Project Description


For over 40 years, the original Shock Pulse Method (SPM) has been very successfully used to obtain a fast, easy and reliable diagnosis of the operating condition of rolling element bearings.

Throughout their lifetime, bearings generate shocks in the interface between the loaded rolling element and the raceway. These shocks ‘ring’ the SPM transducer which outputs electric pulses proportional to the shock magnitude. Unlike vibration transducers, the shock pulse transducer responds at its carefully tuned resonance frequency of about 32 kHz, which allows a calibrated measurement of the shock pulse amplitudes.

Measuring dBm/dBc

Shock pulses from the bearing is sent through the material and picked up by the transducer. The transducer converts the shocks to electric signals which are processed to give a carpet value and a peak value.

The shock pulse meter counts the rate of occurrence (incoming shock pulses per second) and varies the measuring threshold until two amplitude levels are determined:

  • the shock carpet level (approx. 200 incoming shocks per second. This level is displayed as dBc (decibel carpet value).
  • the maximum level (highest incoming shock under 2 seconds). This level is displayed as dBm (decibel maximum value). Using a blinking indicator or earphones, the operator can establish a peak value by increasing the measuring threshold until no signal is registered. Because of the very large dynamic range, shock pulses are measured on a decibel scale (1000 x increase between 0 and 60 dB).

Shock pulse amplitude is due to three basic factors:

  • Rolling velocity (bearing size and rpm)
  • Oil film thickness (separation between the metal surfaces in the rolling interface). The oil film depends on lubricant supply and also on alignment and pre-load.
  • The mechanical state of the bearing surfaces (roughness, stress, damage, loose metal particle).

Input data

The effect of rolling velocity on the signal is neutralized by giving rpm and shaft diameter as input data, with ‘reasonable accuracy’. This sets an initial value (dBi), the start of the ´normalized’ condition scale.


The initial value and the range of the three condition zones (green – yellow – red) was empirically established by testing bearings under variable operating conditions. The maximum value places the bearing into the condition zone. The height of the carpet value and delta (dBm minus dBc) indicated lubrication quality or problems with bearing installation and alignment.

Analyzing dBm/dBc (SPM Spectrum™)

By analyzing the shock pulse measurment in a frequency domain, the source of the shock pulses can be determined.

The purpose of ‘SPM Spectrum’ is to verify the source of high shock pulse readings. Shocks generated by damaged bearings will typically have an occurrence pattern matching the ball pass frequency over the rotating race. Shocks from e. g. damaged gears have different patterns, while random shocks from disturbance sources have none.

Signal and measurement

The result from the dBm/dBc measurment is the bearing condition data, evaluated green – yellow – red. A second measurement produces a time record that is subjected to a Fast Fourier Transform (FFT). The resulting spectrum is used mostly for pattern recognition. Spectrum line amplitudes are influenced by too many factors to be reliable condition indicators, so all condition evaluation is based on the dBm or the HR values.

One unit for amplitude in an SPM spectrum is SD (Shock Distribution unit), where each spectrum is scaled so that the total RMS value of all spectrum lines = 100 SD = the RMS value of the time record. The alternative is SL (Shock Level unit), the RMS value of the frequency component in decibel. Alarm levels are manually set for each symptom to show evaluated results in green – yellow – red. Various types of spectra can be produced. The recommended setting is a spectrum with a resolution of at least 0.25 Hz, e. g. 3200 lines over 500 Hz, saving peaks only.

Input data

Pattern recognition demands precise data on the bearing and exact measurement of the rpm. The rpm should be measured, not preset. The factors defining the bearing frequencies are obtained from the bearing catalogue in Condmaster by stating the ISO bearing number.


The frequency patterns of bearings are preset in Condmaster. Linking the symptom group ‘Bearing’ to the measuring point allows the user to highlight a bearing pattern by clicking on its name. Other symptoms can be added when appropriate, e. g. for gear mesh patterns. Finding a clear match of a bearing symptom in the spectrum is proof that the measured signal originates from the bearing.