TI Training home
Anti-tamper techniques to thwart attacks on smart meters
Detecting magnetic tampering using hall-effect sensors Detecting magnetic tampering using hall-effect sensors: Magnetic tamper detection testing
Anti-tamper techniques to thwart attacks on smart meters
3.2 Detecting magnetic tampering using hall-effect sensors: Magnetic tamper detection testing
As we were releasing the TIDA-00839 design, we ran additional magnetic testing to better reflect the magnetic test conditions for e-meters. To perform this test, the TIDA-00839 board was modified to fit into our standard e-meter case. For the magnetic testing, we took the modified 00839 board within the e-meter case and then applied a typical magnet used to tamper with e-meters in order to verify that DRV5033 was able to detect the magnetic field.
In the tests, both north and south poles of the magnet were tested for the x, y, and z directions. In the design, there is a separate hall sensor for each direction where the hall sensor designated U$4 corresponds to the z direction, U$5 corresponds to the y direction, and U$6 corresponds to the x direction. If the magnetic flux density sensed by a particular hall sensor is above the operating point threshold for the DRV5033, which is 6.9 millitesla for this particular variant, then a specific LED would turn on to indicate it.
Once the sensed magnetic flux density is below the release point, then the corresponding LED would turn off. In this design, the three LEDs on the right are used for the three hall sensors. For the magnetic test, a powerful magnet shown on the right was used for testing. This magnet was a cylindrical N45 magnet that was 1.3 tesla. It had a diameter of 63.5 millimeters and a height of 25.4 millimeters.
The magnet was actually magnetized so that the north and south poles were on the cylinder faces. For testing, the magnet's spacers were left intact to add separation between magnet surface and EVM case to further reduce the magnetic field. For performing the magnetic test, we applied the magnet to the following three phase e-meter case. In this test, these magnet placements correspond to the x, y, and z directions.
For the x direction, the distance from the hall effect sensors to the magnet spacer was 75 millimeters for the x direction, 95 millimeters from the y direction, and 34 millimeters for the z direction. Base on the strength and dimensions of the magnet, as well as the distance from the magnet to the corresponding hall sensor, we were able to calculate the theoretical magnetic flux density upper bound at each hall sensor. For the x, y, and z directions, the theoretical magnetic flux density was 22 millitesla, 12.5 millitesla, and 102 millitesla with the higher flux density for the z direction being the result of the short distance from the magnet to the hall effect sensor.
In these conditions, the sensed magnetic flux density for each orientation was greater than the operating point of the DRV5033. As a result, theoretically, we should be able to detect the magnetic field using the DRV5033HA for the test magnet. In fact, even if the magnet's strength is reduced 0.8 tesla, assuming the same case and magnet dimensions, each hall sensor would still be able to detect the resulting magnetic field. For the DRB503X devices, there are various sensitivity options.
A similar set of calculations that can be used for verifying a particular device variant would be the proper sensitivity based off of different magnet or different distances. To test the calculations, we applied the magnet to the case. This video shows the testing performed. For the first test, the magnet is placed in the y orientation with respect to the case. For this test, you see that the yellow LCD turns on as expected.
After testing the y orientation, we then test the z orientation. Once the magnet is placed in the z orientation with respect to the case, the corresponding orange LED turns on as expected. Finally, we test the x orientation. Once the magnet is placed in the x orientation with respect to the case, the green LED turns on as expected.
In the video, you could see the proper hall effect sensor LED turned on for each orientation. We showed that to the DRV5033HA device had the proper sensitivity to sense a strong magnetic field from the applied magnet.
To summarize the magnetic tamper protection part of this training, magnetic tampering is one of the most common ways to tamper with a meter. This is dealt with by first detecting strong magnetic fields so that the appropriate action can be taken, such as applying a penalty fee to the customer for trying to tamper with it. Hall sensors can be used for detecting the strong magnetic fields. When using hall sensors, two through-hole packages and one surface mount package are used to sense across three dimensions.
This will allow sensing strong magnets regardless of how they are oriented on the case. The TIDA-00839 TI design shows how the DRV5033 can be used to detect strong magnetic fields in all three directions. In this design, we do cycle the DRV5033 to reduce the average current consumption to be less than 2 microamps at a sensing frequency of 4 times per second. This low current consumption allows running the design off of a battery for an extended amount of time. By doing additional magnetic tests with this board in our e-meter case, we showed that the DRV5033 had the proper sensitivity to detect magnetic tampering.
April 12, 2017
This module discusses magnetic tampering tests that were performed on the TIDA-00839 reference design to show that the DRV5033 has the proper sensitivity to detect magnetic tampering for e-meters. A video of the conducted magnetic tamper tests is shown. This module concludes by summarizing this module as well as the other module in the “Detecting magnetic tampering using Hall-effect sensors” section of the “Anti-tamper techniques to thwart attacks on smart meters” training series.
This course is also a part of the following series