Fullscreen HTML5Med Res (??MB)HTML4

Rolling Code Keyless Entry System Versatile IR Unit Also Functions As An Alarm Pt.1: By JOHN CLARKE Ideal for keyless entry for doors in cars, homes & industry, this Keyless Entry System features a rolling code to ensure high security. It also has two door-strike outputs, an alarm system & provision to use up to 16 separate keyfob transmitters with the same receiver. L OTS OF DIFFERENT electronic systems have now been developed for keyless entry. These include systems that require a coded electronic key, such as RF and infrared transmitters, RFIDs (Radio Frequency Identification Devices), keypads and swipe cards. There are also units that do not require a coded electronic key and these include fingerprint, face and iris recognition. Regardless of format, electronic keys 40  Silicon Chip usually comprise a small keyfob-style transmitter and a receiver that goes with the door lock mechanism. The transmitter sends a string of data that is unique to each individual lock and this data must match the data stored in the receiver before the lock will be released. The concept is roughly similar to a standard metal key which has a pattern of peaks and valleys along its length. These peaks and valleys must match the tumblers within the lock in order for the lock to open. With any type of lock, there is always a problem of security. Keys can easily be copied, while many conventional RF and infrared transmitters are far from tamper-proof. One technique is to use a special receiver to intercept and copy the transmitted code. Once copied, the signal can then be re-transmitted to the door lock to gain unauthorised entry. In fact, this technique was commonly used by car thieves in carparks and proved very effective against early electronic locking systems. It could also be used to open automatic garage doors and gain access to buildings. Rolling code security Modern transmitters now circumvent this problem by changing their code each time they send a signal. So if an unauthorised person captures siliconchip.com.au the transmitted code, re-sending this code will not unlock the door. This is because the door lock is now expecting a new code based on an algorithm that both the transmitter and receiver have in common. This code changing technique is commonly called a “rolling code”, although it is sometimes also called “code hopping”. It renders copying useless and thus provides a very high level of security. It is also virtually impossible to send a correct code without having a valid rolling code transmitter. This is because of the huge number of code variations possible. Because of its security advantages, a rolling code transmitter forms the basis of the Rolling Code Keyless Entry System described here. In fact, the odds of picking a correct code at random for our rolling code transmitter are one in 1.4 trillion or one in 1012. If you want to know more about rolling code transmissions, refer to the separate panel elsewhere in this article. Main features Our new Rolling Code Keyless Entry System comprises a small keyfob-style transmitter and a separate receiver. The transmitter is small enough to be attached to a keyring and has two pushbutton switches, each capable of sending a separate code. Each time one of the switches is pressed, a small indicator LED flashes to indicate that the transmitter has sent its code. The larger of the two switches activates the alarm functions of the receiver. It arms the unit so that it will sound an alarm should there be unauthorised access. The alarm functions include an electric door strike control (this allows the door to be opened), two alarm inputs (eg, to monitor doors, windows or other sensors), and an arm/disarm output. The door strike can optionally be set to operate on arming, on disarming or on both. In addition, an alarm output is provided to sound a siren if required. The second, smaller pushbutton switch on the transmitter is independent of the alarm. It can be used to operate a separate door strike or some other device connected to the receiver. Such devices can include a light or a siren that can be used as a panic alarm. This can be optionally set to operate momentarily or can be toggled on and siliconchip.com.au Features & Specifications Transmitter • • • • • • Rolling code infrared transmission Small keyfob style case Dual function buttons Randomisation of code parameters feature Synchronising of parameters feature Up to 16 identifications Receiver • • • • • • • • • • • • • • 12V operation Up to 16 separate transmitters can be synchronised Dual function with an independent output Two alarm inputs with exit and entry delays Two door strike outputs Alarm output Arm/disarm output and LED indicator IR receive acknowledge LED Strike 1 operates on arm, disarm or both Strike 2 operates independently with momentary operation or toggle output Arm output invert option Adjustable door strike, entry/exit delay and alarm periods 200-code look ahead feature Transmitter lockout feature off with each switch pressing. The door strike outputs can be set to operate from between 0-64 seconds, while the inputs can include delayed operation from 0-64 seconds. These delayed inputs allow the alarm to be armed while giving the user enough time to exit the door without setting off the alarm. An identical delay period allows the alarm to be disarmed on entry. During the exit delay period, the ARM indicator LED in the receiver unit flashes on and off