Determine Operating Baseline

March 29, 2018

The quality of your product testing is directly affected by the quality and reliability of your shaker system. A good preventative maintenance (PM) practice is taking baseline operating measurements upon delivery of your shaker system. The same tests can then be run on a routine schedule and results compared to ensure the system is continuing to perform according to its standards and specifications.

A test engineer needs a certain level of comfort in knowing that test results are indicative of vibrations experienced by the product and not the result of a mechanical testing component introducing unwanted vibrations into the test. Any changes in the signal measurements over time can point to a potential problem in the system, such as a failing testing component, which should be investigated and repaired or replaced (if necessary) to maintain the integrity of test results.

There are a several tests which can be utilized to determine the operating baseline of a shaker system, including:

It is important to understand the mechanics of each component in the system and how they affect the overall outcome of the test. For each of the following tests, it is recommended to run baseline operating tests for several system configurations, beginning with bare table. After measurements from the bare table test are recorded, a new test should be run with measurements recorded for each component (ie: head expander, fixture, etc) being affixed to the testing system. By observing the response of each mechanical addition to the system, test engineers can determine how energy is being affected by the addition of each component. When the product is mounted, a direct comparison can be made between the total energy contribution of the testing system components and the energy contribution of the product. The initial recordings from each system configuration should be used as a benchmark and compared against all future recordings. 

When an individual component is adversely affecting the system, a test engineer can easily identify that component by running baseline operating tests in each system configuration, starting with bare table again. The component affecting the system will be easily exposed by the measurements as it is introduced into the configuration.

Test engineers should always ensure they are comparing “apples to apples” by keeping the physical setup of the system for all tests as identical as possible.

Sine Sweep with THD+N   *Recommended*


Figure 1: A shaker manufacturer’s system capability rating

We recommend this method of shaker validation on the basis that many shaker manufacturers rate their systems according to a sinusoidal vibration peak force test, as shown in Figure 1.  Therefore, this test can also be run for shaker acceptance.

Sine Sweep THD-1

Figure 2: Sine sweep from 5Hz to 2000Hz showing the response signal and the drive voltage required to control the test. Record snapshots like this to monitor changes over time.

Another reason for recommending this method is that a sinusoidal test will sweep through a range of frequencies (using tracking filters), accurately informing the test engineer of vibrations across the frequency range of the shaker capabilities (see Figure 2). The Total Harmonic Distortion (THD) will show at all the frequencies how much harmonic content there is in a signal.  The larger the amount of harmonic content in a signal, the larger the % THD value will be.  This would be indicative of noise or undesired vibration in the signal.  As the mechanical components of the shaker degrade over time, more noise will be induced into the system and the % THD value will increase. THD+N will show the harmonic content plus the noise floor of the signal.

  1. Prepare for a bare-table test.
  2. Configure a sine sweep to cover the full frequency range of the shaker. If you are unsure, 5Hz to 2,000Hz is a good place to start. Make sure you are not exceeding the shaker ratings.
  3. Ensure tracking filters are on. This is important to eliminate harmonic data and obtain a more accurate measurement.
  4. Run the test.
  5. Record the control (acceleration vs frequency) signal.
  6. Record the drive output (drive vs frequency) signal.
  7. Record the drive voltage min and max values during the sweep.
  8. Record the THD or THD+N.
  9. Add one mechanical component at a time and repeat steps 1-8. Continue until all mechanical components have been affixed.
  10. Save measurements for comparison against future recordings.

The output drive voltage will show several different variations as the shaker degrades.  In running the same test, you may see an increase in voltage requirement, new resonances being compensated for, or shifts/increases in the existing resonances. All are signs that use of the shaker over time has caused degradation.

Sine Sweep THD-2

Figure 3: Open loop sine sweep from 5 to 2000 Hz. Note the flat drive signal at 50 mV.

The same test can be run open-loop, excluding the controller from the response to verify that the controller is not interfering with the shaker and amplifier measurements. This is accomplished by sending a fixed voltage to the shaker and ignoring the response, as displayed in Figure 3.

Sine Sweep THD-3

Figure 4: Overlay of closed loop and open loop test results. The closed loop drive (bottom-red) is an almost perfect inversion of the open loop response (top-green).

The closed loop and open loop data are essentially the same, as shown by overlaying the control and drive signals from both. The open loop response is almost a perfect inversion of the closed loop drive, as displayed in Figure 4.

The goal of running a random test is essentially the same as a sine sweep – to display the response of the shaker across the operating frequency range.  This is not as accurate as the sine sweep method because it does not display the actual response of the system; it displays an average response of the system.  Instead of G’s vs frequency you get energy density vs frequency.

  1. Prepare for a bare-table test.
  2. Create either a standard random profile, such as the popular NAVMAT profile, or a flat random test (e.g. 0.001 G2/Hz from 10 to 4,000Hz). If there is a standard profile the laboratory runs all the time, it may be desirable to run the baseline measurement with tha standard profile. Otherwise a flat acceleration band across the frequency range is usually a better option.
  3. Run the test.
  4. Record the voltage density (drive output vs frequency).
  5. Record the PSD (response acceleration vs frequency).
  6. Add one mechanical component and repeat steps 1-5. Continue until all mechanical components have been affixed.
  7. Save measurements for comparison against future recordings.

Held Sine Tones
This is a simple baseline which provides shaker response data only for individual frequencies. This test can be run for several frequencies with each system configuration.

  1. Prepare for a bare-table test.
  2. Run a sine tone on the shaker. For an Electrodynamic shaker this could start in the 30Hz range, as it is usually outside of the resonance of the system and a low enough frequency to produce a measurable displacement. For a Hydraulic shaker this might start in the 2Hz to 10Hz range.
  3. Note the amplitude and gain settings.
  4. Run the test.
  5. Record the input signal.
  6. Verify the displacement setting with a displacement wedge or other tool. We typically use a 0.1″, 30Hz Sine Dwell in conjunction with a displacement wedge on an Electrodynamic shaker. As the frequency increases, the displacement will become more difficult to physically measure, so it is impractical to verify displacement at every frequency.
  7. Repeat steps 1-5 for various frequencies at 1/3 octave steps to obtain several data points across the operating frequency range of the shaker system.
  8. Add one mechanical component and repeat steps 1-7. Continue  until all mechanical components have been affixed.
  9. Save measurements for comparison against future recordings.

Signal Generator and Oscilloscope or Spectrum (Signal) Analyzer
This is similar to the Sine Held Tones method discussed above, except independent equipment replaces the controller. This test can be run for several frequencies with each system configuration.

  1. Connect a signal generator to the amplifier.
  2. Connect the (response) accelerometer to a signal analyzer or oscilloscope.
  3. Follow steps 1 through 9 of the Held Sine Tones method above.