Fullscreen HTML5Med Res (??MB)HTML4

Driveway Monitor Pt.1: By JOHN CLARKE Based on a Honeywell magneto-resistive sensor, this Driveway Monitor provides an audible and visual indication when a vehicle is detected. Alternatively, it can be made to activate a remotecontrolled mains switch to turn on lights etc for a preset time. O UR DRIVEWAY MONITOR will alert you to any vehicle arriving in your driveway and it’s equally useful on a farm, detecting vehicles passing through a gate. Several methods can be used for vehicle detection, including infrared and ultrasonic beam set-ups. However, infrared and ultrasonic beams are prone to false triggering so a typical vehicle detection system relies on the very small changes in the Earth’s magnetic field caused the presence of a vehicle. Fig.1 shows a representation of the distortion in the Earth’s magnetic field caused by the presence of a vehicle. Our previous Driveway Sentry units published in November 2004 & August 2012 used a coil of wire to detect changes in the Earth’s magnetic field when a vehicle passed over it. This coil could either be laid under the driveway or concealed in the expansion joints, if that was possible. While this arrangement can work well if you are installing a new driveway, you wouldn’t want to jack-hammer an existing concrete driveway to lay a cable under it! Our new Driveway Monitor sidesteps this problem by using a Honey26  Silicon Chip well magneto-resistive sensor, as used for magnetic field sensing in electronic compasses. It’s so sensitive that it doesn’t need to be located under the driveway; somewhere alongside the driveway is sufficient. The magneto-resistive sensor is teamed with a sensitive instrumentation op amp and a PIC microcontroller which outputs a coded 433MHz signal. This is picked up by a companion 433MHz receiver unit with various optional outputs. Once triggered, the receiver unit flashes a green or red LED and sounds a piezo alarm. It will even tell you which way the vehicle is heading, since a different distinct sound is made by the piezo transducer for each direction, while a third tone indicates vehicle movement in either direction. In addition, the green LED flashes for vehicles entering the driveway, while the red LED flashes for vehicles exiting the driveway. Alternatively, the LEDs and piezo transducer can be omitted and a couple of reed relays fitted to the receiver PCB instead. These can be used to trigger a UHF remote-controlled mains socket (via its remote), a wireless doorbell remote or perhaps even the remote for a motorised gate opener. The detector circuitry is installed in an IP65 case (115 x 90 x 55mm) that’s dust tight and able to withstand wet weather. This would normally be mounted alongside the driveway, either on an adjacent fence, wall or post. Power for the detector comes from a single 1.25V AA NiMH cell that’s recharged using a solar cell panel (the same as those used with solar garden lights). No mains power is necessary. On the other hand, if you don’t want to use solar power and there is access to undercover mains power, a small 9-12V DC plugpack could be used to recharge the NiMH cell. The companion receiver unit is housed in a small plastic case and is powered from a 12V DC plugpack. It can be placed where it can be readily heard and seen, if you intend using it purely as an audible/visual indicator. Alternatively, it can be placed out of sight if you intend using it to trigger a remote controlled mains socket or some other device. Block diagram Fig.2 shows the block diagram of siliconchip.com.au Fig.1: how the Earth’s magnetic field is disturbed by a vehicle travelling along a driveway. These disturbances are detected by the magneto-resistive sensor used in the Driveway Monitor. +5V +5V SWITCHED 3 SET/RESET STRAP OUT– SENSOR1 OUT+ Each time the sensor detects a large change in the surrounding magnetic field, its magneto-resistors need to be reset by means of an internal “strap coil” which provides a strong magnetic field. Hence the strap coil is subjected to a short reset signal from the micro. And in fact, before a measurement can be made, a “set” signal must also be applied, again by the microcontroller siliconchip.com.au IC1 INSTRUM OUT AMP REF 5 –IN 4 2 +5.5V Q2 P-CHAN MOSFET 1 µF Strap coil LOW PASS FILTER 7 Vdd Vset 6 AN2 IC2 MICROCONTROLLER PWM The detector PCB carries the magnetoresistive sensor and is mounted in a waterproof case near the driveway (note: prototype PCB shown). the detector circuit. It’s based on the Honeywell HMC1021S one-axis magneto-resistive sensor, instrumentation amplifier IC1, PIC microcontroller IC2 and Mosfets Q1 & Q2. The magneto-resistive sensor is essentially a Wheatstone bridge of four resistors (which are affected by magnetic fields). The bridge is connected across a 5V supply and a change in the local magnetic field changes the resistor values so that the voltages at the sensor’s outputs move up or down. This shift in the output terminals is monitored by differential instrumentation amplifier IC1. Its output feeds microcontroller IC2 via a low-pass filter. If the magnetic field around the sensor changes, the micro sends an appropriately coded signal to a 433MHz transmitter module (not shown on Fig.2). +IN S LOW PASS FILTER G RB1 D Q1 N-CHAN MOSFET D S RB0 Vss G Fig.2: block diagram of the detector circuit. The