at a 1-second rate. At the end of the exit delay, this LED indicates that the unit is armed by flashing briefly once every second. This conserves power and increases its effectiveness when it comes to attracting attention. An Acknowledge/Power LED is also included in the receiver. This normally flashes with a very short duty cycle. However, when the receiver picks up a signal from the transmitter, the Ack/ Power LED flashes at a very high rate. It also shows if the received code is invalid by momentarily blinking off and on. If the code is correct, the receiver responds to the signal. The transmission range is about 4m which should be sufficient for most purposes. Note, however, that it will not work if the IR receiver is in direct sunlight. Setting it up Before using the Infrared Rolling Code Alarm, both the transmitter and the receiver must be set up correctly. First, each transmitter must be given a separate identity ranging from 1-16. This is selected using link options on the transmitter board but note that no two transmitters should be given the same identity. Second, the transmitter must be randomised. This changes the initial rolling code and algorithm parameters to ensure that the transmitter code is going to be unique. The third step involves synchronising the transmitter and receiver. This process involves sending the rolling October 2007  41 Parts List Receiver 1 PC board, code 01510071, 61 x 122mm 1 UB3 plastic utility box, 130 x 68 x 44mm 5 2-way PC-mount screw terminal blocks (5mm or 5.08mm pin spacing) 1 SPST vertical mount micro tactile switch with 0.7mm actuator (S1) 3 3-way pin header terminal strips (2.54mm spacing) 4 2.54mm jumper shunts 3 PC stakes 1 25mm length of 0.8mm tinned copper wire 2 10kW horizontal trimpots (code 103) (VR1,VR2) Semiconductors 1 PIC16F88-I/P microcontroller programmed with irrcroll.hex (IC1) 1 78L05 low-power 5V regulator (REG1) 1 38kHz infrared receiver (IRD1) 2 BD681 Darlington NPN transistors (Q1,Q2) 2 BC337 NPN transistors (Q3,Q4) 1 16V 1W zener diode (ZD1) 4 1N4004 1A diodes (D1-D4) 1 1N5404 3A diode (D5) 2 3mm red LEDs (LED1,LED2) Capacitors 5 100mF 16V PC electrolytic 3 100nF MKT polyester 3 10nF MKT polyester 1 1nF MKT polyester Resistors (0.25W, 1%) 4 10kW 1 220W 2 2.2kW 2 100W 2 1kW 1 10W 2 680W code parameters to the receiver, as described next month. You can synchronise from 1-16 transmitters, provided each has a different identity. Also included is a facility to prevent any or all transmitters from operating the receiver once they have been synchronised. This “lockout” feature can be useful if a transmitter has been lost and you no longer want it to work with your alarm system. 42  Silicon Chip Test Components 4 red LEDs 4 2.2kW 0.25W 1% resistors Transmitter 1 PC board, code 01510072, measuring 30 x 36mm 1 keyfob remote control case (Jaycar HB-5605 or equivalent) 1 12V A23 car alarm battery (9.5 diameter x 27mm) 2 SPST SMD tactile switches 6 x 6 x 3.85mm (S1,S2) 1 TO-3P transistor silicone insulating washer cut to 20 x 24mm 5 PC stakes 1 25mm length of 0.8mm tinned copper wire Semiconductors 1 PIC16F628A-20/SO 18-lead SOIC microcontroller programmed with irxmroll.hex (IC1) 1 MC78M05 DPAK 5V regulator (REG1) 1 MMBT100 SOT-23 SMD NPN transistor (Q1) 1 MMBT200 SOT-23 SMD PNP transistor (Q2) 2 1N4148 diodes (D1,D2) 1 3mm infrared emitting LED (LED1) 1 green gull wing style surface mount LED (2.2 x 2.2mm) (LED2) Capacitors 2 1mF monolithic ceramic 1 100nF monolithic ceramic Resistors (0.25W, 1%) 2 10kW 2 22W 3 1kW A transmitter identity can be locked out individually but if you don’t know the identity of a lost transmitter, all identities can be locked out. The transmitters that are to be used with the receiver are then re-synchronised. Transmitter circuit OK, so much for the background details. Let’s now take a look at how the circuit works, starting with the transmitter – see Fig.1. IC1, a PIC16F628A microcontroller, forms the heart of the transmitter circuit. The circuit might look quite simple but there are a lot of “smarts” hidden inside the PIC micro, including the software necessary to generate the rolling code. Under normal conditions, switches S1 & S2 are open circuit and transistor Q2 is off, so no power is applied. This is done to ensure long battery life. If power were continuously applied, the current drawn from the battery would be around 4mA because of the quiescent current of the 5V regulator. Conversely, pressing either S1 or S2 connects the 12V battery to the input of regulator REG1 via diode D1 or D2. A 22W resistor is included in series between the battery and the switches to limit the initial charging current into the 1mF bypass capacitor at REG1’s input. This minimises wear on the switch contacts. When power is applied to REG1’s input, its output delivers a regulated +5V rail to IC1. As a result, the micro powers up and runs its internal software program. One of the first things the program does is check which switch was pressed (this happens after a short delay to make sure the switch is fully closed). In operation, the program can decide if S1 or S2 is pressed because of the 10kW resistor connected between S2 and the micro’s RA4 input. It works like this. Initially, RA4 is set low by the program. This pin is then made open circuit so that it can be pulled high if switch S2 was closed. However, if S1 was closed instead, the RA4 pin will stay at 0V. By checking the voltage on RA4, the program can thus determine which