Can DAQmxRegisterEveryNSamplesEvent function be used for the Counter Input Channels

Question

Hi, All,
I have an application to count the digitial pulse number. I want to know the time of the coming pulses which start from 1 and increased 4 later, like 1, 5, 9, 13..... The time interval between each pulse is not a fixed value. So I tried to use  DAQmxRegisterEveryNSamplesEvent and DAQmxCreateCICountEdgesChan functions. But afterI called the DAQmxStartTask function, it always failed.  The board I used is the NI-PCIe-6320. Here is part of my code.
DAQmxErrChk (DAQmxCreateTask("",&m_taskhandle));
DAQmxErrChk (DAQmxCreateCICountEdgesChan(m_taskhandle,"Dev1/ctr0","",DAQmx_Val_Rising, 0, DAQmx_Val_CountUp));
DAQmxErrChk(DAQmxRegisterEveryNSamplesEvent(m_taskhandle, DAQmx_Val_Acquired_Into_Buffer, 4, 0, EveryNSamplesCallback, this) );
DAQmxErrChk (DAQmxStartTask(m_taskhandle));
I don't know the reason. Can Anyone give me some help. Thanks.
Yang
Solved!
Go to Solution.

Answer 1

Hi, John,
What's I what to do is that I have an external digital signal, when the signal goes from low to high, I need do some operations. So I just want to know the time when the digital signal goes from low to high.  The frequency of the digital signal is from 0 - 20000Hz. My first plan is to use a counter to count the pulse rising edges. When the number of edge increases one, I do the designed operations. In this method, I can't guarantee I can catch each pulse rising point. Because sometimes the counter will increase more than one in the on-demand mode. For example, the previous content of the counter is xxxx4, but the next reading the content may become xxxx6.
After I learned some ANSI C  codes, I found that the ChangeDetectionEvent can help me found time when the signal goes from low to high. So I wrote the following code: (I set the digital signal to P0.0)
DAQmxErrChk (DAQmxCreateTask("",&taskHandle));
DAQmxErrChk (DAQmxCreateDIChan(taskHandle,"Dev1/port0/line0","",DAQmx_Val_ChanPerLine));
DAQmxErrChk (DAQmxCfgChangeDetectionTiming(taskHandle,"Dev1/port0/line0","",DAQmx_Val_ContSamps, 1));
DAQmxErrChk (DAQmxRegisterSignalEvent (taskHandle, DAQmx_Val_ChangeDetectionEvent, 0, Callback , dialog));
DAQmxErrChk (DAQmxStartTask(taskHandle));
int32 CVICALLBACK Callback (TaskHandle taskHandle, int32 signalID, void *callbackData){
CNIPCIE6320Dlg* dialog = static_cast<CNIPCIE6320Dlg*>(callbackData);
dialog->m_event_num++;  // Here is to calculate how many events has been triggered
return 0;
Yesterday I used this method to get the pulse rising time. I found the m_event_num is close to the ground truth (In the simulation, I know how may rising pulse I generated).  In one two hours test, the offset is almost 400 pulses. But today I did the same tests again. I found the m_event_num got a big difference from the ground truth. The voltage of the whole system today maybe smaller than yesterday, because today the power is from an invertor while yesterday the power is from the electricity supply. So will the lower voltage supply will causing some event missing. Or if PC system became busy will ignore some callings of the event callback function? That's two possible reason will affact the final result. I am not sure. Do you have any experience about this problem.  Or do you have some better suggestion to get the pulse rising time. Thanks again.
Regards,
Yang