output from the magnetoresistive sensor (Sensor1) is amplified by differential op amp IC1 which then feeds the AN2 input of microcontroller IC2 via a low pass-filter. IC2 then processes the amplified sensor signal and also drives Mosfets Q2 & Q1 via its RB0 & RB1 outputs to provide set and reset signals to a strap coil in the sensor. In addition, IC2’s PWM output applies an offset voltage to the REF input of IC1, so that IC1’s pin 6 normally sits at 2.5V (half-supply). (see the panel on pages 30-31 for further details). The set and reset pulses appear at IC2’s RB0 & RB1 outputs respectively and these drive Mosfets Q1 & Q2. The resulting pulses are fed to the strap resistor via a 1µF capacitor. Each time IC2 takes RB0 low for a set pulse, Q2 switches on and current flows through the 1µF capacitor and the set/reset strap to ground (0V). The 1µF capacitor charges to +5.5V and Q2 then switches off . The amplified sensor output (Set) is then read at IC2’s AN2 input. Conversely, for the reset pulse, Q1 is switched on when RB1 is taken to +5V. The 1µF capacitor then discharges through the set/reset strap with current now flowing in the opposite direction than for the set pulse. Q1 is then switched off to end the reset pulse and the amplified sensor output (Reset) is again read at the AN2 input. IC2 then needs to apply an offset voltage to IC1 so that its output at pin 6 normally sits close to 2.5V and thereby ensure that its output swing is symmetrical. IC2 calculates this offset voltage by averaging the readings after the set and reset pulses. It then generates a pulse width modulated signal at its PWM output and this is fed via a low-pass filter to pin 5 (REF input) of IC1. The PWM signal generated by IC2 switches between 0V and 5V at 7.8kHz. Its duty cycle is automatically adjusted after each measurement to correct for any offset changes from both the sensor and IC1 due to temperature changes. IC1’s output is also low-pass filtered, to reduce any voltage ripple at IC2’s AN2 input, due to the PWM signal, to July 2015  27 +5V AMPLIFIED OUTPUT DUE TO CHANGE IN FIELD STEADY STATE LEVELS UPPER THRESHOLD AMPLIFIED SENSOR OUTPUT LOWER THRESHOLD +2.5V TRACKING THRESHOLDS DETECT DETECT 0V Fig.3: this diagram shows how IC1’s pin 6 output changes when a vehicle comes close to the sensor. It either decreases and then rises as the vehicle approaches (as shown here) or it increases and then falls, depending on the orientation of the sensor and the direction of the vehicle. A vehicle is detected whenever the amplified sensor output exceeds the slowly-averaged upper and lower thresholds set by IC2 in response to IC1’s steady-state output. a very low level. This allows microcontroller IC2 to detect voltage changes from the sensor as low as 5mV without being swamped by noise and ripple. Detecting a vehicle Fig.3 shows how IC1’s output changes when a vehicle comes close to the sensor. As can be seen, it either decreases and then rises as the vehicle approaches or it increases and then falls, depending on the orientation of the sensor with respect to the Earth’s magnetic field and the direction of the vehicle. This enables the microcontroller to determine the vehicle’s direction. A linking option on the PCB tells the microcontroller which is the entry direction and which is the exit direction. IC2 detects a vehicle when IC1’s output rises above an internally-generated upper threshold or falls below a lower threshold. These two thresholds are set equidistant above and below the steady-state amplified sensor output. Main Features •  Remote vehicle detection with adjustable detection sensitivity •  Vehicle detection LED indication •  Vehicle direction detection •  Solar panel and NiMH cell powered •  Transmits vehicle detection via UHF link to the receiver •  Typical UHF range: 200m in open space •  Eight possible UHF transmission identities to allow for multiple Driveway Monitor pairs to be used in close proximity •  Selectable vehicle entry only or exit only detection or both entry and exit detection •  Optional non-directional indication •  Vehicle direction detection setting to cater for detector positioning and driveway orientation   Over range indication (flashes red & green LEDs in detector unit) alternately. • •  Diagnostic setting •  Receiver has audible and visible indication of vehicle detection •  Receiver produces different sounds for exit and entry unless non-direction detection is selected on the detector unit •  Detection sampling rate: typically 300ms •  Set & reset pulses: every 10s 28  Silicon Chip In operation, these upper and lower thresholds track the sensor’s amplified output at a slow rate to compensate for any output changes with temperature (as well as slow magnetic field changes) over time. So if the sensor’s amplified DC output falls, the thresholds will also fall. On the other hand, the tracking is slow enough to ensure that any quick changes in IC1’s output level (ie, due to vehicle movement) will exceed the thresholds for brief periods. These are shown as the “detect” periods on Fig.3 and when the thresholds are breached, the microcontroller determines that a vehicle has been detected. Note that in order to conserve the battery, the detector circuit doesn’t continuously monitor changes in the magnetic field. Instead, both the sensor and IC1 are powered for