How-to set pre-specified time for amplitude measurement vi?

Question

In acquiring data at 100 Hz, I'm writing this to file while using the Amplitude & Level measurement VI to decide when the data has peaked. I've built arrays to accumulate the data but I don't see a simple timing scheme to continue this for one or five minutes.
This is a GPIB instrument read over a GPIB-USB-B adapter in LV 7.0 running on Windows XP Pro. The computer is a Pentium 4, 2.4 GHz, with 512 MB RAM.
Attachments:
Input_fluke_with_peak_array_2620.vi ‏329 KB

Answer 1

So I do not have the Fluke or the driver to see how your code runs. From what LabVIEW tells me with the wire size, it looks like you are reading one point at a time from the instrument. You will need to address this so you can get useful data. It is not that reading one point at a time is bad--it just slow. Often it is more desirable to collect more than one point at a time to get a fixed dt between the points. I would focus on seeing the data in a graph first.
Once you are able to get the data comming in a numeric array then you can focus on termiating the acquistion when a peak is found. There are several ways to define a peak. You can say it is a point that is larger than the surrounding points but this will only show a local maximum. Noise will often cause a peak to be undesirably found. Another method for peak detection takes in account hysterisis. This says that if a high value is n units higher than later points, it must have been a peak. Look for hysterisis on NI's site for a better description. Another method for peak detection says that any values above x is a peak. This is useful if you want to collect data that is rising until 6V. LabVIEW has some VIs for peak dection on arrays. I recommend playing with them but remember you will need to decide what works best in your application.
Now that you have data in the form of an array and a way to analyze it, you are in business. Well almost. How much time can elapse between aquiring data and analyzing it. If it is not critical, you can aquire 1000 points, analyze it, determine if you need to aquire more (use a while loop) and iterate or not. If the timming is more critical then you can use a model for VIs where there are two while loops in parallel. In one loop, you can aquire data and feed it to a queue. This allows data to be transfered to the second loop that can dequeue the data and analyze it. When the analysis indicates the data acquision loop can be triggered to stop when it returns with data. This method allows for data to be streamed off the device as fast as possible and then later processed in another loop while the VI is waiting for more data. It is very slick and difficult to code. Look in the LabVIEW shipping examples for queue to see examples of queue use. This data processing loop can write to file as the data is avaliable (slow) or it can save the data and write to the file at one time once the peak has been detected (fast). You can even add logic to only write data up to the peak. (the ease of this depends on how to detect the peak).
I do not know if I have helped or not. It is a lot of stuff to consider. I recommend looking for a simmilar example program. This can really jumpstart any project. Look in the shipping examples and use the advanced search page. I cannot say enough good things about how well that searches NI's site for example programs, knowledgebase articles, discussions, etc.