Time consume by Daq Task

[Newbee_2]

Hi everybody

I am doing a normal DAQ operation(I,e DAQ Create Virtual channel, DAQ start, DAQ write, DAQ stop) I have not incorporated any delay so I expect a spike but all I am getting a pulse which is 7-9mS in duration. Is it normal condition or some problem in DAQ module(I am using NI-9264 AO module). Please give some suggation...

Thanks in advance....

[aeastet]

Post your code or a picture of your code so that we can see what you are doing.

[DFGray]

Please explain yourself better as well.  Is the pulse you are referring to a pulse in processor activity or a pulse on the output of your AO line?  The AO output can not change faster than the output of the device is capable (about 40µs for the 9264).  In addition, the 9264 has no internal timers, so the timing on this is dependent upon what you plugged it into - cRIO chassis, DAQ chassis, single-slot USB chassis, etc.  The more info you give us, the better we can help you.

[Newbee_2]

Thanks for spending some time and Yes Mr. DFGray pulse on output of the AO line and I Know it can not change faster than 40 us and if changes in about 100 uSec it is good for me but it is more in my case and i plugged DAQ chassis. I am also attaching my VI with this (this is simple version of the actual VI but it is also generation a clock pulse of 7-9 mSec actual VI there is delay of 5 mSec which is acceptable)

[GerdW]

Hi Newbee,

why do you create and destroy the channel in each iteration?

Wouldn't it be better to create the channel once before the loop starts and to destroy it after the loop has finished?

[Newbee_2]

In the Main Code the DAQ channel is created and destroyed only once( As I said before it is only a sample) and we are also using the other AO line for different Analog output(9264 has 16 AO line). Can that make the signal generation delay because some time I am getting 500 uSec... 

[DFGray]

Your example is using digital output, but your board is an analog board.  I would recommend you use an analog waveform for your output.  That way you can precisely determine the pulse width using the waveform timing characteristics.  You may want to set up timing and triggering, as well, depending upon what else you are doing.

Your startup delays are going to be variable, depending upon what you have done before you run the VI.  If you need to load the drivers, it will take awhile.  If the drivers are in memory and the VI has been run once, the delay will be a lot less due to several layers of caching (the operating system, DAQmx, and LabVIEW).  The delay from when you fire the run VI to when the output appears should be fairly constant.  Note that unless your operating system is LabVIEW RT, 30ms to 1s random delays are a fact of life.  They do not always happen, but they do sometimes and you need to prepare for them.  The best preparation is to do as much of your timing and triggering at the board level as you can.

Let us know if you need more help.

[Newbee_2]

Thanks sir but when I run the VI for the first time the Pulse is OK but as the VI is running for a longer period of time this problem since to fairly common as some time when n no of pulses are their the last pulse are the main problem as it is high for random no of time( once I measured 12 sec). What could be problem is it task is not properly cleared or O/P of ni board is a not switching properly. I am not able to rely on it