LED modelling problem

glenn_nha

Joined Feb 4, 2012
1
I have tested a LED and found the Vf and If and plotted the graph. Now i want to model this in pspice for this i need to alter the default values of Is,N,Rs,Ikf in order to match the default LED characteristic with my own. As of now i have got only advice to do this by trial and error is there any other method please advice
Thanks

PaulEE

Joined Dec 23, 2011
423
I have tested a LED and found the Vf and If and plotted the graph. Now i want to model this in pspice for this i need to alter the default values of Is,N,Rs,Ikf in order to match the default LED characteristic with my own. As of now i have got only advice to do this by trial and error is there any other method please advice
Thanks
Well,
An LED is just a diode that lights up. It stands to reason that you could apply a diode model with the given data.

For example, I think "Is" is scaling current in your list of symbols, "N" is the ideality factor (I THINK), Rs is the series resistance, and Ikf I'm not sure about.

The diode equation is Idiode = Is e^ (diode voltage/ nkT) where nkt is the ideality factor, k is boltzmann constant, and t is temperature in kelvins.

That product is the thermal voltage, which is estimated to be 25 - 26 mV.

So, with your Vf, I characteristics, you should be able to plug them into the diode equation and solve for the unknowns. Also, look into what Ikf is...diode current?

http://en.wikipedia.org/wiki/Diode_modelling Look at Shockley Diode Equation in Large-signal modeling...

Hope that helps.

EDIT: further digging suggests that the Ikf parameter has to do with gain variation with differing current levels, particularly with BJTs. Perhaps you can use the default diode parameter for a worst-case approximation.