03-09-2015 09:58 AM
With respect to requested sample rate, VS actual sample rate.
For example, lets say that i request 1000 samples with a rate of 1000 samples/second. The sample should cover a period of 1 full second. If i measure an accelerometer at 15 HZ on a calibrated vibration shaker, i recieve back approximatly 9 full sinusoids, instead of 15. I have verified the speed of the shaker with a seperate system, and also a strobe. I Have placed code to time the interval between the analog callbacks and it is on the order of 605 ms +/- 1 ms per analog callback. The strange thing is that i still recieve back 1000 samples, except its at a rate of 1652 HZ (because i am recieving approx 9 cycles instead of 15). Is there an error in the sample project? Am i misapplying this example? Is there a driver error in the .NET SDK?
Its as if the driver is 'scrunching' the measurments into a 605ms window instead of the full second.
This problem can be reproduced using the example named ...NI-DAQ\Examples\DotNET4.5\Analog In\Measure Acceleration\ContAccelSamp_IntClk_AnlgStart\VB
-Comment out the triggering.
-Set rate and nSamples to 1000 (FYI nSamples input is actually hard coded at 1000 in the example)
-Run and continuous aquisition begins.
I need to aquire continuous data, seemless if possible, in 1 second long blocks for the application i have. Please help.
Additional values to show the timing can be placed in AnalogInCallback are shown in the attached rich text file.
Computer: 64 Bit windows 7 professional, niDAQ mx version 14.1
I have also reproduced on a 32 bit system.
03-09-2015 01:56 PM
Hardware is:
cDAQ 9174 - 4 slot compact Daq USB
Slot1: NI9207 (8 channel voltage, 8 channel current)
Slot2: NI9235 (8 ch 1/4 bridge)
Slot3: Empty
Slot4: NI 9234 (4 ch IEPE accelerometer)
03-09-2015 02:33 PM
Hi Alan,
The 9234 has a minimum sampling rate of 1.652kS/s, which is probably what you're seeing. This is shown on page 18 here:
http://www.ni.com/pdf/manuals/374238c.pdf
John Moody
Applications Engineering
National Instruments