I am tasked with replacing a very old Solatron frequency response analyzer with a new one based on LabVIEW software and NI hardware. The primary purpose of the equipment will be to generate and analyze Bode plots. I thought some of you might have some suggestions as to the best way to get started. What is the best way to do frequency analysis in LabVIEW? Is the functionality built-in, or do I need to look for 3rd-Party products? What are the best DAQs to use in this project?
As you can see, all of this is very new to me. Any help would be appreciated. Thanks!
Solved! Go to Solution.
I am not familiar with the Solatron instrument but generally to measure frequency response one generates an excitation signal which covers the frequency range of interest and measures the ouput of the device under test over the same range. Certainly all of those things can be done with LabVIEW and appropritate hardware. I have been doing that since the LabVIEW 1 days.
But before anyone can advise you on specific processes or hardware, the details of your requiements must be indicated. Is the frequency range 1-100 Hz or 50 MHz to 10 GHz? Is the excitation signal a voltage, a current, or something else like mechanical vibration? How large does the excitation signal need to be? What is the form and amplitude of the response signal? How fast does it need to cover the entire frequency range? Are there noise or interference signals mixed in with the response? Are there safety or equipment protection issues?
Do you need to reproduce all of the Solatron's capabilities or only a subset? Do you need to add capabilities that the old instrument did not have?
Very good questions. We will be dealing with analog voltages, but I don't have specfics as to the frequencies and amplitudes at this point. I will get answers asap.
I finally have answers to your questions.
I need to analyze frequency responses from ~0-50Hz. Voltage amplitude looks like it will be in the 0-0.5V range. The end result will be a Bode plot looking at Amplitude Ratio and Phase Angle as a function of frequency. I need to compare observed values against limits set by the requirements.
Is that enough information to start a discussion? Thank you for your patience with me. I am climbing a steep learning curve here.
That information is quite useful.
Almost any DAQ device can sample signals with that frequency and amplitude range. The lowest frequency you can analyze will be limited by the duration of the measurement. If you measure for one second, then you will have one full cycle at 1 Hz. While it might be possible to get some information about lower frequencies, it will be quite difficult. At the high frequency end the Nyquist criterion requires you to sample at > 100 Hz. Since you want phase information, sampling at a higher rate, such as 1 kHz would be better.
You need to measure both the input and the output so you will need a DAQ device with at least two channels.
LabVIEW comes with several Fourier transform VIs and related functions that can be used for the Bode plot calculations.
So it really does seem that a relatively basic DAQ will work. In researching NI's approach to Fourier Transforms, there is much talk of high speed digitizers and such. Should I conclude that the low frequency range I am working with makes less expensive options viable?
Yes, the less expensive otions should work fine for your application. Of course, NI does not mind if you want to spend more money!
Other things to consider in selecting a DAQ device include the method by which it is connected to the computer (PCI, PXI, USB, Ethernet, WiFI, ...), the package or enclosure, the type of connections to the system being measured, power, weight, and so on. Your local NI sales representative can be very helpful in making sure you consider all the relevant issues.
This is going to be for an industrial application, and I don't have the budget for a PXI system, so I will probably go with a rack-mount PC with PCI slots. NI still sells enough PCI cards that will do the trick. Thanks very much for your help!