a brief period every 300ms and this is when IC2 samples IC1’s output. Circuit details Now take a look at Fig.4 which shows the full circuit of the detector unit. The top section can be regarded as a more detailed version of the block diagram of Fig.2 but it also includes 433MHz UHF transmitter module TX1, PNP transistors Q3 & Q4 and a switching power supply. Q3 & Q4 are respectively driven by RB2 & RA3 of IC2 and switch power to the sensor, IC1 and the UHF transmitter module every 300ms, as mentioned above. The switched 5V supply rail from Q3 is decoupled using a 1µF electrolytic capacitor and a 100nF cer­amic capacitor. The power supply uses a TL499A switchmode step-up regulator (REG1), a linear 5V regulator (REG2) and an LMC6041 micropower op amp (IC3). As stated above, it’s powered from a 1.25V NiMH AA cell that’s topped up by a solar panel. Alternatively, by cutting a track on the PCB and installing resistor R1, a 9V or 12V DC supply can be used to maintain cell charge if a mains supply is available. The 5V rail from REG2 directly powers microcontroller IC2 and is switched to sensor 1 and IC1 by Q3 and to the TX1 module by Q4. Sensor 1 draws a current of about 5mA each time it is briefly powered up. The Out+ and Out- terminals are fed to the IN- and IN+ inputs of IC1 via ferrite beads, while the 1nF bypass siliconchip.com.au Q3 BC327 C SENSOR1 HONEYWELL HMC1021s 100nF* 2 100nF* GAIN 1 +Rg IC1 AD623 OUT REF 1 –Rg 2 –IN 4 FB2 6 8 2.2k 1 1 µF 5 9 1nF* LK1 ENTRY SR+ 5 Q2 IRF9540 SET/RESET STRAP SR– EXIT 220Ω 6 1 µF S G SWAP 12 LK3 11 470 µF 1 µF 10 10V LOW ESR D D Q1 IRF540 +5.5V 13 LK2 6 10Ω 7 λ A PWM IC2 PIC1 6F8 8 PIC16F88 RB7 15 4.7k 2 RA3 E B RB6 IDENTITY VR3 10k 17 16 RB5 Vcc TX1 RB4 RA1 Q4 BC327 C TP1 RA0/AN0 RB0 K λ AN2 18 433MHz TX MODULE DATA RB1 ANT Vss 10Ω S 2.2k 3 RA4 RB2 Vout 100nF LED1 MCLR RA6 22k 3 4 14 Vdd 7 +IN 10k 100nF B 10k 3 4 VR1 500Ω 8 OUT+ 1 µF 1nF* FB1 OUT– +5V E 5 G GND * CERAMIC L1 470 µH SEE TEXT D1 1N4004 R1 (1W) TO + SOLAR PANEL – CON1 A +5.5V 3 SW REG K IN2 CUT TRACK IF R1 USED 1.25V NiMH CELL SW IN REG1 TL499A SW CUR CTRL K D2 1N4004 470 µF 4 OUT REF 6 IN 2 220 µF 100nF 6 IC3 A 10 µF 2 4 330Ω +5V 3 7 10V LOW ESR OUT GND VR2 1M 1nF 10V LOW ESR 7 5 REG2 LM2936Z-5.0 IC3: LMC6041 8 GND PGND TP5.5 100k TP GND 1N4004 A SC  20 1 5 LM2936Z K DRIVEWAY MONITOR DETECTOR K A Q1, Q2 BC 32 7 LED IN B OUT GND E G C 433MHz Tx MODULE D D ANT Vcc DATA GND S Fig.4: the full circuit diagram for the detector unit. It includes all the elements shown in Fig.2 and also shows LED1 (for exit and entry indication) and the 433MHz transmitter (TX) module which is driven by IC2’s RA1 port. Q3 briefly switches power to the sensor and IC1 at 300ms intervals while Q4 briefly switches power to the TX module, to minimise current consumption. Power comes from a 1.25V NiMH cell topped up by a solar cell. Switching regulator REG1 steps up the voltage to produce a 5.5V rail for Q2 & Q1, while REG2 produces a regulated 5V rail for the rest of the circuit. capacitors to ground attenuate any RF signals. In addition, a 100nF capacitor bypasses the IN+ and IN- inputs to provide further RF suppression. IC1 is an Analog Devices AD623 differential amplifier and it draws about 300µA from the 5V supply. Its gain is adjusted using trimpot VR1 and can be varied from about 201 times when VR1 is set to 500Ω (its maximum) up to about 1000 times when VR1 is set to 100Ω. IC2’s PWM signal is fed to pin 5 (REF) of IC1 via a low-pass filter consisting of a 22kΩ resistor and a siliconchip.com.au 100nF capacitor. The filter sets the roll-off frequency to about 72Hz and this effectively removes a considerable amount of the 7.8kHz PWM switching frequency. However, by itself that is not effective enough to remove sufficient PWM ripple and so IC1’s output also has low-pass filtering, using a 2.2kΩ resistor and 1µF capacitor. Microcontroller IC2 (a PIC16F88) converts the voltage at its AN2 (pin 1) input into a digital value using a 10-bit A-D converter. This gives a resolution of close to 4.9mV. The variation available for the PWM output also has 10-bit resolution, allowing IC1’s offset voltage to be set in 4.9mV increments. Set & reset signals Mosfet Q2 is driven by IC2’s RB0 output and provides the set pulse drive current, while Q1 is driven by RB1 and provides the reset pulse drive for Sensor1’s set/reset strap. This strap is a coil with about 7.7Ω resistance and it produces the high magnetic field required to realign the elements in the sensor along the “easy” axis (see the further description of the July 2015  29 How A Magneto-Resistive Sensor Works The Honeywell HMC1021S sensor used in this project is a one-axis type. In essence, this means that it only reacts to changes in the horizontal component of the Earth’s magnetic field (assuming that it is installed on a vertical PCB. Fig.5 shows the basic construction of this type of sensor which comprises four identical resistive elements arranged in a Wheatstone bridge configuration. Each element is basically an NiFe (nickel-iron) thin film that changes its resistance in response to changes in the magnetic field passing through it. Whether an element increases or decreases its resistance with magnetic field strength depends on its orientation and the magnetic field polarity. Fig.5 shows how the elements in the sensor are arranged. Two diagonally