fitting curve to custom equation

I want to use Johnson-Cook equation, so, I have experiment data about three strain rate. (0.001, 1000, 2000) equation : y=(A+B*x^N)*(1+C*log(strain rate))

I should find A, B, C and N from curve fitting. But with the curve fitting toolbox, I got wrong answer.

In this case, how can I fit the curve?

[dpb]

Can't estimate four parameters with only three data points, sorry. You could choose a value for one of the three but would still have no extra DOF so if there's any noise in the acquired data the model estimated could be expected to be less than desirable.