I currently am required to use a device that is old. It is a darlington driver that has no known well defined equivalent model (SG2815). When in saturation, the output VCE of the device is spec'd at 1.8V nom and 2.1V max.

The question is, is it valid to use the output transfer characteristic to determin the equivalent series resistance of the device in saturation? For instance, the output characteristic shows a 1V saturation voltage with a collector current of 300mA. That defines a 3.33Ω resistor.

Would a valid simple output model be an ideal voltage source of 1V with a series resistance of 3.33Ω for that collector current?

Any help appreciated.

Thanks!

 

Probably not.

You want to use a resistor value that represented the amount by which Vce will change if the current changes. So estimate the slope of the characteristic curve at that point.

 

Is this for a Spice type simulator? For that you could use two transistors of near equivalent voltage and current rating connected in a Darlington configuration. That should give a reasonable simulation.

 

Yes this is for a Saber Sim. We have tried to model this based upon the spec sheet simplified equivalent circuit, but, the best that we could get is a Vsat of 1.0v. The company that currently makes the device claims to not have the equivalent transistor information as they are fourth owner of the design, and the origonal designers are long gone.

The specific need is to be able to change the Vsat value to confirm operation of a couple of the paths in parallel which include such devices. This is necessary to demonstrate one of the paths hogging current from the other path which could result in a over current condition in the current hog. That is why we chose to use the voltage source - source resistance approach.

 

I agree with Crutschow. It shouldn't take long to find a suitable combination of transistors to demonstrate current hogging.

 

You might try adding a diode or two, or a zener, in series with the collector of the two-transistor model to get your Vcesat up where you want it. Heck, for that matter you could probably add a battery. The key is to not add something to the model that is going to act as ballast since what you are trying to demonstrate will be defeated by such approaches (which includes using a resistor as you were originally proposing).

 

I think that I have uploaded the spec sheet just in case you want to take a look at it. If you can suggest a transistor pair that works, I would be forever grateful. Bear in mind that the ability to force the Vce-sat to a voltage is necessary.