THE VR BLOG

Over the years we have helped many customers with their challenges and difficulties with vibration tests. Some problems are more complicated than others. Nevertheless, there are some common problems that are found in vibration testing and their solutions are relatively simple.

NOISE FLOOR

Problem

In our day-to-day lives, we recognize there is a lot of background audio noise, from the hum of a fan in the computer to the conversations of others in an adjoining room. All the various background sounds we generically call "white noise".   

That same background noise is found in digital circuits and is known as “noise floor”. Figure 1 illustrates the background noise floor for a particular vibration test.

The following are items that contribute to the noise floor: 

• Ground loops/60 cycle noise
• External EMF noise
• Circuitry Noise (shot noise of P-N junctions; thermal noise)

Figure 1: Noise floor across a spectrum from 1 kHz to 20 kHz. Notice a few resonances that stand out above the noise floor. The resonances near 5 kHz is easily identifiable, whereas the resonance near 15 kHz just stands above the noise floor.

Solution

Complete removal of noise is impossible but noise can be reduced. A few relatively simple solutions to minimize the noise (and thereby reduce the noise floor) are as follows:

• Remove ground loops
• Use high quality components in your circuit boards

ACCELEROMETER SENSITIVITY ISSUES

Problem

There are many problems that can arise related to transducers. Sometimes a test will not work or poor results are obtained because of an issue related to the transducer. Some of those issues are:

1. Incorrect Sensitivities

   • When setting up a test it is necessary to enter sensitivity values for the transducer. Suppose an accelerometer is used that is calibrated to 10mv/G but the test engineer mistakenly enters the sensitivity as 100mv/G. In this case if the accelerometer detects a 10mV value it will display a 0.1G acceleration value when it should display a 1G acceleration. The results will be off by a factor of 10. This, of course, is a drastic difference but it is possible that a value is entered that is off by just a little bit from the certified value. This can lead to errors, howbeit, small ones.

2. No Power to accelerometer

   • Many accelerometers require external conditioning or power from an external source. If a test engineer forgets to give the accelerometer power, then there will be no reading made by the accelerometer. There is an easy fix for this but it is a rather common error. It is akin to turning on your computer but neglecting to plug it in. When the computer doesn’t do anything, it is always wise to check the power cord. It is uncanny how often a lack of power is the cause of many troubles with electronic devices.

3. Improper accelerometer for test

   • One other common accelerometer-related problem is that an improper accelerometer is used for a test. Suppose the controller has a 5V max input. If it uses a 100mV/G accelerometer, then the accelerometer will saturate with a high Q resonance that measures greater than 50G – being unable to collect the data above 50G. 

 Solution

     The solutions to these accelerometer sensitivity issues are fairly simple:

1. Use TEDS (Transducer Electronic Data Sheets)

   • TEDS are stored in the memory of the IEPE accelerometer. The calibration data and characteristics of the accelerometer, such as its' calibrated sensitivity, calibration date, serial number, manufacturer, and calibrated frequency range, are all stored in the memory of the IEPE accelerometer. This information can easily be imported into the controller software so that engineers will not be able to accidentally enter incorrect sensitivity values.

2. Select a different sensitivity accelerometer 

   • Engineers must wisely select accelerometers based on the circumstances of the test, seeking to ensure that the accelerometer will not be saturated. 

      

CABLE ISSUES

Problem

Vibration test can develop issues and poor results due to problems with the cables.

1. Cable Whip

   • When a cable is not properly secured it can vibrate wildly due to the vibrations of the test. This whipping action can introduce additional noise-like data into the results (triboelectric effect).

2. Cable not connected securely to accelerometer

   • A number of years ago, I ran into some results from a test that didn’t make sense. After examining various possibilities, it became apparent that there was a problem with the cable connection to the accelerometer. This cable connection issue led to discontinuities in the data.

Solution

     The solution to these cabling issues are fairly simple:

1. Secure Cables

   • Cables should be secured so they cannot whip around during a vibration test. Electrical tape is commonly utilized to secure the cable in position during a test.

2. Properly connect cables

   • Cables must be tightly fastened to the accelerometer. Care must be exercised to ensure that the cable is fastened to the accelerometer so that the cable pin is not bent or damaged.

ACCELEROMETER MOUNTING ISSUES

Problem

Sometimes problems arise in vibration testing that is related to the fact that the accelerometer or transducer is not properly mounted to the DUT (Device Under Test). If the accelerometer is not properly mounted, it is impossible to obtain valid data. Two common sources of improper mounting are:

1. Mounted in wrong direction

   • On the surface, it doesn’t seem likely that one would mount an accelerometer in the wrong orientation but it happens more often than you might expect. A single-axis accelerometer may measure accelerometers in the vertical plane. However, an engineer may be using a shaker that vibrates in the horizontal plane (slides back and forth). More often than one might expect, the accelerometer in such a situation is mounted in a way as if it were measuring vertical motion. Consequently, the data obtained (if any) will not make sense or be valuable.

2. Not securely mounted

   • Another easily fixable problem is an improper method of mounting the accelerometer. The best method of mounting the accelerometer is to use the stud mount method. Sometimes the accelerometer is tightly mounted. Other times the stud mount method is not available and an adhesive is used. A cyanoacrylate glue would be the next best method, but sometimes engineers settle for a wax adhesive because of its convenience. Using a wax adhesive can lead to poor contact between the surfaces, especially under certain temperature situations, and consequently cause poor results.

      

Solution

The solutions to these mounting issues are fairly simple:

1.      Confirm the accelerometer is orientated correctly

   • Engineers should confirm the direction in which the accelerometer measured the acceleration and be sure to mount the accelerometer in the correct orientation.

2. Mount with the best possible mounting method

   • Engineers need to be aware of the best mounting methods for an accelerometer. They should not use the method that is most convenient for them if it is not also a method that produces satisfactory results.

AMPLIFIER ISSUES

Problem

Another common area of errors in vibration testing are those associated with the amplifier.

1.      Gain Cranked Up

   • A very dangerous situation occurs when the amplifier gain is cranked up and the test engineer adjusts the equipment. Test engineers should always get in the habit of TURNING DOWN THE GAIN on the amplifier in-between tests. If the gain is turned down, then even if the current rises rapidly to the amplifier/shaker, no significant current can be sent to the shaker. When the gain is turned down, a damaging blow to the shaker is prevented.

2. Gain Off or Dialed Down

   • Many times, in my early years of testing I would have the entire test set-up and ready to run. When I would initiate the running of the test, the test would not begin and I would receive some error message. Many of those times it was because I failed to turn on the amplifier and turned up the gain.

Solution

The solutions to these amplifier issues are fairly simple:

1. Turn down the gain

   • Whenever a test is not in progress, and especially when an engineer is adjusting the test set-up and cabling, the amplifier gain should be turned down. This will prevent any damage to the shaker system. 

2. Turn on the gain and turn it up

   • Whenever a test is ready to go, and no more adjustments will be made, it is then necessary to turn on the gain and turn it up.