This design aims to create a powerful alarm system that can promptly notify the parties involved in emergencies, even if they are not nearby, and automatically trigger the alarm. Compared to ordinary alarms, this design focuses on remote voice alarms. By connecting appropriate sensors, it can form a multi-functional alarm system including burglar alarms, fire alarms, and gas leak alarms, and can also implement some intelligent control functions, giving the alarm system more powerful and complete functions to meet people's needs for safety alarms. This system is based on the telephone network but is independent of the telephone and will not affect the normal use of the telephone. It is operated through voice prompts, making the human-computer interaction user-friendly. System Principle and System Block Diagram The system mainly includes an automatic telephone off-hook and on-hook circuit, a DTMF signal transceiver circuit, a voice prompt circuit, an alarm circuit, a keypad display circuit, a human body signal detection circuit, an encoding circuit, a wireless transmission circuit, and a single-chip microcomputer control circuit as the core. The system structure block diagram is shown in Figure 1. We set the alarm part as the main working part of this system, that is, real-time monitoring of the safety situation in the room. In the software, this is represented as a main loop, which is interrupted only when there is a ringing signal or a set signal to execute the corresponding operation. Here, we need a sensor that is sensitive to infrared radiation from the human body and resistant to interference (such as from small animals). Therefore, we selected a passive pyroelectric infrared detector and covered its radiation surface with a special Fresnel filter to significantly control environmental interference. The setting section stores the number to be dialed when an alarm is triggered and sets the owner's authentication password. For system simplicity, an LCD screen is used for display. Figure 1 shows the system structure block diagram. When the infrared human body detection circuit detects a human intrusion, the encoding circuit encodes the address of the detection probe and transmits it to the wireless receiving circuit via the wireless transmitting circuit. After CPU decoding, the LED displays the alarm address, and simultaneously issues an audible and visual alarm or dials a pre-set phone number to the owner. The signal detection circuit mainly consists of the pyroelectric infrared detection probe SD02 and the BISS0001 signal processing circuit. The signal detection circuit is shown in Figure 2. A pyroelectric infrared sensor, equipped with a filter lens and impedance matching field-effect transistors, detects infrared radiation from the human body non-contactly and converts it into an electrical signal. This signal is then pre-amplified by operational amplifier N1 in the BISS0001 and amplified in the second stage by operational amplifier N2, raising the DC potential to the built-in voltage Um before being sent to a bidirectional amplitude discriminator composed of comparators N4 and N5 to detect the valid trigger signal Us. Since the built-in voltages UH ≈ 0.7UDD and UL ≈ 0.3UDD, when UDD = 5V, ±1V noise interference can be effectively suppressed. N3 acts as a conditional comparator. When the input voltage Uc is less than the built-in voltage UR (≈ 0.2UDD), N3 outputs a low level, blocking Us from being passed to the next stage. When Uc > UR, N3 outputs a high level, opening AND gate N7. If the rising edge of the trigger signal Us arrives at this time, a delay timer can be started, and Uo outputs a high level. The selection of the comparator threshold is crucial; a threshold that is too low can easily lead to false alarms, while a threshold that is too high results in low sensitivity. Figure 2 shows the signal detection circuit. During the timing period Tx, output terminal 2 of BISS0001 is at a high potential, causing transistor VT1 to saturate and conduct, with its collector at a low potential. This signal is sent to the encoding and wireless transmission circuit, composed of a microcontroller and a wireless transmission circuit, connected to the P0.0 port of the microcontroller used for encoding. The microcontroller encodes the probe and transmits it wirelessly. At the end of Tx, BISS0001 enters the blocking period Ti, its output terminal becomes low, the transistor is cut off, and its collector becomes high. Pin 1 (terminal A) of BISS0001 is connected to the power supply, enabling the signal detection circuit to be repeatedly triggered. The timing interval Tx can be determined by the resistor and capacitor connected to pins 3 and 4 of BISS0001. The elevation angle of the signal detection