switch was pressed and initiate the correct function codes for that switch. The 10kW resistor is necessary to limit the current into the internal clamping diodes at RA4 when S2 is closed. In practice, the positive clamp diode will conduct, clamping the RA4 input to 0.6V above the +5V supply. This protects the input from damage. Diodes D1 & D2 protect the regulator from reverse polarity should the battery be inserted the wrong way around. These diodes also isolate the switch outputs from each other, so that the RA4 input will only go high if S2 is pressed. If S1 is pressed, the 12V at REG1’s input reverse biases D2 and so siliconchip.com.au Fig.1: a PIC16F628A microcontroller forms the heart of the transmitter circuit. It contains all the software necessary to generate the rolling code and drives an infrared LED (LED1). is blocked from reaching RA4. Next, the program sets RA2 at pin 1 of the micro high. This output drives the base of NPN transistor Q1 via a 10kW resistor. As a result, Q1 switches on and this in turn switches on transistor Q2. This action latches the supply to regulator REG1, even if switch S1 or S2 is released. This is necessary to allow time for the rolling code calculations to be made and stored without interruption, otherwise the code may become corrupted. It also ensures that the rolling code is transmitted in its entirety. The next stage in the program involves calculating the code and storing the values. This calculation is based on the previously transmitted code and uses an internal algorithm. Once calculated, the new code appears at outputs RB0-RB5 which in turn drive an infrared LED (LED1). The 22W resistor in series with LED1 limits the current to a safe value. In operation, LED1 is driven using 100mA pulses at a rate of 38.46kHz. A high (or a “1”) is transmitted as a 512ms-long burst of 38.46kHz signal, followed by 512ms of no transmission. Conversely, a low (or a “0”) consists siliconchip.com.au of a 512ms period of no transmission followed by a 512ms burst of 38.46kHz signal. LED2 is the Transmit LED and is driven by output RA3 during code transmission. Basically, RA3 goes high each time there is a “1” in the transmitted code, and low each time there is a “0”. As a result, LED2 flashes to mimic the transmission code. Transmitter identity Transmitter identity is selected using the LK1-LK4 link connections to RA1, RA0, RA7 & RA6. As shown, each individual input can be connected to either the +5V supply or the ground supply (0V) but not to both or the supply will be shorted. The number of possible combinations is 16. Each of these inputs is initially tied to +5V on the PC board (via thin PC tracks) and this selection is Identity 1. The other 15 identities are selected by breaking one or more of these connections to the +5V rail and connecting them instead to an adjacent 0V rail. We’ll talk more about this in the construction. In-circuit programming Five-pin header CON1 is provided on the circuit to allow for In-Circuit Serial Programming (ICSP) of IC1 using a PIC programmer. Alternatively, we have developed a surface-mount converter board that will allow IC1 to be programmed directly using a PIC programmer. We’ll publish the details on this next month. The ICSP connections on the transmitter are also used to run the randomisation and synchronisation functions using a bridge between pins 3 & 5 and 3 & 4 respectively. IC1 runs at a nominal 4MHz, as provided by an internal oscillator. This oscillator has a 1% tolerance and its accuracy is sufficient for this application (ie, there’s no need for a crystal oscillator). However, because the oscillator frequency can vary with temperature, we have included a means for the receiver to lock onto the transmitter’s clock rate, so that variations over a long time period do not matter. By the way, the transmitter uses several surface-mount components so that the circuit will fit into a small keyfob case. These surface-mount parts include IC1, REG1, Q1, Q2, LED2, S1 & S2. The remaining parts are standard through-hole component types that are October 2007  43 Standby current: 0mA Total transmit current: rolling code transmission = 35mA for 80ms; synchronise = 35mA for 100ms; randomisation = 10mA. Infrared transmit frequency: 38.46kHz Code transmission rate: 1.024ms Encoding: a high (or a 1 bit) is transmitted as a 512ms burst of 38.46kHz infrared signal followed by 512ms of no transmission. A low (or 0 bit) is transmitted by a 512ms period of no transmission followed by a 512ms burst of 38.46kHz infrared signal. Rolling code: sends four start bits, an 8-bit identifier, the 48-bit code plus four stop bits. The start bits include a 16.4ms gap between the second start bit and the third start bit. Synchronise code: sent as two blocks. Block 1 sends four start bits, the 8-bit identifier, a 32-bit seed code and four stop bits. Block 2 sends four start bits, a 24-bit multiplier, the 8-bit increment and 8-bit scramble values, and four stop bits. The start bits include a 16.4ms gap between the second start bit and the third start bit. Code randomisation: alters the multiplier values, the increment value, the scramble value and the seed code at a 40ms rate. Infrared transmission range: 4m inputs – Input 1 & Input 2 – and these connect to the RB5 and RB6 inputs of IC1 via 2.2kW resistors. Each input is also bypassed using a 100nF capacitor to filter out transients and thus prevent false triggering of the alarm. When these inputs are open, both RB5 