Non-Contact Sensor Performance Report Abstract The 30mm non-contact sensor (Encoder) was subjected to a variety of tests outside of the recommended usage parameters. The separation distance, planar tilt, and axial misalignment were all varied in turn to gauge the sensor s response to the application magnet. The output of the sensor was then compared to the response of the reference potentiometer. It was found that even in usage conditions that would damage or destroy mechanically coupled sensors, such as 0.5000 inch separation distances or 45 planar tilt, the non-contact sensor would continue to function within a 10 percent margin of error. It was also found that the non-linear sensor response tended to be highly regular and repeatable, thus making these non-linearities easier to compensate for. Based on the results of the acceleration tests it can be concluded that the sensor does not have any appreciable lag when used under standard operating conditions. Additionally, it was found that even at extreme distances response lag was kept below 8msec. Lastly, by plotting the sensor s motion versus the potentiometer s motion hysteresis curves were generated. From this series of tests it was determined that the 30mm sensor does not exhibit any hysteretic behavior under normal operating conditions. MBerggruen 7/18/2012
PAGE 1 of 19 Non-Contact Sensor Performance Report Overview The 30mm non-contact sensor was benchmarked in three out-of-specification scenarios in order to gauge its performance outside of the provided recommended usage envelope. The face-toface separation distance, planar tilt, and axial misalignment parameters were investigated. Additionally, the 30mm non-contact sensor was actuated with a switched motor in order to gauge the sensor s response to rapid accelerations. The non-contact sensor was subjected to hysteresis testing in order to parameterize any hysteretic behavior. These tests were conducted over two ranges of motion. Methodology The general testing procedure consisted of driving a reference potentiometer at a constant speed. The potentiometer was fitted with the standard 12mm ring application magnet. This magnet was used to engage the 30mm non-contact sensor. Data was collected using a PC oscilloscope. The standard testing conditions used are shown at right; these parameters were used whenever appropriate. Separation Distance Standard Testing Conditions Mechanical Application Magnet 12 mm ring Rotation Speed 30 rpm Separation Distance 0.0625 inch Sensor Orientation Horizontal Planar Angle 0 PicoScope Configuration Channel A (Blue) Channel B (Red) Input Range Time per Division Potentiometer Sensor 5 Volts 500 ms/division Face-to-face separation distance was measured from the surface of the application magnet to the sensor face using a graduated platform. Separation distance was increased from 0.0000in to 0.5000in in increments of 0.0625in. For reference, with the standard application magnet, it is recommended to maintain a separation distance between 0.0100in and 0.5000in with values closer to 0.1875in being better. Planar Tilt Planar tilt was measured by adjusting the sensor mount using the graduated platform. The sensor and application magnet were then adjusted to be as close as possible. Planar tilt was increased from 0 to 45 in increments of 5. For reference, with the standard application magnet it is recommended that the planar tilt be kept to 30 or less. Axial Misalignment For the axial misalignment trials spacers were used to offset the axis of the sensor from that of the application magnet. Axial misalignment was increased from 0.0000in to 0.2500in in increments of 0.0625in. For reference, under normal operating conditions it is recommended that the axial misalignment be kept between 0.0000in and 0.1000in.
PAGE 2 of 19 Test set-up used for separation distance (measurement in red) Test set-up used for planar tilt (measurement in red) Test set-up used for axial misalignment (measurement in red) Acceleration The test set-up consisted of a motor driven reference potentiometer mounted opposite the noncontact sensor. The potentiometer was fitted with the standard 12mm ring application magnet. This magnet was used to engage the 30mm non-contact sensor. The motor would then be rapidly energized using a benchtop power supply. The separation distance was increased from 0.0000in to 0.2500in in increments of 0.0625in. Hysteresis The standard procedure for this test was to manually actuate the potentiometer through a given angular range and then return back through the same range in the opposite direction. Prior to each trial the sensor-potentiometer azimuth was adjusted to minimize the separation of the two signals. Two hysteresis tests were conducted, the first manipulated the sensor-potentiometer system over a small angular range (~90 ), and the second manipulated the sensor-potentiometer system over a larger range (~290 ). Results The results of the separation distance testing begin on page 3; planar tilt results begin on page 8; axial misalignment results begin on page 13; acceleration results begin on page 16; hysteresis results begin on page 18. For the planar separation, planar tilt, axial misalignment and acceleration tests the reference potentiometer is plotted in blue; the sensor s response is shown in red. For the hysteresis testing forward rotation is plotted in red and backward rotation is plotted in blue. Note that left/right phase shifting results from not synchronizing the potentiometer and sensor s motion (azimuth misalignment). This phase shifting only affects the left/right position of the wave and does not have any impact on the shape of the resultant waveforms. This was done intentionally in order to separate the potentiometer and sensor waveforms for convenient viewing.