opposite elements are orientated one way, while the other two are orientated in the opposite direction, in the Wheatstone bridge. Because of its sensitivity to magnetic field direction, a magnetoresistive sensor is often called an “anisotropic magneto-resistive” sensor or AMR. The term “anisotropic” simply means directional. In operation, a supply voltage is applied between the top and bottom of the bridge (ie, between its Vb and GND terminals), so a current flows through the elements. If a magnetic field is absent, the OUT+ and OUT- terminals are both at half supply (ie, Vb/2). By contrast, if a magnetic field is present, two diagonally opposite elements will decrease in resistance while the other two diagonally opposite elements will increase in resistance. As a result, the OUT+ and OUT- terminals Vb BRIDGE CURRENT MAGNETIC EASY AXIS PERMALLOW THIN FILM OUT- OUT+ MAGNETIC SENSITIVE AXIS GND Fig.5: the magneto-resistive sensor consists of four identical thin-film elements arranged in a Wheatstone bridge configuration. Two diagonally opposite elements are orientated one way, while the other two face in the opposite direction. will change voltage by equal amounts but in different directions. In other words, one terminal will rise above half-supply by a certain amount, while the other will fall below half-supply by an equal mount. Offset voltage That’s the basic theory of the sensor operation. In practice though, real sensors have an offset voltage between OUT+ and OUT- in the absence of a magnetic field. That’s because when the sensor elements are made, there PERMALLOY (NiFe) MAGNETO-RESISTIVE ELEMENT RANDOM MAGNETIC DOMAIN ORIENTATIONS SET MAGNETISATION EASY AXIS SENSITIVE AXIS AFTER A SET PULSE RESET MAGNETISATION EASY AXIS SENSITIVE AXIS 30  Silicon Chip AFTER A RESET PULSE Fig.6: the set and reset pulses applied to the strap coil inside the sensor align the magnetic domains in the resistive element along the easy axis. will always be small variations between them, thereby causing an imbalance in the bridge. In addition, this offset voltage will vary with temperature. Another problem is that the magnetic domains in the sensor elements can move out of alignment in the presence of strong magnetic fields. Basically, the domains in the elements must be orientated along what is called the “easy” axis and this is the alignment that the magnetic domains are set to during manufacture. Fig.6 shows the general idea. Correct easy axis alignment is necessary to ensure maximum sensitivity of the sensor to magnetic fields. The most sensitive direction for magnetic field detection is perpendicular (ie, at a right angles) to the easy axis. Any external magnetic field (or a portion of that field) that is not parallel to the easy axis will cause the magnetic domains to rotate away from the easy axis and this alters the resistance of the sensor element. Conversely, when the magnetic field is removed, the magnetic domains return to their easy axis alignment, provided that the magnetic field does not exceed the specified operating range for the sensor. Set & reset pulses In practice, this all means that the sensor will periodically need to be set and reset using a high magnetic field, to realigns the magnetic domains along the easy axis. This set and reset procedure is achieved by applying a pulse current to the strap coil incorporated within the sensor. The set and reset currents used are opposite in polarity. When the coil is driven with one current polarity, it produces a magnetic field that aligns the domains in one direction along the easy axis. Reversing the current direction through the sensor’s coil then aligns the domains in the other direction (ie, it rotates them by 180°). Fig.7 shows the effect of the set and reset pulses on the sensor’s output. As shown, a brief set pulse produces a large output from the sensor, due to the large magnetic field produced by this pulse. Following the set pulse, the output from the sensor goes down to Vset which is the voltage difference between Out+ and Out-. This voltage is shown as being above the offset voltage (Voff) of the sensor and is produced in response to an external magnetic field. siliconchip.com.au Vset Voff OFFSET TIME Vcc/2 Vreset + SET SET AND RESET PULSES RESET – Fig.7: this diagram shows the effect of the set and reset pulses on the sensor’s output. Note that the output polarity switches after each pulse. A reset pulse then follows, after which the output from the sensor goes to Vreset. This again is the voltage difference between Out+ and Out- and is now below Voff. Note that after a set pulse, a subsequent reset pulse switches the polarity of the sensor’s output voltage. Similarly, after a reset pulse, a set pulse switches the output polarity back again. Fig.7 only applies for one direction of the magnetic field. If the field is reversed, then the polarities of Vset and Vreset are also reversed. In other words, Vset will be lower than Voff after a set pulse, while Vreset will be higher than Voff after a reset pulse. By contrast, the sensor’s offset voltage (Voff) is unaffected by magnetic field variations – it’s only the sensor’s output that varies. In the absence of a magnetic field, the offset Voff would simply be the difference between Out+ and Out-. In practice though, the device operates in the presence of the Earth’s