LabVIEW

cancel
Showing results for 
Search instead for 
Did you mean: 

First time implementation of DAQ: need suggestions on timing setting

Hello,

I am first time planning to use a DAQ to acquire the analog signal as shown in the image signal and want to take samples in a period of a second to check how many peaks are present in this time interval.But quite confused with the timing setting in the DAQ assiatance like what the generation mode,samples to write,rate represent and what are the values I need to use to get a samples in a time interval of second where the post process of finding the peaks will be done from thses samples from the image shown above.

0 Kudos
Message 1 of 5
(2,568 Views)

For general information, I would suggest looking at the help information for the DAQmx Timing VI, which would tell you about how some of the functionality in the express VI operates, which you could adapt for your application

 

http://zone.ni.com/reference/en-XX/help/370469AA-01/lvdaqmx/mxtiming/

 

For more specifics, you would need to provide more information about how you want to sample and measure your signal, such as sampling rate, finite or continuous sampling etc...

 

--dave

0 Kudos
Message 2 of 5
(2,518 Views)

Hello PatanGova,

 

I will explain how the timing settings in DAQassistant works for you. Exactly how you will set them up is up to you and the rate you need to sample in to detect these peaks.

 

1 Sample (On Demand) means software timed, you have to place the DAQassistant in a While loop to be able to read multiple sampels. However this will probably not be fast enought looking at your picture over your original scope.

 

N Samples is a finite mode, you will measure a signal and measure the the number of Samples to Read at the rate you specify. For examples you set the rate to 1kHz and you set to read 1000 samples. That would be 1000 samples scattered evenly over 1 second. Because the time it takes to do 1000 samples at 1kHz is this formula (rate/samples to read).

 

Continuous samples is very much the same as N samples, just that you continuously sample. So you set how many samples you want to read each time and at what rate. As long as you set it to the same number you will have 1 seconds worth of data. How many peaks you detect will depend on your sample rate.

 

You need to samples at least twice as fast as the real signal to be able to detect the frequency of the peaks. If you would like to see a nice view over the peaks and see how they change over time you need to sample maybe 10 times faster then the peaks are occuring.

 

If you need further assistance please post more information about the sample rate you measure on your scope in your attached picture. It would also be nice to know how good resolution you want over your peaks and what frequency you expect to get them.

Best regards
JM, LabVIEW CLA
0 Kudos
Message 3 of 5
(2,507 Views)

I'd like to add that while true Nyquist states you have to sample it twice as fast, real world signals don't always hold up to that and a better rule of thumb is 10 times to be able to understand the wave.  So if your peaks happen once 1ms (1KHz) then you should sample it at 10KHz.

0 Kudos
Message 4 of 5
(2,501 Views)

Nyquist Theorem is very much true, but as I explained that is just to detect the frequency and if he needed a nice view over the peaks he needs to sample faster and NI's recommendation is 10 times faster, but if he needs more then 10 points/peak he needs to sample that much faster that he needs resolution of his peaks.

Best regards
JM, LabVIEW CLA
0 Kudos
Message 5 of 5
(2,499 Views)