Magneto-inductive sensors offer a myriad of other possibilities especially in smart parking technology design.
Driving around looking for a place to park your car is a ‘hit-and-miss’ endeavour. In this situation, a device in your car that can locate available parking spaces in the area and guide you right to it is must have.
The above example is only one situation and need for an application using a magneto-inductive (MI) sensor in its design. There exists a myriad of other possibilities for this type of sensor, only limited by the designer’s imagination.
The MI sensor IC
PNI has an MI sensor IC, the MS2100, consisting of two MI sensors making up a two-axis sensing solution plus a control ASIC. We will use this sensor as an example in the following tutorial. The following is how an MI sensor works and what sets the PNI solution apart from anything else out there that I have seen in performance.
There are many ways to employ an MI sensor but the usual circuit includes an L/R oscillator configuration with the MI in the feedback of a Schmitt trigger (Figure 1).
Figure 1: Typical MI sensing circuit design.
The magnetic field parallel to the coil is shown as HE. H is the total magnetic field that the MI sees and is a function of two elements: the magnetic field formed by the current, I, in the circuit and the external magnetic field. The equation H = k₀I + HE in which k₀ is a constant depending upon the physical parameters of the particular sensor.
For Figure 1, it is assumed that there is a “0” value at the Schmitt trigger input, A; i.e., 0 or some value less than the trigger value. The output then inverts to a logic “1” at a voltage we will call VS. Now that VS across the voltage up across the MI sensor until the point A voltage reaches the Schmitt trigger threshold voltage, VH, and is seen as a logic “1” at Point A, thereby making the output a logic “0”. Now the voltage across the MI sensor is driven down and the result is a sustained oscillation (Figure 2).
Figure 2: The waveforms for the oscillator circuit. The current, I, follows the voltage waveform at point A.
Now here is the underlying principle to PNI’s magneto-inductive sensing technology. PNI designs its MI sensors with a solenoidal geometry coil which gets wrapped around a high-permeability magnetic core. The inductance of such a highly permeable material varies with the applied magnetic field. So the sensor’s inductance, μ, is a function of H, the magnetic field (Figure 3).
Figure 3: A graph shows the inductance, μ, of the highly permeable material vs. H, the magnetic field.
In Figure 1, Rb is the bias resistance and it, along with the voltage VS on the Schmitt trigger, are chosen so that the sensor’s magnetic field is in the non-linear region of the permeability curve shown in Figure 3.
Now let’s look at the voltage output when this circuit is driven with a positive or a negative bias without an applied external magnetic field (Figure 4).
Figure 4: The sensor circuit performance curve with no externally applied magnetic field. The period of the oscillation is the same when biased either positively or negatively.
Now we apply an external magnetic field, HE which can also be the Earth’s magnetic field, and we see that both the positively and negatively biased curves shift in the same direction. It can be seen that when the circuit is positively biased, the shift causes the inductance to increase and when it is negatively biased, the inductance now decreases. This effect causes τ, the period between cycles, to increase for the positively biased circuit and decrease for the negatively biased (Figure 5).
Figure 5: The sensor circuit operation with an applied external magnetic field is shown here.
Now, if we measure the time to complete a fixed number of oscillations or periods which occur during the forward and reverse polarity directions, then take the difference between these two values, we can derive the strength of that external magnetic field.
Although there are existing products out there in the industry, accurate and timely vehicle detection is lacking. Enter the PNI PlacePod, an IoT-enabled, smart parking sensor that facilitates on- or off-street parking in either a private or municipal parking management facility. PNI claims that it is the industry’s most accurate magnetic sensor system for vehicle detection. The system combines PNI’s 3-axis RM3100 geomagnetic sensor and the SENtral-A2 coprocessor with algorithms for accurately detecting the presence or absence of a vehicle in a particular parking space (The device has two states: empty and vehicle parked) (Figure 6).
Figure 6: PlacePod offers end-to-end encryption, multiple nodes and easy installation. It also integrates with third-party networks.
A major reason PNI’s design stands out among the rest of the field is that these sensors and mature algorithms are designed for ultra-low power consumption and are continuously in an on state, as opposed to other systems which go into a sleep mode to conserve power—this is the time when you can miss that a vehicle has either come or gone from the parking space. I find this and the following reasoning to be a very compelling reason for designers to select this solution over others in the market from what I see out there in the industry today.
Some other features that set this device apart are:
- Stability over temperature is critical, especially in varying outdoor temperatures. Temperature stability is inherent to the device because of the reverse forward-biasing nature of PNI’s MI circuit which athermalises (i.e., becomes independent of temperature or thermal effects) the output since the changes in the bias resistance or the inductance over temperature are experienced in both the forward and reverse directions and hence are cancelled.
- The device has essentially a digital output which produces a clearly defined value directly related to the applied magnetic field. Many other devices have an analog output which would need the addition of an op amp and A to D converter externally.
- The device is capable of a high resolution of 10nT using the basic circuit. Other devices have a noise floor which limits their resolution, or are more costly in order to get better resolution using external components.
- The device’s low power consumption, due to the MI technology, is inversely proportional to the sample rate. With an 8Hz sample rate, power consumption is typically 1.5mW; at a 300Hz sample rate, it is 7.5mW. A magneto-resistive sensor (MR) solution will typically consume 15mW to 30mW. A 10-year battery life is possible depending upon the configuration and distance of the device from the gateway.
- This device has no need for reset like an MR technology which often needs a high current pulse to be sent through that sensor in order to reset the magnetic domains within the core material. This reset pulse also generates a large peak power of several watts which may cause problems in other devices in the system.
- There is no hysteresis to affect accuracy in the MI sensor due to the forward reverse bias nature of the MI circuit.
The system uses BLE for simplicity of the wireless set up and calibration, as well as providing diagnostics and software updates using a mobile or a desktop app. It is IoT-enabled with LoRaWAN 1.0 compliant compatibility using the built-in, long-range 915MHz LoRa-based radio module for communication (uses sub-GHz ISM bands in North America and Europe) with a LoRa-enabled gateway.
The device can be either embedded into the ground or inserted onto the surface of a street, road or into parking spaces in multi-level/single-level parking structures since it has a robust polycarbonate housing (Figure 7).
Figure 7: An example of high accuracy, smart parking sensor technology.
Finally, the device is environmentally tested to MIL-STD-810G and NMEA Type 4 IP67 (IEC60529).
First published by EDN.