and RB6 are held high (ie, at +5V) via internal pull-up resistors. In practice, this means that you can use normally-open (NO) or normallyclosed reed switch and magnet assemblies to trigger the inputs. If you use an NO switch, the input will normally be high and the system will trigger if a switch is closed. Conversely, if an NC switch is used, the input will normally be pulled low but will go high if the switch is opened. Basically, any change in level when a reed switch opens or closes will be detected and sound the alarm at the end of the entry period – provided that the receiver is in its armed state. Note, however, that the alarm will not sound if the receiver is still within its exit delay period. Receiver Door strike outputs Specifications Transmitter Supply Current: 7.6mA typical when armed and with no external devices powered. Strike 1 period: adjustable from 0-64 seconds in 0.25s steps approximately. Strike 2 period: adjustable from 0-64 seconds in 0.25s steps approximately. Input 1 delay: adjustable from 0-64 seconds in 0.25s steps approximately for exit and entry delays. Input 2 delay: adjustable from 0-64 seconds in 0.25s steps approximately for exit and entry delays. Alarm period: adjustable from 0-128 seconds in 0.50s steps approximately small enough to fit onto the PC board. Receiver circuit Refer now to Fig.2 which shows the receiver circuit. It’s built around infrared receiver IRD1 and PIC microcontroller IC1, the latter operating at 4MHz to match the transmitter’s frequency. Once again, much of the complexity is hidden by the software programmed into the microcontroller. IRD1 only has three leads but inside it comprises a complete infrared detector and processor. First, it receives the 38kHz infrared pulse signal from the transmitter and amplifies this to a constant level. This signal is then fed 44  Silicon Chip to a 38kHz bandpass filter to remove any 50Hz or 100Hz mains signal and other noise. It then demodulates the signal to produce a serial data burst at IRD1’s pin 1 output. This serial data signal from IRD1 is fed to the RB4 input of IC1 via a 100W resistor. A 1nF capacitor filters out any transients. IRD1 is powered from the receiver’s +5V supply rail. A 100W resistor and a 100mF capacitor provide supply decoupling and filtering, to prevent the receiver from producing false signals due to power line changes. As well as the IR receiver, there are two other inputs to the PIC microcontroller. These are alarm sensor When an IR signal transmission is received, the output from IRD1 is processed by IC1. This then drives Darlington transistors Q1 & Q2 as appropriate to control the door strike outputs (ie, Strike1 & Strike2). As shown, Q1 & Q2 are driven via 680W resistors from IC1’s RB0 and RA2 outputs respectively. Diodes D1 & D2 clamp the voltage produced by the door strike solenoid to the supply rail when the transistor is switched off. Transistors Q1 & Q2 are both BD681 Darlington types and can be used to drive loads up to 1.5A. A typical electric door strike only draws about 800mA at 12V. The other two outputs are the Alarm & Arm outputs and these are controlled by transistors Q4 & Q3 (both BC337) respectively. Q4 is driven by IC1’s RB1 output via a 220W current limiting resistor. However, the base current is sufficient for the transistor to remain fully saturated for a 200mA load and this is ideal for many piezo sirens. Similarly, transistor Q3 is driven via a 10kW resistor from IC1’s RB2 output. Q3’s collector provides the Arm output and this can be used as a toggle output to set a second alarm system. Typically, you would use a 1kW pull-up resistor between the Arm output and the +12V rail, so that the siliconchip.com.au D1–D4: 1N4004 D5 1N5404 +12V IN INPUT 1 IN K ZD1 16V 1W 0V REG1 78L05 10 K A +5V OUT +11.4V VR1 10k GND 100 F 16V A 100 F 11 RB5 100nF AN1 2.2k 12 RB6 RB2 100nF INPUT 2 RB1 100 RB0 100 F IRD1 IC1 PIC16F8810 I/P RB4 3  1 100 1nF 2 IRD1 +5V LK1 – + LK2 – + LK3 – 3 1 2 AN0 10nF 10nF + 10k 10k 10k 10nF RA2 RA3 4 15 RA7 RB3 RA5 TPG K D4 K D1 K D2 A A A RB7 100 F TP2 ARM 17 ALARM 18 10k* 8 Q3 E B 220 7 STRIKE 1 C B Q4 E 680 6 STRIKE 2 C 1 LEDS C B 680 Q1 C E B Q2 1k 3 1k 2 +5V 9 13 GND IN + – Vss 5 LK4 ARM INVERT A ACK/ POWER A  LED1 K K INFRARED ROLLING CODE RECEIVER Q3, Q4: BC337 E B C K ZD1 A OUT ARM  LED2 S1 A SC K 78L05 D1–D5 2007 A E SYNCHRONISE RA6 +11.4V * REDUCE TO 1k IF ARM OUTPUT USED TO DRIVE A RELAY RA4 16 D3 A TP1 14 Vdd K 100 F 100nF 2.2k VR2 10k Q1, Q2: BD681 B C C K E Fig.2: infrared receiver IRD1 and PIC microcontroller IC1 are the main parts in the receiver. IRD1 picks up and demodulates the infrared transmissions, while IC1 decodes the data and drives the various outputs. level can swing between 0V and 12V. Alternatively, Q3’s collector could be used to drive a relay coil. In this case, the 10kW base resistor will need to be reduced to 1kW so that the transistor can remain in saturation while driving a 285W 12V relay coil. The unit can be optionally configured with Q3 either on or off when armed. This is set using link LK4. When LK4 is in the “+” position, Q3 is on when the unit is armed and off when disarmed. In this case, the RB3 input is held at +5V via an internal pull-up resistor within IC1. Moving LK4 to the “-” position pulls RB3 to ground and changes the sense of the Arm output. In this case, Q3 is off when the unit is armed and on when disarmed. siliconchip.com.au LED 2 indicates the state of the unit. It’s driven from the RA4 