data acquisitioni sampling

[heang]

hello

i was using the DAQ assistant to acquire analog voltage. the sampling rate is 1000Hz. the program written is to determine impulse force exerted by rolling object. the impulse force may have 2, 3 or 4 peak depends on the length of the rolling object. different rolling object will have different length. what my question is will the sample in daq assistant can be kept changing according to the length of rolling object. it is done so because if the sample is set too high, 2 subsequent rolling object which suppose 2 have 2 peaks, cannot be differentiated from one rolling object which has 4 peaks. how can it be solved?

the sampling rate i use is 1000Hz as mentioned before. so to determine the peak location(in time), should the index of peak to be divided by 1000? how if the sampling rate used is 2000Hz? the index will be divided by 2000?i know there is knowledge about that. i try to use higher sampling rate to avoid aliasing and to utilize the optimum features of PCI 6036E, but what i found is that the result is getting weird reading.(as i have a comparison result which i obtained before)

regards

[Jeff_T.]

Hi Heang,

I am looking forward to working with you on this application.  Before I can make any recommendations can you please clarify a few points?  My main question has to do with the "impulse force" caused by the "length" of the rolling object.  Can you please clarify what you mean by the "length" of the object and how the length causes different peaks?  I am assuming that these peaks are in the time domain (as opposed to the frequency domain) but I do not understand how they represent impulse force of rolling objects.  Thanks Heang I will be watching for your response.

Best Regards,

Jeff Tipps
Applications Engineer
National Instruments

[heang]

hello
thanks for the reply

the rolling object can be for example, vehicle. the impulse force is from the wheel.it might have 4 axle vehicle. 3 axle or even 2. it all depends on the length of vehicle. if the sample is too small, 6 axle vehicle will be shown in the graph as three 2-axle vehicle. if the sample is set too large, 3 vehicle(continuos flow) will be shown as one 6-axle vehicle.

it requires the changes of the sample.

regards

[Jeff_T.]

Heang,

I hope I am answering your question, but you can change the "samples per channel" for each iteration of your while loop.  I have attached a JPG screenshot to demonstrate.  Please continue posting to this forum if I am misunderstanding you or if you require further assistance.

Regards,

Jeff Tipps
Applications Engineer
National Instruments

[heang]

hello

thanks for the reply
what i was asking is that at first the daq assistant is placed to acquire signal. before signal is processed, the length of that rolling object is now known. should i set the sample higher or lower?i do not know whether it should be set any sample before getting to know its length. as the program is used to determine various rolling object automatically.
what i previously did is after processing, calculate for its length of the rolling object, then feedback node to the sample or add shift register(by using selector and comparison to choose, for example <0.5m the sample is set to 1500. and >1.0m the sample is set to 2500. )
however, it appears to be lagging. it seldom goes to 2500 as it has constraint.( i do not know the constraint). I think logically that the sample is processed, then know its length, then feedback to determine its sample., but i know it has lagged.


i hope u understnad what i said.

thanks

regards

[Jeff_T.]

Heang,

I think it would be helpful if we return to what your program is designed to do.  It is my understanding that your program is designed to calculate the length of each vehicle based on the number of "spikes" or axles that you detect.  I am assuming that you know the speed of each vehicle and the distance between each axle.  Am I correct in these assumptions?

If the spacing between the axles (and between vehicles) is unknown, then we cannot tell from the graph alone if 4 spikes means a long 4 axle vehicle or two short two axle vehicles.  We will need some other parameters.

As far as the "lagging" goes, it might have something to do with the amount of time that it takes to acquire 2500 samples. When your rate is set to 1 kHz and your samples is set to 2500, it takes the DAQ device 2.5 seconds to acquire that many samples and return it to LabVIEW.  If you specify 100 samples at 1kHz, it takes the DAQ device 0.1 seconds to acquire the data and return it to LabVIEW.  Do you feel the duration of the lagging is related to the size of your "samples to read" field?

Regards,

Jeff Tipps
Applications Engineer
National Instruments

[heang]

hello
sorry for the late reply

what your understanding is correct. however, the 'sample' seems to be important for me to determine whether it is long vehicle or 2 continuous vehicle. i do not know whether the problem arises from the algorithm or the sample i need to acquire.

anyway.
thanks for the help

i will review  the written algorithm and only if i am fail, i will return to you. have a nice day
again, thanks for the assistance

regards