Numerical intergration of a set of measured data points

[homim4]

Hi,
I have faced the following question. In our lab we perform different measurements on Transistors. We program a scope and that controls the tests. For one of our tests we would like to calculate the total charge Q. Mathematically this is given by
Q=∫ dt i(t), where i(t) is given by i(t) = v(t)/R, i(t)= current, v(t)=voltage, R= resistivity ( constant value), for a time interval [t1, t2]
We cannot measure i(t) directly but we can measure v(t). This means
Q=∫dt v(t)/R.
We have noticed the following
1. if we define Q1 = ∫dt ( v(t)/R)
2. or if we define Q2= (∫dt v(t))/R
Mathematically these two integrals should produce the same result. But we see completely different results. I wonder if this depends on the numerical methods used in integration or has another reason.

Thank you for your help.

 

[marcusl]

If R is truly a constant, then the answer cannot depend on when you divide. You must be making a mistake in your calculations.

 

[Integral]

You have not told us how you are evaluating the intergrals. Without that bit of information there is no way we can hope to help you.

 

[homim4]

Thank you for your feedbacks,
1. R is constant and I expect that the integrals should be the same but they are not.
2. I do not know what integration method is used. This is a tester connected to a circuit we can enter the function that we want to evaluate, with limits of integration.
I have been thinking if we use spline then difference between the two way of integration may come from approximation using (v(t)/R) all the way instead of just multiplying (1/R) at the end.