Integrator added DC offset

[Dennis D.]

Using express VIs I generated a sine wave and integrated it twice. The integrator VIs added a DC offset to the result.  Can anyone tell me why?  The VI is included.

[johnsold]

Dennis,

I think that the issue is related to sampled data versus continuous mathematics. If you change the format of your frequency control knob to allow any real value you will see that for most values the average offset is zero as expected. For integer multiples of 100 as you had it configured you see the offset effect. If you look at the actual data points used, I think you would find a set which creates the unusual effect. It is similar to aliasing.

Lynn

[mikeporter]

Because mathematically, that's what happens.

When you integrate a waveform you start by calculating a running sum of the values. In a sine wave that is evenly distributed around zero, this results in another sine wave where the minimum value is zero because you have just as many points greater than zero as you do less than zero.

Now, If you then integrate that waveform the result is a waveform that is steadily trending upwards, because again, the first step in the integration process is to calulate a running sum of the value, and because the input values never go negative, the output waveform will trend steadily upwards.

Mike...