magnetic field. In summary, we need the set and reset pulses to realign the magnetic domains in the sensor to ensure maximum sensitivity. As a bonus, this also provides a means to calculate the sensor’s offset siliconchip.com.au voltage (Voff) and thus compensate for it. That’s done by simply adding the Vset and Vreset values together and dividing by two. Calculating Voff at regular intervals then allows us to compensate for offset changes with temperature. Why compensate for offset? But why do we need to compensate for the sensor offset? The reason is that changes in the sensor’s output in response to magnetic field variations are quite small and so we need to amplify its output. Assuming a 5V supply (as in this circuit), the output varies by around 2.5mV, depending on the sensor’s orientation within the Earth’s magnetic field (approximately 50μTesla or 0.5 Gauss). By contrast, the sensor’s offset could be up to 11.25mV. So if a 2.5mV signal is amplified by say 500 to obtain a 1.25V signal, the 11.25mV offset voltage would also be amplified by 500 to a level of 5.6V. That means that unless we compensate for the sensor’s offset voltage, the amplified signal could result in the amplifier’s output being pegged at either the positive or 0V supply rail, with no resulting change in level due to magnetic field variations. workings of the magneto-resistive sensor in the panel at left). As shown on Fig.4, Q2’s source is connected to the +5.5V supply rail via a 220Ω isolating resistor and decoupled using a 470µF low-ESR capacitor and a 1µF MKT capacitor. The set pulse current is applied to the strap via the 1µF capacitor as it charges when Q2 turns on, while the discharge current from this 1µF capacitor provides the reset pulse when Q1 turns on. Both the charge and discharge peak currents are in excess of the 500mA minimum required for this operation. The accompanying oscilloscope traces (Fig.8 & Fig.9) show the set and reset pulses. In each case, the top trace is the drain voltage of Q2 & Q1, while the lower trace is the pulse applied to the set/reset strap of the sensor. Note that the set pulse is a positive voltage while the reset pulse is negative. Note also that there is a small amount of “dead time” between when Q2 is switched off and Q1 is switched on. This ensures that they aren’t both on at the same time (however briefly) which is necessary to prevent a momentary short across the decoupled supply rail. In operation, RB0 & RB1 of IC2 drive the Mosfet gates every 10s and both the set and reset pulses decay away over time. These pulses produce a magnetic field in the sensor, so the amplified sensor signal from IC1 is checked by IC2 only while both Q1 & Q2 are switched off to ensure that only variations in the Earth’s magnetic fields are detected. Detection & link options As previously mentioned, the voltage fed by IC1 to AN2 of IC2 is compared against high and low threshold voltages that track AN2’s voltage at a slow rate. Whenever AN2’s voltage varies, the thresholds are adjusted up or down by 4.8mV every 1.5s. How­ ever, a moving vehicle will cause AN2’s signal voltage to vary by considerably more than 4.8mV in much less than 1.5s and so the thresholds will be exceeded. IC2 detects whenever AN2 goes below the lower threshold or above the upper threshold and drives a bi-colour red/green LED (LED1). The green LED lights for five seconds if AN2’s voltage initially goes below the lower threshold, while the red LED lights for 5s if it goes above the upper threshold. July 2015  31 Fig.8: this scope grab shows how the set pulse for the strap coil is generated. Each time IC2’s RB0 output briefly goes high, Mosfet Q2 switches on and the commoned Mosfet drains go high as shown by the orange trace. The bottom green trace shows the resulting positive-going set pulse that’s then applied to the strap coil via the 1μF capacitor. During this time, the detection process is disabled. At the same time, a vehicle detection signal is sent to the receiver circuit by the 433MHz transmitter (TX) module, depending on the linking options selected for LK1, LK2 & LK3. LK1 is used if you want entry (arrival) notifications to be transmitted, while LK2 is installed if you want exits (departures) to be transmitted. Either LK1 or LK2 can be installed, or both can be installed to warn of both arrivals and departures. LK3 is the “swap” link and is used to set the unit so that it correctly identifies the vehicle’s direction (entry or exit). As stated, this direction indication initially depends on the orientation of the driveway and which side of the driveway the detector unit is mounted on. If the directions are incorrect, it’s Fig.9: the following reset pulse is generated when RB1 subsequently briefly goes high. This turns on Mosfet Q1 and the commoned Mosfet drains are then pulled to 0V as shown by the orange trace. A negative-going reset pulse (green trace) is then generated as the 1μF capacitor discharges just a matter of installing the link. Installing LK3 simply swaps over the exit and entry transmission codes that are sent to the receiver and the detection LED colour. Non-directional signalling Yet another link option (not shown on Fig.4) forces the Driveway Sentry to send a non-directional signal to the receiver unit, instead of separate entry and exit signals. That’s done by installing a link between LK1 & LK2 to short pins 12 & 13 of IC2. The receiver unit then simply indicates that a vehicle has passed by the detector without indicating its direction. Yet another option is to install a link between LK2 & LK3 to short pins 11 & 12 of IC2. This is a diagnostic connection and we’ll describe this in greater detail next month. Fig.10: the top trace in this scope grab shows the reference voltage applied to pin 5 of IC1, while the bottom trace shows the filtered output from pin 6 that’s fed to the IC2’s AN2 input. The reference voltage is about 180mV above the half supply of 2.5V to compensate for the sensor’s offset. 