This is not about programming but about methodology.
How do you want to detect a peak? If your signal is noisy, you cannot use derivatives, or you have to filter data (with a gaussian filter or with a local mean of data), but this however will include a delay in the filtered data, as when the signal starts increasing the mean will be lower than the actual value. It will also flattent your peaks if they are really sharp.
So, here, you have to decide. How is your noise? If your noise as a defined max amplitude, you can use thresholds to detect when a signal starts increasing or lowering. If not, you'll need to filter the data and what could be a nice solution is to filter both predictable data and measured data, so the error introduced in filtering will be similar to the two graphs, fixing your timing issues.
Whatever solution you choose, you can then programm it on LabVIEW, but what you have to do first is take a piece of paper and a pencil and think about what is your problem and how you can resolve it.
Hope that helps
Oh, and, LabVIEW has no benefits in developping examples that will serve one single customer. If lot of people had the same issues, there will be an example. If you cannot solve your specific problem, then you need to pay for a developper ^^
Message Edited by Swieser on 12-16-2008 07:17 PM
