VehicleTechnology.org  
sae-uk.org
 
     
<-Back   Next ->
 
 
INDEX
_
_
_
-
 
 

pick-up to receive the reflected waves. Two reflected pulses are received in quick succession. The first comes from the reference surface as this travels the shorter distance. The second comes from the measurement surface. The more viscous the oil, the more energy is absorbed, making this reflected pulse smaller in amplitude. The ratio of the pulse amplitudes therefore provides a measure of the oil's viscosity.

Interface
The original hardware was replaced with a PC-based oscilloscope/signal generator, which is connected via USB.  The unit was supplied with a DLL file to allow communications with application software. In the LabVIEW application, this was embedded in a “Call Library Function” node wherever there was a need to send or receive data, either to configure the unit or read the sampled values. Figure 2 shows the GUI. The Waveform Graph shows the reflected reference and viscosity pulses. The vertical-line cursors define the measurement windows within the pulse, the horizontal lines giving the average value within those windows. The user can control these positions using the slider controls under the Graph to best match the measurement window to the pulse shape.

The controls in the top-left corner control the operating  frequency. By adjusting the frequency manually, the user can identify the range of frequency of interest for more detailed analysis. This is also useful for setting measurement window boundaries to appropriate positions such that the correct parts of the waveform are measured across the frequencies of interest. The user can then specify the start, end, step size and number of readings to average for an automatic sweep of frequencies that are logged to a results file (CSV format) specified at the bottom left of the screen. This file contains time and date information, allowing data to be tied to a particular test-run of the sensor. This can be important, as the sensors exhibit an initial settling behaviour where their response changes over a number of thermal cycles. The file can then be loaded into a spreadsheet program for further manipulation and to produce graphs. The “WRITE TO FILE” button allows individual manual readings to be captured to the results file.

The calculated ratio, which decreases with increasing viscosity, is displayed on the left-hand side of the GUI. The raw value is displayed, along with a version that is smoothed using a Moving Average filter. Below this is the oil temperature, measured using a thermocouple. This is logged alongside the viscosity data, and a temperature chart on a separate Tab Control page shows a historical display of the temperature. This serves as a utility when stepping the operating temperature to show when the oil temperature has stabilised at the next setpoint.

 


Implementation

The range of controls and displays that are available for creating the ideal front panel are matched by the  functions used to process the data. This application uses various inbuilt functions that were easily configured to extract the required data samples for processing. The use of Formula Nodes  then allowed complex algorithms to be entered directly for those parts where wiring a block-diagram would have been less practical.

Advantages of the automated system
 The advantages of the system over the former manual implementation are:

  • A direct measurement of signal amplitude is obtained, without the potential inaccuracies of human judgement in obtaining a reading from an oscilloscope.
  • Multiple readings at each measurement frequency can be averaged quickly.
  • It is feasible to make many more readings by reducing the frequency step size between measurements.
  • There is no risk of human error in entering the data into the spreadsheet.
  • The ratio value is calculated and logged directly, along with a timestamp and other diagnostic information about system configuration.
  • The windows on the waveform where data is processed can be moved dynamically. Each measurement run is reduced from several hours to a few seconds, making the easurement less prone to temperature variations during the run.
  • It is possible to refine the system towards an auto-tuned version and visualise the data thereby assisting the development process towards an auto-tuning version that would be required in production. E.g. setting frequency or window position based on some other input, such as temperature.

Summary
The automated system provides a marked improvement in speed and accuracy compared to the manual system it
replaced. It was possible to obtain a greater volume of higher quality data to better understand the sensor's behaviour over the required frequency and temperature ranges thereby assisting in the development of self-tuning algorithms for production systems.

 
 
 
 
 
Site Designed by: Versatile Solutions