Labview help

[MarineOne]

Hello,

I need to write a VI to run a life test fixture. It has one load cell
transducer on it which gives me a sine wave type signal into my IO
card. This sine wave is around 1 hz throught the test which runs for
a few days.

Here is my delima. I want to gather data and save it to my hard drive
but I don't want to fill up my drive. I was thinking of sampling the
transducer at say 1000 samples per second to define the wave. All I
am interested in are two things:

1. How the peak to peak distance changes over the life of the test
(few days).
2. How the mean of the peak to peak shifts over time. It starts at
zero and shifts away from zero as the test goes on.

I was thinking I should only save the data I need and not the whole
sine wave since I don't th
ink I could write the sine wave to disk fast
enough anyway at a sample rate of 1000Hz.

Any ideas how I should aproach this problem?? I just completed the NI
LBI and LB Basic II training coarse. Now I need to apply what I
learned but want to make sure I head in the right direction.

Many thanks.........Joe

[Nadav_Chernin]

1. Rule of Niquist says that sampling of signal must be with frequincy more than 2 maximal freq. of sampling signal. For your signal that around 1hz, sampling rate must be not less than 2 hz. As you can to sample with frequency 10 hz(10 samples on period, this is much more than enough).
2. If you interested only in 2 things: peak to peak distance and mean of peak to peak, you can use peak detection vi on LV toolbar and save only this data to your hard disk.

[Mellik]

Also, if you don't want to exhaust your hard-drive constantly, I
suggest you create a loop with a shift register to write blocks of
data to the disk. e.g. 10kb blocks or so.
This is only important, if you have a lot of data coming in
constantly.
You can of course choose the block size according to cpu and ram in
use.

> 1. Rule of Niquist says that sampling of signal must be with frequincy
> more than 2 maximal freq. of sampling signal. For your signal that
> around 1hz, sampling rate must be not less than 2 hz. As you can to
> sample with frequency 10 hz(10 samples on period, this is much more
> than enough).
> 2. If you interested only in 2 things: peak to peak distance and mean
> of peak to peak, you can use peak detection vi on LV toolbar and save
> only t
his data to your hard disk.