shift register problem

[bunnykins]

Dear LabVIEW forums,

I am having difficulties programming a structure that appears to work in theory on paper.  The structure or procedure is based on finding the cut-off frequency (unknown but I can tell you its approx.80Hz) of a low-pass filter (which is embedded in a sub-vi).  The sub-vi filter has two inputs (input voltage amplitude and frequency) and one output (voltage amplitude). 

My idea and structure is to use a half-split method to search for the cut-off frequency based on starting from 10MHz frequency (initial value) then... progressively halving the frequency until the amplitude ratio (output/input) > 0.7071 (i.e. the -3dB cut-off point) then...

to adjust the new test frequency to be 1.5 x the frequency value when amplitude ratio (output/input) > 0.7071 (i.e. bringing the frequency value above but closer by half of the previous frequency value that was above cut-off frequency).

Then re-test and adjust accordingly in same manner until the test-frequency is within 1Hz of the cut-off frequency based on amplitude ratio (output/input) = 0.7071

The attached vi appears to partially work whereby the initial half-split run down to 76Hz works however it unexpectedly oversteps down to 38Hz (it should have gone back up to 114Hz for the re-try frequency (i.e. 1.5 x 76Hz = 114Hz).

I have tried various combinations of program however I'm just going round and round in circles now.

Any help and assistance would be greatly appreciated.

Regards,

[BretD]

you are wiring the wrong value from your case structure to the frequency input of your sub vi. Wire the divided value not the incoming value to the frequency input and it will work fine. I tried it both ways and your sub VI still gave a value of 81 point something Hz, which should be about the right value.

[BretD · Accepted solution]

I also changed a few things around so that it gives a better approximation, like making the multiplied value from 1.5 to 1.1 and lowering the initial frequency count so that the result is more accurate. If this isn't sufficient just let me know 😃

[bunnykins]

Many Thanks BretD,

I appreciate your help and assistance.  I am new to LabVIEW and tend to get very frustrated with it at times.

I selected the half-split method since I thought this would be quicker and more efficient than say a ramp method from 0Hz upwards in 1Hz steps.

The initial frequency count will have to be set at 10000000 (10MHz) since later I have to figure out a way of scanning through various similar Sub-vi's that have "unknown" cut-off frequencies withing the range up to 10MHz reporting the results automatically (sort of a batch operation).

Again many thanks for your help.

Regards,