Mike Hertz, senior field applications engineer at Teledyne LeCroy, discusses a new test methodology for accurately characterizing automotive axle torque, which combines dynamic parametric measurements with numerical script processing
By Mike Hertz, senior field applications engineer, Teledyne LeCroy
Torque sensors consisting of strain gauges are often used in mechanical engineering research and development to convert mechanical load into electrical signals which can be measured and quantified. Accurate gauging presents measurement challenges to electronic instruments which interpret the electrical output of these mechanical devices.
Figure 1: Block diagram of hardware configuration for torque measurement.
A typical system for torque conversion (Figure 1, above) consists of a half-shaft from an automobile front axle. The inboard joint side of the constant velocity shaft attaches to the transaxle assembly, and the opposite (outboard) side connects to the wheel. Signal electronics develop an mV-per-unit-force output from the strain gauges, which are combined and converted into a digital signal which is transmitted to a digital telemetry receiver (or demodulator). The receiver box generates an analog output proportional to torque from the shaft, which is typically connected to a measurement instrument such as an oscilloscope for measurement display and interpretation.
Measurement process flow and special operators
Electronic measurement devices such as oscilloscopes conventionally receive input voltage signals and report results in various units (such as torque) by employing rescale math operators and unit conversions. Reporting numerical results is common, but persistent graphical results of intermediary mechanical benchmarks are typically not available. In order to provide more advanced mechanical information, a new methodology has been discovered through the incorporation of dynamic in-line measurements.
Figure 2: Simplified oscilloscope internal block diagram shows processing path of signal.
Before exploring the use of dynamic in-line mechanical measurements, an illustration of the traditional oscilloscope measurement process flow is depicted in Figure 1. The analog waveform (torque sensor output) enters the acquisition system via an amplifier stage, and is subsequently digitized and streamed into acquisition memory. The processing functional block shown in the diagram operates on stored acquisition memory and calculates automatic measurement parameters along with its other analysis. In the case of stock measurement parameters such as rise time or pulse width, the processing functional block operates directly on the stored waveform with measurement results fully available within the application, enabling further analysis and the accumulation of measurement statistics to be performed with subsequent acquisitions.
While existing oscilloscope measurement parameters and mathematical operators provide an array of analysis options, when the need arises for performing complex calculations outside the sphere of available oscilloscope measurements selections, operators have traditionally opted for a measurement technique in which data acquired by the scope application is transferred to a separate application for analysis to be performed. Recent advancements now allow for execution of a user-defined algorithm directly within the measurement application. The implementation of a dynamic torque measurement enables computational results from a script to be executed inside of the measurement system in real time, with waveform data sample points acquired by the oscilloscope becoming immediately available to the script. This method forms a dynamic in-line collaborative measurement process.
Figure 3: VBS code sample for conversion.
Figure 3 (above) shows sample source code for implementing a torque meter implemented on an oscilloscope with an in-line Visual Basic (VB) script. The input source to the script is the electrical output from the receiver, and sparsing and filtering is performed on the input to denoise and linearize the sensor value. Next, roof and floor operators are used to retain both the positive and negative extremes applied to the half shaft. Conversion factors are applied to scale the electrical readings into in-lb of torque, and both maxima are dynamically relabeled as successive torque values are obtained. The larger magnitude of the two maxima is reported as the P5 measurement parameter numerical value, and parameter aliasing is applied to report the result as a torque meter value.
Function F1 in Figure 4 (below) is a zoom of the channel input, while functions F3 and F2 are the roof and floor of the input source. The timebase descriptor indicates a time per division setting of 20 microseconds per division, and the roof and floor functions have each retained the maxima of positive and negative torque respectively for the 8,801 sweep values occurring at the time of image capture. Numerical readings in in-lb are pinned to the roof and floor operators by the in-line script, and the values of the two maxima are continually updated by each subsequent acquisition. Parameter P5 displays the torque meter calculation with its parameter alias applied. By incorporating a novel use of dynamic in-line parametric measurements into this electro-mechanical system, a new way to quantify automotive torque can be realized.
Figure 4: Dynamic in-line realization of torque meter calculation
May 9, 2017
