Q & A: Calibration Verification

May 30, 2019

Verifying Controller When Accelerometers Are Already Calibrated

It is critical to properly verify the calibration of all the measurement components of a vibration testing and measurement system. Calibration verification ensures that the hardware is accurately measuring the signals provided.

Both the VR9500 and ObserVR1000 combine analog and digital circuitry to accurately and reliably measure the electrical signals generated by various types of sensors. However, all electronic systems are subject to some form of fatigue and drift due to normal use, and VR systems are no exception. While we use only the highest quality components and implement a rigorous testing system to minimize this drift, it is impossible to eliminate.

An extreme example would be hardware experiencing a 5% drift. For an accelerometer calibrated perfectly at 10mV/G, that 5% drift would have the same effect as changing the sensitivity by 5%, meaning the measurement is now 10.5mV/G. This may seem relatively small, but for a 50mV signal, the correct measurement is 5G and the post-drift measurement is 4.8G, a significant error. As the amplitude increases, this difference continues to grow. At full scale, a 10mV/G accelerometer can measure approximately 500G of acceleration. A 5% drift at this level results in 476.19G measurement, almost a 24G error.

The same type of expanding measurement error behavior can occur with the VR9500 or the ObserVR1000. It is just as critical to verify this measurement hardware as it is to verify the calibration of an accelerometer. If you don’t verify the measurement of the controller, how do you really know if you are running the test you think you are?

 

Does Calibration Verification Run Only in Sine Test Mode? What about Random?

The calibration verification of a hardware unit ties the output directly to all the inputs with a simple BNC cable. The hardware and cabling create a very linear system, so there is no variance in loss or damping across the frequency ranges tested and measured. The system is also time-invariant since there is no shift in period or duration between the output signal generated and the input signal measured.

The superposition principle states that, for all linear systems, if signal A generates a measurement X, and signal B generates a measurement Y, then signal A+B = measurement X+Y.  Extending this to waveforms, when a Fast Fourier Transform (FFT) converts a time-domain signal into the frequency domain, it breaks the signal into a series of narrow bandwidth tones, spread evenly across the desired frequency. These narrow bandwidth tones are, essentially, sine tones with varying amplitudes being measured. Therefore, in the linear calibration verification system, a random signal is simply an additive set of sine tone signals. Verifying the calibration using a set of sine tones is the same as verifying the calibration using a random signal.

Even more significantly, the RMS calculations for DMM measurements of a random waveform will not match the RMS calculations for the VR9500 measurement of a random waveform. That is because the DMM measures across a frequency range that will not exactly match the frequency range of a random test driven by a VR9500. This issue does not exist for sine waveforms, making Sine test mode the obvious choice for calibration verification.

 

What Options Are Available For Calibration and Calibration Verification of Vibration Research Hardware Units?

Vibration Research Headquarters

Calibration, which is adjustments to measured voltage readings, of Vibration Research hardware can only be done at VR headquarters or at an approved sales and support office.

There are several options for the Calibration Verification of Vibration Research hardware:

The Vibration Research lab offers two levels of calibration verification:
1. An ISO/IEC:17025-2005 accredited calibration: the laboratory is accredited by A2LA to meet the requirements of ISO/IEC:17025-2017. Calibration certificates and reports are provided with the stated measurement uncertainty for each measurement taken.
2. A NIST traceable calibration certificate: no measurement uncertainty is included in the final report
Vibration Research offers software for automated Calibration Verification:
This method requires either a permanent or annual calibration verification license
The software is already activated if the hardware is part of a current Upgrades & Support Agreement
Additional equipment is required: see the Procedure lesson for details
Perform the calibration verification with an in-house, customized procedure:
This method is acceptable: Vibration Research recommends modeling any custom procedure after the manufacturer recommendations.

 

If calibration verification shows that a hardware unit is no longer generating in-tolerance measured voltages, then that unit will need to return to Vibration Research headquarters for calibration adjustment and any necessary repair. Please contact vrsales@vibrationresearch.com for more information.