "compact model" element in the SPICEs & Spectre I've used).

If you can get at the model definition, it might express things
in an informative way. But LTSpice also has a bunch of obscured /
encrypted models to it.