Using Prime 9.0
Hello,
I have a large matrix of time and current data for 6 different channels. The even columns (0,2,4,6,8,10) contain the time data for the odd current data columns to the right (1,3,5,7,9,11) - Excel image shown below for visual clarity.
I am trying to sum up the current data of the 6 channels into a 2 column matrix (time, current) adding the currents at the appropriate times. The time data is in microseconds with time steps of ~.006-.007 μs and the channels time data overlap with start times shifted by 7.5 μs. For example, Channel 0's time data spans from 7.5 μs to 41 μs, Channel 1's time data spans from 15 μs to 48.5 μs, and so on. I wrote the matrix data to an excel file and attached.
Excel Image:
MathCAD matrix:
It's kind of hard to explain verbally, but basically the channels have time and current data that overlap, and I want to add the currents of the overlapping times accordingly.
I have a user-defined function that simply adds up the odd current data columns together, but doesn't take into account the time data at all...
Let me know if further clarification is needed, and I appreciate any help with this!
ACCEPTED SOLUTION
Accepted Solutions
So you simply want to add up the six sine-looking signals?
Doing it with the values in the table might be problematic.
channel 0 might have values at time stamp 80 ms and 82 ms but channel 1 might have timestamps 79 ms, 81 ms and 83 ms.
So I guess you will want to interpolate?
My suggestion would be to create interpolating functions (linear interpolation with "linterp" might be all thats needed), limit the function to the appropriate time range for each channel and then create the sum of all six functions.
Would this approach do the job?
Of course you now could sample the sum signal using time steps of your choice if you really need a vector of values.
Here is what I was talking about. Not sure if its what you are looking for, though
Prime 9 file attached
Hi, it might be easier to understand with a simple small scale example. I attached the Prime sheet. In this case, the output matrix contains the whole time range from 0 to 1.5 seconds, and the current data is summed together from all the channels - constant value of 1 A for simplicity. When I said "overlapping" before, I meant the portion of the channels time data of the same range. For example, Channel 0 is from 0 to .9 s, and Channel 1 is from .2 to 1.1 s, so .2 to .9 s are "overlapping" time values and the current data at each matching time point should be added up. In this example, at the time of 1.0 seconds, the output matrix has a value of 3 Amps since Ch1, Ch2, and Ch3 all contain the 1.0 time value which have values of 1 A each.
In the original application, the time steps between the channels might not be exactly uniform between the different channels. Likely within .002 μs, if not the exact values.
Hope that makes sense.
So you simply want to add up the six sine-looking signals?
Doing it with the values in the table might be problematic.
channel 0 might have values at time stamp 80 ms and 82 ms but channel 1 might have timestamps 79 ms, 81 ms and 83 ms.
So I guess you will want to interpolate?
My suggestion would be to create interpolating functions (linear interpolation with "linterp" might be all thats needed), limit the function to the appropriate time range for each channel and then create the sum of all six functions.
Would this approach do the job?
Of course you now could sample the sum signal using time steps of your choice if you really need a vector of values.
