Showing results for 
Search instead for 
Did you mean: 

LMS Test.Lab: Remove mean/average rpm from torsional data?


I have torsional rpm data (taken with a high pulse per revolution) that I would like to remove the mean/average rpm from in the time domain data.


Is there a good way to do this in LMS Test.Lab?  I tried doing a high pass filter, but did not like the results.


In the picture attached, I would like to center the rpms around 0 rpm, or close to it.








Re: LMS Test.Lab: Remove mean/average rpm from torsional data?

Siemens Legend Siemens Legend
Siemens Legend

Rather than use a filter, you might want to consider (using a 120 pulse/rev example):


  1. Calculate the torsional data from the original pulsetrain with the high pulse per revolution (say 120 pulse/rev) -> Produces data with torsional fluctuations and a mean rpm
  2. Calculate the torsional data again from the original pulsetrain, but with 1 pulse per revolution (skip 119 pulse/rev) -> Produces a mean/average rpm trace without torsional fluctuations
  3. Subtract the resulting time traces of step 1 and 2 -> Leaves the torsional fluctuations in time domain with no mean/average rpm (in example below, would subtract the green trace (with pulses skipped) from the blue trace (used all pulses)rpm120v1.png



Here are the steps in Time Signal Calculator:


1. First calculate the torsional data with a high pulse per revolution (for example, use “pulses_per_rev” of 120, and “pulses_to_skip” of 0), using either Time Signal Calculator command below:


  • TACHO_PULSE_TO_RPM (from data channel containing pulse train)
  • TACHO_PULSE_TO_MOMENT (from tacho channel data)

2. Next calculate the mean/average rpm with “pulses_per_rev” of 1, and “pulses_to_skip” of 119.




3. Subtract the two traces from step 1 and 2.


The Time Signal Calculator menu could look like this:






Final results should look like below: