FRF not averaging TF correctly

[mhughes]

I am trying to use LV's built in FRF to perform transfer function averaging.  However it does not appear to average multiple sets of data correctly.

In LV 7.0 I wrote a transfer function sub VI, which calculates the TF for only 1 set of data at a time and does not average.  I am trying to implement the FRF function to do the required TF averaging. 

I can read in the first set of data to the TF sub VI that I wrote, as well as the FRF. Both output the same TF magnitude value, 48.236 (which is expected since its only 1 set of data).  When I read in the next set of data, the TF magnitude for that set is 47.2095.  Thus, I am expecting a average TF magnitude of (48.236+47.2095)/2=47.2228. However, the FRF tells me the average TF magnitude is 26.7809.

My averaging parameters cluster is as follows:
Vector Averaging
Linear
10
One Shot

I have one theory, but I cannot test it.  My code was originally written in LV 7.0. I recently upgrade to LV 7.1.  However, so that I was able run the VIs I wrote in LV 7.0, I needed to copy parts of my old LV 7.0 fdirectory to my LV 7.1 directory; and for some reason I can not the run the FRF which comes with LV 7.1 (I get "unable to execute" errors), so therefore I am using the FRF function which I had bundled into my library of VIs which I wrote in LV 7.0.  Could using the LV 7.0 FRF in LV 7.1 cause this averaging error?

[JLS]

Hello,

I don't think the problem is due to your VI version upgrade from 7.0 to 7.1.  I do not know what FRF refers to, but if you'd like to average sets of data, we can implement that very easily (use the sum array values function in the numeric palette to get the SUM; use the array size function from the array palette to get the SIZE, and SUM/SIZE will be the average).  If you think you have found a bug, however, with an existing function, please let me know which function, or better yet, attach a simple program which reproduces the problem.  In either case, please let me know what FRF stands for, and I'll try to help you with whatever is going on.

Thank you,

JLS