Home > news > Content
Questions about the acquisition frequency of NI data acquisition cards
- 2020-06-24-

Regarding the acquisition frequency of the NI data acquisition card, it is not equal to the actual acquisition time interval.

1. First understand a basic concept, the sampling rate of the capture card depends on the main time base frequency. For example, the main time base of the sampling card NI9234 I use now is 13.1072Mhz, so our available sampling rate:

2. Second, we need to introduce a frequency division coefficient: used to divide the sampling clock frequency you want from the time base frequency. You can get this frequency division coefficient through the interface DAQmxGetSampClkTimebaseDiv(taskHandle,&uint32data);

When you are not the 31 available sampling rates, the frequency division coefficient you get is not 0. You can get this from the interface, but I don't know how to calculate the frequency division coefficient. The actual sampling rate has such a relationship with the sampling rate you input.

In my test, when the sampling rate is 20020, the frequency division factor obtained is 512. The main time base is 13.1072Mhz, and the actual sampling rate in the ni capture card is 25600; your sampling time is 0.00003906 at this time.

At the same time, during the test, I found that the frequency division coefficient is also 0 when the 31 available sampling rate ranges are not much different. If you know the specific situation, I hope the great god will tell you.

Because the conversion accuracy is limited, quantization errors will inevitably occur when the analog value is digitized. The quantization error is determined by the converter and its error, noise and non-linearity. When the input signal and the counter time base are different, a quantization error occurs. According to the matching degree of the phase of the input signal and the time base of the counter, the counter has the following three possibilities:

  • Lost 2 edges-the counter did not recognize the first rising edge and the last falling edge of the time base, and the count value is 1 less than the expected value.

  • 1 missing, identify 1-the counter identifies the first rising edge or the last falling edge of the time base, and the count value is consistent with the expected value.

  • Identify 2 edges—The counter identifies the first rising edge and the last falling edge of the time base, and the count value is 1 greater than the expected value.

For example, if the counter time base is 20 MHz and the frequency of the input signal is 5 MHz, the measured value may be 3, 4, or 5 depending on the quantization error. The measurement frequency corresponding to the measured value is 6.67 MHz, 5 MHz and 4 MHz, with an error rate of 33%.