output of IC1 via a 1kW resistor and flashes when the unit is armed. There are two different flash styles. During the entry and exit delay periods, the LED flashes with a 50% duty cycle (ie, it is on for half the time and off for half the time). However, at the end of the delay period, it flashes on for only 4% of the duty cycle (ie, each flash is very brief). Other link options Links LK1, LK2 & LK3 are included to provide further options. For example, LK1 can be tied to either the +5V rail or to 0V, or it can be left open. These three options determine how the Strike1 output operates. Basically, Strike1 can be set to operate when the unit is armed, when it is disarmed or on both arming and disarming. In operation, the software programmed into the PIC micro decides where the link is inserted by running a few tests. First, it takes the RA7 output high (5V) and then sets the RA7 pin as an input to read the voltage. If the voltage is now low, then the link must be in the “-” position. However, if the input remains high, then the link is either in the “+” position or is open circuit (it remains high when the link is open because of the charge on the associated 10nF capacitor to ground). To test if the link is in the “+” position or open, the RA7 pin is made an output again and is driven low (to 0V). The RA7 pin is then changed October 2007  45 Note: transistor Q2 mounts under 10kW resistor These two larger-than-life-size photos clearly show how the parts are mounted on the transmitter board. You will need a fine-tipped soldering iron (2mm diameter or less) and a magnifying glass to do the assembly. Fig.3: follow these parts layout diagrams to build the transmitter board. Note that you have to set the transmitter’s identity before installing IC1 (see text) and don’t forget transistor Q2 – it goes under a 10kW resistor, just below S2. to an input and the level checked again. If the voltage is now high, then the link must be in the “+” position. Conversely, if the voltage is low, then the link is open. The 10kW resistor in series with RA7 is there to prevent shorting when this pin is taken high and low with a link in position. LK2 sets Strike2’s operation for either momentary operation or for Capacitor Codes Value mF Code IEC Code EIA Code 100nF 0.1mF 100n 104 10nF .01mF   10n 103 1nF .001mF   1n0 102 toggle operation. This link pulls the RA5 input either to +5V when it is in the “+” position (momentary) or to 0V when it is in the “-” position (toggle). Note that this link cannot be left open because the RA5 pin can only be used as an input. LK3 is used in conjunction with trimpots VR1 and VR2 to set the various time periods. These include the Strike1 and Strike2 momentary on periods, the entry and exit delays for Input1 and Input2, and the alarm period. As shown on Fig.2, trimpots VR1 & VR2 are connected across the 5V supply and their wipers connect to analog inputs AN0 and AN1 respectively. The voltage applied to each analog input is converted to a digital value within Fig.4: this enlarged track section shows the locations of links LK1LK4 on the transmitter board. The transmitter identity is changed by breaking one or more of the thinned link connections to the +5V rail and bridging them (with solder) to the adjacent 0V rail instead. the software and it is these values that determine the timeout periods. Synchronise switch Switch S1 is the Synchronise switch and this connects to the RB7 input. This input is normally held high via an internal pull-up resistor but when S1 is closed, it pulls RB7 to 0V. Basically, S1 is used to synchronise the receiver with the transmitter. It is also used when setting the time periods. In addition, if S1 is closed during power-up, it selects the transmitter identity lockout function. Power supply Power for the circuit is from a 12V supply such as a battery or DC plugpack. Diode D5 provides reverse Resistor Colour Codes (Transmitter) o o o o No.   2   3   2 46  Silicon Chip Value 10kW 1kW 22W 4-Band Code (1%) brown black orange brown brown black red brown red red black brown 5-Band Code (1%) brown black black red brown brown black black brown brown red red black gold brown siliconchip.com.au Table 1: Transmitter Identity Identity LK1 LK2 LK3 LK4 1 2 3 + + + + + + + + - + + + + + + + + - + + + + + + + + - + + + + + + 4 5 6 7 8 9 10 11 12 13 14 15 16 + + - polarity protection and is rated at 3A so that it can handle the currents that may be drawn by an electric door strike and siren. The 10W resistor and zener diode ZD1 provide transient protection, with the zener clamping voltages over 16V. The 10W resistor limits the current through ZD1 to a safe level. Following ZD1, the supply is filtered using a 100mF capacitor and applied to 3-terminal regulator REG1. The resulting regulated +5V rail is then used to power IC1 and the infrared receiver (IRD1). Power on/off indication is provided by LED1 which also acknowledges the infrared signal. Normally, LED1 flashes with a 4% duty cycle about twice per second. However, when an infrared signal is received, it flashes at the infrared reception rate. LED1 also flashes with an even duty cycle for a short time at the end of synchronisation and if the infrared signal is incorrect. Construction The Infrared Keyless Entry Alarm is built on two PC boards: a receiver board coded 01510071 and a transmitter board coded 01510072. We’ll start with the transmitter assembly which is the trickier of the two. In order to fit in the keyfob case, the transmitter board measures just 30 x 36mm and uses lots of surface-mount components. However, these are not too difficult to solder in, provided you have a soldering iron tip that is just siliconchip.com.au The keyfob case is supplied with the key switch covers mounted as shown here. This assembly must be removed. Fig.5: once the switch covers have been removed, the flanges are ground down using 180-grit sandpaper, so that only the tops remain (see text). 