32  Silicon Chip IC2 determines which links have been installed by first pulling its RB5, RB6 & RB7 inputs high (ie, to +5V). Its RB4 output is then pulled low (0V) and the RB5-RB7 inputs checked to see if any of these are also now low. If so, then a jumper link must be installed on that particular input. Determining if there is a connection between RB7 & RB6 or between RB6 & RB5 is only slightly more complicated. It’s done by first making RB6 an output and RB4 an input. RB6 is then taken both low (0V) and high (5V) and RB7 & RB4 checked to see if either one follows RB6. If an input follows, then there is a jumper link between it and RB6. 433MHz UHF transmitter TX1 is the 433MHz transmitter module. Its supply line is switched by Q4 and this transistor is turned on by IC2’s RA3 output whenever transmission is required (ie, Q4 turns on when RA3 goes low). Trimpot VR3 is also connected to the +5V supply rail when Q4 turns on. Its wiper is monitored via IC2’s AN0 input and the set voltage is included in the UHF transmission as identity information. This voltage then needs to match the voltage set on a similar trimpot in the receiver unit in order for the transmission to be accepted (ie, in order for pairing to take place). There are eight valid voltage ranges that can be set using VR3 to select one of eight different identities. As a result, up to eight different detector and resiliconchip.com.au Parts List: Detector Unit ceiver pairs can operate independently in close proximity. Conserving the battery As already noted, Sensor1 draws about 5mA and the AD623 amplifier (IC1) about 300µA from the 5V supply when connected via Q3. That’s a total of 5.3mA from the 5.5V output of REG1 and means that around 25mA would be drawn from the single 1.25V AA cell that powers everything (taking into account power conversion and efficiency). Because of this, a number of steps have been taken to minimise the power consumption. First (and as previously mentioned), Sensor1 and IC1 are only powered up each time a measurement is required and that’s done for only about 20ms at 300ms intervals. This 20ms duration was chosen to give sufficient time for the filters at IC1’s reference (REF) input and at its output to settle (ie, much longer than the lowpass filter time constants of 2.2ms). As a result, the power on/off ratio is 1/15 and so the average current drawn from the 5V supply is just 5.3mA x 1/15th = 353µA. The 433MHz UHF module draws 10mA when powered, while VR3 draws a further 500µA. However, they draw very little power overall, since they are only powered up when a UHF transmission is required (ie, when a vehicle is detected). Even if a vehicle stops next to the detector, the unit will quickly stop transmitting as the upper and lower thresholds catch up to the voltage on IC2’s AN2 pin. Further power is saved by shutting down microcontroller IC2 so that it is in sleep mode for most of the time and drawing a maximum current of just 11µA. This current is much lower than when actually running its internal program and drawing up to 2.8mA. In operation, IC2 is woken up for 20ms every 300ms by a watchdog timer that runs while it is in sleep mode (ie, 2.8mA is drawn for just 20ms every 300ms). This means that IC2’s average current is just 187µA. An additional power saving has been made by having IC2’s RB4 output normally set high, so any jumper links that are inserted do not cause the internal pull-up current to flow. This can save up to 1.2mA if all the links are in place. In operation, RB4 is taken momentarily low when the link connections need to siliconchip.com.au Detector Unit 1 PCB, code 15105151, 104 x 78mm 1 IP65 polycarbonate enclosure, 115 x 90 x 55mm 1 single AA cell solar panel & wiring 1 AA cell holder 1 NiMH AA cell 1 powdered-iron toroidal core, 15 x 8 x 6.5mm (Jaycar LO-1242) 1 2-way PCB-mount screw terminal with 5.08mm spacing 1 UHF transmitter (TX1) (eg, Jaycar ZW-3100) 1 3-way DIL pin header strip (2.54mm spacing) 3 pin header shunts 1 18-pin DIL IC socket 3 8-pin DIL IC sockets (optional) 1 cable gland for 3-6.5mm cable 7 PC stakes 2 No.4 x 6mm self-tapping screws 4 M3 x 6mm screws 2 5mm ferrite beads 2 100mm cable ties 1 50mm length of tinned copper wire 1 750mm length of 0.5mm- diameter enamelled copper wire 1 170mm length of light duty hook-up wire 1 500Ω miniature horizontalmount trimpot (code 501) (VR1) 1 1MΩ miniature horizontal-mount trimpot (code 105) (VR2) 1 10kΩ miniature horizontal-mount trimpot (code 103) (VR3) be checked but again the overall average current is quite small. Power supply circuit The single 1.25V AA cell’s output is stepped up to 5.5V using step-up regulator REG1. Regulator REG2 is then used to derive the 5V rail. This