What are the "blind spots" of the Ebers-Moll bipolar junction transistor model?

Question

I saw this sentence from one of @Dave's excellent answers:

If the collector is open-circuit, all of this current flows out the base connection. But as long as there's at least a small positive bias on the collector-base junction, most of the current is diverted to the collector and only a small fraction remains to flow out of the base.

Although I can understand why it's happening, it seems that the Ebers-Moll equations can't predict this, which really concerns me. I want to know what else can't the EM model explain, so that I won't fall into the pit when facing such situations.

As I know so far, "such situations" include

1. High frequency
2. \$V_{bc}=0\$
3. Early effect

Any supplements?

Answer 1

The EM model is a very simple and limited model. If it covered everything there would be no place for NXP's Mextram model, the Gummel-Poon model or the Hicum model.

As a circuit designer you must always be aware that a model is just a model. A model is just a simplified version of reality. So you can never trust a model 100%. How much you can trust a model depends on how you're using the transistor, under what conditions the model parameters have been derived and many other factors.

It is typical for beginner designers to either trust the model 100% or to worry about model accuracy so much that they don't dare to trust anything. It takes experience to judge these things properly.

The trick is to make your circuit designs such that the actual properties of the transistor are less relevant. For example, say I do not trust the Early voltage prediction of the model. Then I would try to use a cascode so that the Early voltage does not matter much anymore.

Oh and a final note: For BJTs in general modeling is quite good. For MOSFETs there are far more issues regarding modeling. Yet still we can make working circuits :-)