2mm in diameter or finer. A magnifying glass (or, preferably, a “maggie lamp”) is also required to check your soldering, while a length of 1.5mm de-soldering braid (solder wick) would also be useful for cleaning up any excess solder that may flow between connections. Fig.3 shows the parts layout on the PC board. The first step is to check the PC board carefully for any breaks in the copper or shorts between tracks. Repair any faults that you do find (rare these days), then check the shape of the board. It should have a curved front edge and a small circular cut out at the other end. In addition, there should be two slots for the battery clips. Next, check that the PC board fits neatly into the base of the keyfob case. If it does not fit, it’s just a matter of filing it neatly along the edges until it does. Setting the identity Before mounting any of the parts, it’s first necessary to set the transmitter’s identity but only if more than one transmitter is to be used. If more than one transmitter is used, then each will require a unique identity. As supplied, the PC board initially Above: the finished transmitter board inside its keyfob-style plastic case. Power comes from a 12V A23 car alarm battery. Note how the keyswitch covers are mounted on the lid, using a 20 x 24mm silicone washer – see text. Right: the two keyswitch covers are attached to the 20x 24mm silicone washer as shown here. Use silicone sealant to “glue” them in place. The keyfob lid can be used as a template to position them correctly. October 2007  47 Fig.6: follow this diagram to mount the parts on the receiver board. Use a socket for the PIC microcontroller and take care to ensure that all polarised parts are correctly orientated. The infrared receiver module (IRD1) can either be mounted on the board (as in the prototype) or connected via shielded cable (see diagram next month). ties links LK1-LK4 to the +5V supply rail. This is Identity1, or ID1. If only one transmitter is to be used, then you don’t have to do a thing – just leave it at the default identity (ID1). If you do wish to change the identity, it’s just a matter of altering one or more of the links as shown in Table 1. You do that by breaking the link’s thinned connection to the +5V track and connecting it to the adjacent 0V track instead via a small solder bridge. Make sure, however, that a link connection is not made to both the +5V and 0V tracks. We have labelled the +5V connection with a plus (+) sign and the 0V connection with a minus (-) sign. It is important to select the identity now because the +5V track section cannot be accessed when IC1 is in place. The +5V connections should only be broken with a sharp craft knife and, once broken, should not be resoldered. That’s because IC1 would no longer sit properly on the board, making it difficult to solder its pins. The selected identity should be marked on the back of the PC board using a marker pen. For example, if the identity is 2, write ID2 on the PC board. This number can also be written on the back of the keyfob transmitter case, in the indentation provided. If you are building the unit from a kit, then IC1 will be supplied preprogrammed. If not, you will have to program the PIC yourself using a suitable programmer. As previously mentioned, we have provided two programming options, the first of which is to use the in-circuit programming connector on the PC board. Alternatively, you can build and use the surface mount adaptor board to be described next month, so that IC1 can be programmed out of circuit. Parts assembly Except for a single wire link, all parts for the transmitter mount on the copper side of the PC board. Don’t install the link yet though – that step comes after you install IC1. To install IC1, position it on the board with its pin 1 at top right – see Fig.3 (pin 1 is indicated by a small adjacent dot in the body of the IC). Carefully adjust it so that its pins line up with the tracks and use a clothes peg (or some other small spring-clamp) to hold it in position. That done, solder a couple of diagonally opposite pins, check that everything is correct, then remove the peg and carefully solder the remaining pins. The main thing to watch out for here is unwanted solder bridges between adjacent tracks. If this does happen, use some solder wick to draw up the excess solder to clear the short. A magnifying glass will be handy here to inspect your work. Note that pins 6-9 and 10-11 are connected together anyway, so solder between these pins is OK. Once the IC is in, you can install the link beneath it on the other side of the board. This link must sit flat against the board, otherwise the board will not sit down in the case correctly. The remaining surface mount components – Q1, Q2, LED2 & REG1 – can Resistor Colour Codes (Receiver) o o o o o o o o No.   4   2   2   2   1   2   1 48  Silicon Chip Value 10kW 2.2kW 1kW 680W 220W 100W 10W 4-Band Code (1%) brown black orange brown red red red brown brown black red brown blue grey brown brown red red