second regulator helps remove any switching noise from the output of the step-up regulator and provides a well-regulated 5V supply to power Sensor1, IC1, IC2 and the 433Hz TX module. In greater detail, REG1 is a TL499A step-up regulator. In operation, current flows through inductor L1 each time REG1’s SW IN output (pin 6) switches low. When this reaches a peak value, Semiconductors 1 Honeywell HMC1021S oneaxis magneto-resistive sensor (Sensor1) 1 AD623AN instrumentation amplifier (IC1) 1 PIC16F88-I/P microcontroller programmed with 1510515A. hex (IC2) 1 LMC6041IN CMOS micropower op amp (IC3) 1 TL499A power supply controller (REG1) 1 LM2936Z-5.0 low dropout 5V regulator (REG2) 1 IRF540 N-channel Mosfet (Q1) 1 IRF9540 P-channel Mosfet (Q2) 2 BC327 PNP transistors (Q3,Q4) 2 1N4004 1A 400V diodes (D1,D2) 1 bi-colour LED (two lead) LED1 Capacitors 2 470µF 10V low-ESR electrolytic 1 220µF 10V low-ESR electrolytic 1 10µF 16V PC electrolytic 1 1µF 16V PC electrolytic 3 1µF MKT polyester 3 100nF MKT polyester 2 100nF ceramic 1 1nF MKT polyester 2 1nF ceramic Resistors (0.25W, 1%) 1 100kΩ 2 2.2kΩ 1 22kΩ 1 330Ω 2 10kΩ 1 220Ω 1 4.7kΩ 2 10Ω the SW IN output is switched off and the stored energy in the inductor is fed via an internal diode to the pin 8 output. This output is then filtered using a 100nF MKT capacitor and a 220µF low-ESR capacitor. The 330Ω resistor between pin 4 of REG1 and ground sets the peak current through the inductor to about 300mA. Voltage regulation is achieved by sampling the output voltage using a resistive divider (VR2 and 100kΩ) and then feeding this sampled voltage to the reference input at pin 2. In this case, the inductor switching rate must be adjusted so that pin 2 is kept at 1.26V. This means that for a 5.5V output, the voltage divider needs to reduce the 5.5V down to 1.26V and that’s done July 2015  33 Parts List: Receiver Unit 1 PCB, code 15105152, 79 x 47mm 1 433MHz UHF receiver (RX1) (Jaycar ZW-3102) 1 12V DC plugpack rated at 100mA 1 UB5 or UB3 case (see Pt.2) 1 8-pin DIL IC socket 1 PCB mount DC socket with 2.1 or 2.5mm centre pin to suit plugpack plug (CON1) 6 PC stakes 1 170mm length of light-duty hook-up wire 1 20mm length of 1mm-diameter heatshrink tubing 2 10kΩ miniature horizontal mount trimpots (code 103) (VR1,VR2) Semiconductors 1 PIC12F675/I-P microcontroller (programmed with 1510515B. hex (IC1) 1 78L05 5V regulator (REG1) 1 1N4004 1A diode (D3) Capacitors 2 100μF 16V PC electrolytic 1 100nF MKT polyester Resistors (0.25W, 1%) 1 1kΩ 1 100Ω with VR1 set to 336kΩ. The voltage from the divider is then buffered using op amp IC3 (which is configured as a voltage follower) before being fed to pin 2 of REG1. This buffer stage allows the use of higher-value divider resistors than would otherwise be the case and this was again done to minimise power consumption. REG1’s 5.5V output appears at pin 8 and is used to drive regulator REG2. The 5.5V rail from REG1 is also as the supply for the set/reset pulse generator circuit based on Mosfets Q2 & Q1. REG2 is a low quiescent current and low drop-out regulator. Its low dropout specification means we only need to provide 5.5V for the regulator to fully regulate to 5V. By contrast, most standard regulators require at least a 6.5V input to regulate to 5V. Average current The average current drawn from the 5V supply is around 550µA. However, the current drawn from the AA cell is much higher than this. That’s because the AA cell has an output of just 1.25V and this is stepped up to 5.5V before 34  Silicon Chip Extra Parts For Version 1 (Relays & Mains Remote Control) 1 UHF remote controlled mains switch (Altronics A 0340, Jaycar MS-6145, MS-6142) 1 UB3 box 130 x 68 x 44mm 2 SPST DIP 5V reed relays (Altronics S4100A, Jaycar SY4030) (Relay1,Relay2) 2 1N4148 diodes (D1,D2) 3 2-way pin headers 3 2-way pin header plugs 1 100Ω 0.25W 1% resistor 2 M3 x 9mm tapped spacers 4 M3 x 6mm tapped spacers 12 M3 x 6mm screws OR 6 M3 x 6mm screws AND 6 M3 x 6mm countersunk screws 120mm x 6-way rainbow/IDC cable Extra Parts For Version 2 (Audible & Visual Indication) 1 UB5 box, 83 x 54 x 31mm 2 M3 x 9mm tapped spacers 4 M3 x 6mm screws 1 piezo transducer (Jaycar AB3440, Altronics S 6140) 1 green high intensity LED (LED1) 1 red high intensity LED (LED2) 1 1kΩ 0.25W 1% resistor being regulated to 5V. So you would expect the current drawn from the AA cell to be some 5.5/1.25 = 4.4 times higher, assuming that the TL499A regulator’s step-up efficiency is 100% which, of course, it isn’t. At a more realistic 70% efficiency, the current would be expected to be 6.3 times higher. And that means that the calculated total average current drawn from the AA cell is 3.5mA. In practice, we measured a current drain of close to 3mA in our prototype. That means that a 2000mAh AA NiMH cell would last for about 28 days without recharging. The solar panel we tested charged the cell at 20mA in mid-morning autumn sunlight and that is more than sufficient to maintain the cell’s charge. Diode D1 provides protection if the solar panel is connected with the wrong polarity, while D2 provide reverse polarity protection if the 1.25V cell is inserted in its holder the wrong way around. Finally, resistor R1 is included to provide current limiting if a 9V or 12V mains plugpack is used instead of a solar panel to recharge the battery. This resistor is normally