Creating FRF from a sine sweep, without generating the signal

[FNG]

Hi, my problem is that I am trying to create a FRF of a vibrating structure, however the equipment we are using to excite the structure will not allow us to input a generated sine signal from a national instrument board, instead another computer will generate the sine sweep input. Thus I am going to try and recreate the sine input signal (for the FRF) from the output of the shaker table control accelerometer.

We currently have the sound and vibration toolkit and a PCI-4472, I have had a good look at the sine sweep tools contained within the toolkit, and I can see that it would be very complicated to modify these to do what I want.

I'm hoping someone can either tell me a simple way to achieve what I need or at least point me in the right direction.

Thanks in advance.

[TimM(AE)]

Hi FNG

What are you actually trying to do with the NI board ?

I have found these links

This is a kb for Audio Measurements 

http://zone.ni.com/devzone/conceptd.nsf/webmain/CE46CCEDEBE745FA86256CC3007F9C06 

This is a link to another forum which refers to FRF

https://forums.ni.com/t5/LabVIEW/Curvefitting-after-FRF-function/m-p/277279?requireLogin=False

If I am comply of the point please let me know

Hope they help

Tim

NI UK AE

[FNG]

In answer to the first question, (what am i trying to do), I'm taking measurement from four channels, the first channel is the input into the DUT, the last three channels are the outputs from the DUT. I would like to know how the amplitude, frequency and phase of the input signal affects the amplitude, frequency and phase of the output signals. Thus I require a graph of amplitude against frequency (which is built up slowly over time as the input signal sweeps through all the frequencies from 20-2000). I also require a graph of phase against frequency similar to above.

However, the difficulty is that I would eventually like to remove the influence of the input frequency, so that I can have graphs of the output response with respect to a unit input, I don't think it is as simple as just dividing by the input.

The first link was of some use, the second didn't really help.

thanks,

[TimM(AE)]

Hi FNG

I am still not to sure what you are doing but if you are trying to find out what the input signal to the shaker table is then you should be able to link the 4472 straight to the accelerometer.

I am guessing you are using LabVIEW as you have the sound and vibration toolkit. If you are there are lots of example bit of code that you can find under Help >>example finder that will help you with sound and vibration analysis.

Tim

NI UK AE

[johnsold]

FNG,

Can you measure the input signal with another channel of the same device which is measuring the output signals? If you can do this, then you have the actual input signal, sampled at the same rate as the output, which should simplify your later analysis.

Lynn

[TimM(AE)]

Hi

Yeah you should be able to, its is just like measuring any other input signal and as all the DSA boards have simultaneous sampling all for will be sampled at the same rate.

There shouldn’t be any problem

Tim

[FNG]

Yeah I have considered measuring the actual output signal of the other control device, but this isn't much different to measuring the predominant frequency of an accelerometer attached to the shaker table. Also, using the output from the accelerometer removes the need for any scaling (the output from the control device has to go through a big amp).

I'm currently making some progress by using the peak hold function of the spectral analysis express vi. although this is very rough and ready it may just work, If anyone is intersted I'll post the finished code: although this may take a long time as I'm still learning labview.

Message Edited by FNG on 01-17-2006 12:21 PM