brown brown brown black brown brown brown black black brown 5-Band Code (1%) brown black black red brown red red black brown brown brown black black brown brown blue grey black black brown red red black black brown brown black black black brown brown black black gold brown siliconchip.com.au The assembled PC board clips neatly into a standard plastic utility case. The full installation and setting up details for the receiver will be in Pt.2 next month. now be soldered in place. Transistor Q1 has an N1 label on its top, while Q2 has an N2 label instead. These numbers relate to the MMBT100 and MMBT200 types respectively. Don’t get them mixed up. Installing the semiconductors The orientation of the two transistors is obvious – they have one pin on one side of the body and two on the other side. REG1 has a GND tab plus IN and OUT pins that must be soldered to the PC board. The central pin between the IN and OUT pins is left unconnected. Be careful with the orientation of LED2 – its cathode lead is the longer of the two. Next, solder in switches S1 & S2, then install five PC stakes for the ICSP header. These pins are inserted from the non-copper side of the PC board and soldered in position. The pins are then trimmed on the copper side to 3mm in height. On the underside, they are trimmed and filed to 0.5mm. The standard components can now be installed. These must be mounted flat against the PC board or as close to it as possible in the case of the 10kW resistor that straddles Q2. Take care with the orientation of diodes D1 & D2 and note that the tops of the three monolithic capacitors must be no more than 4mm above the PC board. In particular, the two capacitors near REG1 can be laid over at about 45°, while the one adjacent to IC1 needs to have its leads adjusted so it can be pushed down onto the PC board siliconchip.com.au far enough to meet the 4mm height requirement. Cut all the leads beneath the PC board (ie, on the non-copper side) flush with the surface. LED1 can go in next. Its anode lead is the longer of the two (unlike LED2) and this lead must go towards IC1. To mount it, first bend its leads down by 90° exactly 2mm from its body, then insert the leads into the PC board. Finally, push the LED all the way down onto the PC board, solder the leads and cut them flush with the underside. Note that a small circular notch is required in the rim of the keyfob base for the LED to sit in. This can be made using a small rat-tail file. When this notch has been made, file a matching notch in the top half of the keyfob case. Battery terminals The battery terminals are installed by first placing the PC board in the base of the case. That done, the terminals are slid into position and soldered. Make sure that the terminal with the spring is located as shown in Fig.3. Switch cover modifications The key switch covers that are supplied with the keyfob case have to be modified to suit the two switches on the PC board. As supplied, the two switch covers are already secured in place in the keyfob lid. This assembly must be removed and the covers carefully ground down to 1.5mm thick – see Fig.5. This is done by placing some 180-grit sandpaper onto a flat bench and sanding the switch covers until they are flat on their base. That done, cut out a 20 x 24mm rectangular piece from a silicone TO-3 washer (20 x 24mm) to make a new switch cover assembly. It’s then simply a matter of attaching the switch covers to this washer using silicone sealant – see photo. Use the keyfob lid as a template to position the covers correctly. Receiver assembly Now for the receiver. As usual, start by checking the PC board for any defects. Check also that the hole sizes for the screw terminal blocks are correct and enlarge them if necessary. That done, check that the PC board fits inside the specified utility case. File the board edges to get it to fit if necessary but don’t file them too much, otherwise the board will not lock correctly into the wall slots. Fig.6 shows the assembly details. Install the wire link first, then install the resistors. The accompanying table shows the resistor colour codes but you should also check them using a digital multimeter. The diodes and the IC socket can go in next, taking care to orient each with the correct polarity. Follow these with the capacitors, again making sure that the electrolytics go in correctly. The three PC stakes for TP1, TP2 & TPG can then be installed. Depending on your requirements, LEDs 1 & 2 can either be mounted directly on the PC board or mounted October 2007  49 Frequently Asked Questions One question that’s often asked about rolling code systems is what happens if the transmitter is out of range and one of the transmit switches is pressed? Will the receiver still work when the transmitter is later brought within range and the button pressed again? This question is asked because the code the receiver was expecting has already been sent and the transmitter has rolled over to a new code. So how does the system get around this problem? The answer to this is that the receiver will acknowledge a signal that is the correct length and data rate but it will not trigger unless it receives the correct code. So if the signal format is correct but the code is incorrect, the receiver then calculates the next code that it would expect and checks this against the received code. If the code is now correct the receiver will unlock the door. If the code is still incorrect, the receiver calculates