shorted out on the PCB since it is not required when a solar panel is used. However, the PCB track has a section that’s easily cut if the resistor is required (see the construction details next month). Receiver circuit Fig.11 shows the receiver circuit details. It’s based on the 433MHz receiver module, an 8-pin PIC12F675 microcontroller (IC1) and a 5V regulator (REG1). Also shown are the entry and exit LEDs, the piezo transducer and the alternative reed relays. Microcontroller (IC1) monitors the data signal output from the UHF RX (receiver) module and acts when it receives a valid code. The arrival, departure and non-directional signal codes are all different, so that IC1 can discriminate between them. IC1 also monitors trimpot VR1 at its GP4 input. This is the identity adjustment that is divided up into eight separate voltage bands. This voltage needs to match that set in the detector unit before any received signal is deemed valid. Trimpot VR2 has its wiper monitored by IC1’s GP2 input (pin 5). This trimpot sets the alert duration. Alternatively, it sets the time period between relay 1 closing and relay 2 closing (and thus the period for which a remote-controlled mains socket is powered on). In operation, IC1 converts the voltages at GP4 and GP2 to 8-bit digital values. When a valid signal is received, its GP0 and GP1 outputs drive either the piezo transducer and one of the LEDs (LED1 or LED2) or relays Relay1 and Relay2. If used, the latter are wired across the On and Off switch contacts on the hand-held remote that’s used with a remote-controlled mains socket. Piezo transducer The exit (or departure) tone from the piezo transducer is a 440Hz horn-type “bip” lasting about 1s, followed by a 440Hz tone that smoothly increases to 6.8kHz over a period ranging from 1-5s (depending on the setting of VR2). LED1 (exit) also lights while ever the piezo sounds and stays lit for about 15s after the tone ceases. By contrast, the entry (or arrival) tone starts with a 1s 440Hz horn “bip” and is followed by a 6.8kHz tone that decreases to 440Hz (again adjustable siliconchip.com.au D3 1N4004 REG1 78L05 +5V OUT 100nF K IN GND 100 µF 4 2 DATA GP0 GP5 PIEZO TRANSDUCER 100Ω 1 Vdd MCLR +12V 7 RELAY 1 3 GP4 GP2 VR2 10k 5 Vss 8 IDENTITY ENTRY LED2 ALERT DURATION A λ K A 100Ω λ LED1 K D3 1N4004 A A K THESE PARTS USED ONLY FOR AUDIBLE & VISUAL INDICATION D2 1k OFF A 433MHz Rx MODULE K 78L05 LEDS SC DRIVEWAY MONITOR RECEIVER RELAY 2 K TP1 D1, D2: 1N4148 ON D1 EXIT GND K A IN OUT Vcc DATA DATA GND 6 IC1 PIC12F675 GP1 VR1 10k 20 1 5 0V THESE PARTS USED ONLY FOR RELAY SWITCHED OUTPUTS K GND A CON1 16V ANT GND GND Vcc ANT 433MHz RX MODULE +12V IN 100 µF 1k Vcc A Fig.11: the circuit diagram for the receiver circuit. The 433MHz RX (receiver) module picks up the signal from the detector unit and feeds its data output to PIC microcontroller IC1. When a valid code is received, IC1 drives a piezo transducer & activates either LED1 or LED2 to indicate vehicle entry or exit. Alternatively, the LEDs & piezo transducer can be omitted and reed relays fitted instead. These can then be wired across the buttons of a remote control unit, eg, for a remote-controlled mains socket or a wireless doorbell. from 1-5s). This makes it quite distinct from the exit sound, since the tone now decreases instead of increasing. In addition, the entry LED (LED2) lights during the tone and again stays lit for 15s after the tone ceases. The non-directional tone is different yet again. In this case, there is a 1s 440Hz horn “bip” followed by a further 440Hz “bip” lasting from 1-5s. In addition, LED1 & LED2 both light and then flash alternately for 15s after the tone ceases. Relay version For the relay version, Relay1 is first switched on for 500ms, thereby allowing its closed contacts to activate the “On” button on a UHF remote control (eg, for a mains socket). Then, after a preset period ranging from 20s to five minutes as set by VR2, Relay2 is switched on for 500ms to activate the “Off” button on the UHF remote. Both relays are driven via 100Ω resistors, while diodes D1 & D2 clamp any switch-off voltage spikes produced by the relay coils. The 100Ω resistors are there to protect IC1’s GP0 & GP1 outputs. siliconchip.com.au The receiver PCB can be built in two versions (relay version shown here). See Pt.2 next month for the assembly details. The program in IC1 automatically detects if the piezo transducer has been installed or if the relays have been installed instead. It does this by first making the GP1 pin an input and then switching GP0 high. If GP1 goes high immediately after switching GP0 high, then the piezo transducer is connected. That’s because the piezo transducer’s capacitance allows the voltage transition to be coupled through to GP1. Conversely, if GP1 stays low, the software assumes that the relays are connected since Relay2’s coil provides a low resistance path to ground. Power for the circuit is derived from a 12V DC plugpack, with diode D1 providing reverse polarity protection. REG1 then provides a regulated 5V supply for IC1 and the UHF RX module. The 100µF input and output bypass capacitors provide supply line filtering, while a 100nF capacitor provides additional decoupling for the supply going to the microcontroller. That’s all for this month. Pt.2 next SC month has the assembly details. July 2015  35