PAGE 3 of 19 Separation Distance No separation, application magnet in contact with face of sensor 0.0625in (1.59mm) separation
PAGE 4 of 19 0.125in (3.18mm) separation 0.1875in (4.76mm) separation
PAGE 5 of 19 0.2500in (6.35mm) separation 0.3125in (7.94mm) separation
PAGE 6 of 19 0.375in (9.53mm) separation 0.4375in (11.11mm) separation
PAGE 7 of 19 0.5000in (12.70mm) separation
PAGE 8 of 19 Planar Tilt No planar tilt, sensor in-line with application magnet. 5 planar tilt
PAGE 9 of 19 10 planar tilt 15 planar tilt
PAGE 10 of 19 20 planar tilt 25 planar tilt
PAGE 11 of 19 30 planar tilt 35 planar tilt
PAGE 12 of 19 40 planar tilt 45 planar tilt
PAGE 13 of 19 Axial Misalignment No misalignment, sensor in-line with application magnet. 0.0625in (1.59mm) misalignment
PAGE 14 of 19 0.125in (3.18mm) misalignment 0.1875in (4.76mm) misalignment
PAGE 15 of 19 0.2500in (6.35mm) misalignment
PAGE 16 of 19 Acceleration No separation, application magnet in contact with face of sensor 0.0625in (1.59mm) separation
PAGE 17 of 19 0.125in (3.18mm) separation 0.1875in (4.76mm) separation
PAGE 18 of 19 Hysteresis Small actuated range (~90 ) Large actuated range (~290 )
PAGE 19 of 19 Conclusions The 30mm non-contact sensor performed admirably in all three of the out-of-specification operating conditions. During the separation distance test the sensor managed to track the application magnet to within 10 percent of its actual location when separated by 0.5000in. Additionally, even at the 0.5000in position the periodicity of the signal remains accurate. Similarly, when the planar tilt angle was increased outside the recommended operating envelope to a maximum of 45 the sensor managed to track the application magnet to within 10 percent. Periodicity was also maintained in the same manner. As the magnet was moved off-center during the axial misalignment test, a sinusoidal nonlinearity appeared in the sensor response. This was an expected result due to the geometric relation of the ring magnet to the sensor magnet. It was also observed that there was a sharp decline in the sensor s ability to track the application magnet once the center of the sensor moved outside of the application magnet s radius. This limitation could be easily overcome by using an application magnet with a larger radius. The 30mm non-contact sensor s error tolerance allows it to continue operating under conditions that would destroy a mechanically coupled sensor. The non-linear waveforms produced by the out-of-specification conditions manifest themselves in a known and repeatable manner. The repeatability of these waveforms could be used as the basis for software based error correction. Alternatively, this observed non-linearity could be used to alert the end user to potential damage to the sensor s housing or the application magnet s mounting. By comparing the lead/lag relation between the potentiometer and sensor it can be concluded that under normal operating conditions there is no significant delay in the sensor s response to the application magnet s motion (<1ms). Even at the 0.2500in separation distance the reaction time of the sensor is approximately 8ms. Note that the sinusoidal response is characteristic of the sensor s magnet spinning up to track the application magnet. As the sensor s magnet reaches the angular velocity of the application magnet this effect diminishes. This effect can be reduced by either moving the sensor closer to the application magnet or by using a more powerful application magnet in order to achieve a tighter magnetic coupling. The flat hysteresis curves indicate that the sensor did not exhibit any hysteretic behavior in either of the trials.