the next expected code and will do this up to 200 times. If none of these are correct, the receiver keeps its original code but it will not trigger. In fact, the only way to trigger the receiver after this is to re-synchronise it to the transmitter. Of course, a second transmitter will still operate the receiver (provided they have been synchronised in the first place). That’s because this transmitter has a different identity and a different code to the other transmitter. Automatic synchronisation Some rolling code transmitters systems offer automatic synchronisation if the transmitter and receiver lose sync. In these systems, the receiver includes a code “look-ahead” feature as described above but the number of look-ahead codes is usually limited to fewer than 200. What happens is that if the code is not recognised after all the look-ahead calculations externally and connected using wire leads. Be sure to mount each LED with its cathode lead (the shorter of the two) towards the lower edge of the PC board. Similarly, IRD1 can either be mounted directly on the PC board or connected using twin-core shielded cable (see diagram in Pt.2 next month). Trimpots VR1 & VR2 and the 3-way and 2-way pin headers for LK1-LK4 50  Silicon Chip have been made, the receiver changes its synchronisation method. Basically, the receiver requires two separate transmission codes before restoring correct operation. On the first transmission, it calculates the next code it should receive using this received code as the basis for calculation. If the second code sent by the transmitter is the same as the code that was calculated, the receiver operates. The drawback of this latter scheme is somewhat less security since, in theory, two successive transmission codes could be intercepted and recorded. These codes could then be re-transmitted to synchronise and thus trigger the receiver. Calculating the code Another question that’s often asked is how does the receiver know which code to expect from the transmitter, since this changes each time? The answer to this is that the transmitter and the receiver both use the same calculation to determine the next code. They also both use the same variables in the calculation and these variables tend to be unique values that no other transmitter uses. For example, if the calculation for consecutive codes requires the original calculated code to be multiplied by 100 and the number 7 added to it, then both the transmitter and receiver will use these numbers to perform the calculation. Without knowing both the multiplier and the increment value, it would be very difficult to predict the next code. This is particularly true because of the very large numbers involved. The values quoted for the multiplier and increment value are not as simple as 100 and 7 but are 24 bits and eight bits respectively in length. In addition, the code length is 48 bits with as many as 2.8 x 1014 combinations. This reduces by a factor of 200 because of the look ahead feature to a 1 in 1.4 x 1012 are next on the list. That done, install REG1 and transistors Q1-Q4. Q1 & Q2 must be installed with their metal faces towards IC1. Finally, complete the board assembly by installing switch S1 and the screw terminal blocks. Note that the 6-way terminals at the righthand edge of the PC board are made up using three 2-way blocks. These are joined by sliding their dovetail joints chance of striking the correct code – still impossibly long odds. Code scrambling A further complication with the transmitted code is that the code is not necessarily sent in sequence. There are also 32 possible scrambling variations that can be applied to the code. What if the transmitter sends two consecutive codes that are the same and the code is intercepted and re-transmitted to open the lock? This is highly improbable and our rolling code transmitter has safeguards to prevent the same code appearing twice in succession. For each code calculation, a comparison is made between the current and last code. If the code is the same, the code is recalculated after an increment of the code value to ensure successive code calculations diverge. It is this new code that is transmitted. The receiver performs the same recalculation so that the new code will be accepted. Another question concerns the use of different transmitters. Does each transmitter use the same rolling code calculation and if so, wouldn’t the receiver lose its synchronisation if several transmitters were used? The answer is that the receiver will not lose synchronisation, even if one of the transmitters is not generally used. This is because each transmitter operates independently from the others. Only 16 transmitters can be used with a given receiver and each must have its own different identity from 1-16. The identity is built into each transmitter and synchronisation is required for each transmitter. The codes sent by each transmitter are different and the code includes the transmitter identity value. The receiver has 16 different rolling code and calculation parameters, and so each transmitter is treated independently. together before installing them on the PC board. That’s all we have space for this month. Next month, we’ll complete the construction and describe the installation and setting-up procedures, including setting the entry and exit delays. We’ll also describe the optional SOIC adapator board, so that you can program the PIC microcontroller out SC of circuit. siliconchip.com.au