Frequency measurement....

[sammy346]

This might sound like a simple question, but my searches on the NI website haven't been too successful.

I am investigating for my company, what hardware we require to implement some test systems for our products. I have an evaluation license of LV but no hardware.

One of the measurements needed will be frequency (digital), up to 4MHz, but I also need high resolution too!

When setting up a signal acquisition task using MAX, there is a frequency measurement option (acquire, counter input, frequency).

In this way, it seems that there is limited programming required to get a frequency value, I'm sure it probably spits out a floating point number... but what is the resolution of this measurement? 

Is it hardware dependant?

If I use a PXIe-6341 X-Series DAQ will it be different if I use another product range or platform?

On the other hand, I have also read a couple of tutorials including one identifying the 3 main frequency measurement techniques.

There was also one which used 2 counters, one for the input signal (unknown frequency) and the other as a 'gate' to control the acquisisiton time.

I have not been able to find detailed/comprehensive documentation about programming the timers/counters in NI hardware. I am not after specific examples, rather just an overview of the options and capabilities, modes and combinations and their associated performance/resolution.

I think that the NI hardware will have enough performance for what I need, but I need to prove first that it is acurate enough so that we don't need to buy a dedicated frequency counter (piece of test equipment).

Please point me in the right direction.

Thanks in advance!!

[ejkaiser]

It is probably overkill to use a DAQ system for the task. I would recommend to buy an apprpriate frequency counter with a computer interface, eg. serial, USB, IEEE488, etc. Then it reduces to reading a string from the device.

Cheers

Edgar

[johnsold]

Sammy,

The resolution of any digital frequency measurement (whether a counter in a box or a DAQ system) is dependent on the duration over which the pulses are counted.  If you count a 4 MHz signal for 1 second, the result will be 4000000 and the resolution will be 1 Hz.  If you count for 1 ms, the result will be 4000 and your resolution will be 1 kHz.

The accuracy is dependent on the accuracy with which the duration of counting is determined.  This is often derived from the internal timebase of the counter.

There are techniques which allow higher resolution in shorter times with much greater circuit complexity.

Also, if you need to measure low frequencies with high precision (which is not the same as resolution or accuracy), it is usually better to measure the period and calculate the frequency.

If you provide the complete specifications you need, someone can help you understand how that relates to equipment.  NI Field personnel are usually pretty good at helping customers determine what equipment is required.

Lynn

[Brian-G]

Hi Sammy,

For much more information about the different frequency measurement methods and their corresponding error, please refer to the X Series User Manual.  See pages 159-169.

http://digital.ni.com/manuals.nsf/websearch/DB99DF5BA93B63B18625777C00684538

This should help clarify some fundamentals, but let us know if you have any specific questions.  Also, what kind of resolution are you looking for?  And what is the frequency range you need to test in a single sample?

Brian