I recently find there is a sampling problem of the Picoscope which is very annoying.
For example, if I send a 1 kHz sinusoid signal generated by Picoscope itself, when the sampling frequency is set at 1 MHz for a 2-second signal displayed on the screen (i.e. 2 million samples), the actual number of samples is 1’990’450, that is 9’550 points less. I have no problem with that, because it seems a hardware restriction of the Picoscope.
But the actual timing is not right either. I read the data using MATLAB (8-bit Picoscope vertical resolution).
The following data are the values read from MATLAB. T is for time, and A is for the amplitude of the captured signal.
T(1) = -1.0048 μs, T(2) = 0 s, T(3) = 1.0048 μs, ... , T(end) = 2 s + 2.09 μs.
A(1) = -0.0157 V, A(2) = 0 V, A(3) = 0 V, A(4) = 0 V, A(5) = 0.0157 V, ... , A(end) = -0.0315 V.
T(1) is equal to Tstart, and T(end) is equal to Tstart+(Length-1)*Tinterval;
A(1) to A(end) are the corresponding reading of the signal amplitude.
Apparently, by observing the values of T, the signal does not start from 0 second, and does not end at 2 second. It seems that the signal has been stretched slightly. Also, the amplitude does not start at 0 V, or end at 0 V.
Do you have any solutions for this? Thanks.