I have a question that is somewhat related to this previous post:
The difference - the data I have is a ±5V sine wave, with the frequency of the wave relative to my shaker's sweep frequency(COLA signal).
I am attempting perform an order analysis via post-processing. I have the COLA signal in the Tacho group, and I used the Time Signal Calculator for TACHO_PULSE_TO_RPM and got a relatively smooth curve that accurately depicted my shaker frequency sweep (15Hz to 2kHz).
The problem occurs when attempting to perform the actual order analysis. I'm getting an error from the software:
But, the operation isn't completed successfully. I'm sure it has something to do with the Raw Tacho signal. I typically use a TTL signal for tacho pulses and order analysis, but the COLA signal doesn't give me the option of a square wave.
I could try modifying the COLA signal to a DC Voltage that is relative to the frequency, and follow the advice in the link I copied above?
I have also attempted to record the COLA signal directly into the Tacho Channel group, and the channel fails to even see the ±5V signal for some reason. It doesn't give me the options to set the crossing level and features available in the time-signal calculator, so on the Tracking setup tab it's showing a bunch of low-level noise. I was hoping to free up a SCADAS channel for other purposes, and perhaps rerun the sweep to perform the Order Analysis simultaneous to the data acquisition.
All of the above operations are to convert the peaks, along a frequency sweep, to a 2D plot (and make this easier for management to understand). So, I'm trying to convert the peaks of the first order data in this waterfall plot to a 2D Front-Back plot, and since I didn't have order analysis active, I can't add an offset order cursor.
Any help to solve that problem would be helpful.
Solved! Go to Solution.
I got it!
I stuck with using the raw COLA signal, and using the TACHO_PULSE_TO_RPM in the Signal Calculator, to calculate the shaker frequency (unfortunately in RPM instead of Hz), between 900-120kRPM, representing 15-2000 Hz. I did this in the Offline Sine Data Reduction application, instead of directly in Test.Lab.
I then copied the accelerometer response channels I wanted to see fundamental order curves, the raw COLA signal, and the calculated RPM signal, and I was successful at performing the Order Analysis in Time Data Post-Processing.
Here is an example of the 2D plot I was after - from the initial shaker sweep/3D waterfall data:
Now, I'll see if I can divide the RPM channel by 60, to return to RPS, and I'll have exactly the plot I need.
UPDATE: Added a Time-Signal Calc tacho channel with RPM/60, and I can simply right-click on the x-axis and modify the units to Hz. It gives me the option to plot against multiple calculated Tacho channels - nice!