probe can be adjusted within a 120° range, and the actual detection distance can be adjusted by changing the elevation angle. This can be adjusted through actual testing, or by adjusting the adjustable resistor RP in the signal detection circuit to adjust the probe's detection distance. The detection distance of this circuit design is 30m. The encoding circuit and wireless transceiver module consist of a microcontroller and the nRF401 transceiver chip, requiring few external components and resulting in a simple circuit. This design utilizes the newly released nRF401 wireless data transmission chip from Nordic Technologies, Norway, a true microcontroller-based UHF wireless transceiver chip designed for the 433MHz ISM band, employing FSK modulation and demodulation technology. With a high-gain antenna, the transmission distance can reach 3000m. The encoding and wireless transceiver circuits are shown in Figure 3. The P2.0 port of the AT89C51 controls the PWR_UP of the RF chip, setting it to "1" to indicate normal operation mode and "0" to indicate standby mode. P2.2 connects to the CS pin of the RF chip, controlling the transmit and receive frequencies; "1" indicates an operating frequency of 434.32MHz, and "0" indicates an operating frequency of 433.92MHz. P2.1 controls the TXEN pin of the RF chip, setting it to "1" to indicate transmit mode and "0" to indicate receive mode. Figure 3 shows the interface between the MT8880 wireless transceiver circuit, AT89C51, and voice circuit. The DTMF signal transmission/reception circuit uses the MT8880 chip from MITEL, a dedicated integrated circuit for processing DTNF signals. It not only has automatic dialing functions for receiving and transmitting DTMF signals but can also detect dial tones, ringback tones, and busy tones on the telephone trunk line. It is suitable for interfacing with microcontrollers, and the peripheral circuit is simple. The MT8880 has five internal registers: a receive data register, a transmit data register, transmit/receive control registers CRA and CRB, and a transmit/receive status register. In this design, only the transmit data register, transmit/receive control registers CRA and CRB are used to transmit DTMF signals to achieve the automatic dialing function. The data in the transmit data register determines the frequency of the dual-tone multi-frequency signal to be transmitted; therefore, data can only be written to the transmit data register. The two transmit/receive control registers occupy the same address; therefore, the value of the register select bit in CRA determines whether to operate on CRB. The ISD1420 voice chip employs direct analog storage technology, offering excellent recording and playback quality with a certain degree of reverberation. Its peripheral components are simple, requiring only basic resistors and capacitors to form a simple recording and playback circuit. It requires no backup power supply, has a long information storage time, and does not require a dedicated programmer or voice development tool. It possesses strong addressing capabilities, allowing the memory to be divided into 160 segments for management, resulting in a minimum recording and playback time of 125ms. In this design, because four alarm prompt voice segments are needed, each voice segment is set to 5 seconds, with starting addresses of 00000000B, 00101000B, 01010000B, and 01111000B, respectively. The ISD1402's data ports A3, A4, A5, and A6 are connected to the PB0, PB1, PB2, and PB3 ports of the 8255 microcontroller port expansion chip, respectively. A0, A1, A2, and A7 are grounded. PLAYL is connected to pin PB5 of the 8255. SP, via capacitor C14, couples the voice signal to the telephone interface circuit. When switch SB1 is pressed, the recording indicator LED illuminates and recording begins simultaneously. In case of an alarm, the microcontroller controls the DTMF signal transmitting/receiving circuit to automatically dial a telephone. After the call is connected, the microcontroller sends instructions to the ISD1420 to specify which recording segment to play and the playback command based on signals from different detectors. The ISD1420 then sends the voice signal to the telephone interface circuit. After playback is complete, the microcontroller sends a hang-up command, ending the alarm. The D0-D3 ports of the MT8880 are connected to the PA0-PA3 ports of the 8255, CLK2 to PA4, R/W to PA5, RSO to PA6, CS to PA7, and IRQ to the T0 port of the 89C51 main control circuit processor, used to record the number of various pulses. Signals from the voice circuit are sent to the telephone line via R44. Relay K is used to control on/off switching. The base of the transistor is connected to the P1.2 port of the 89C51 main control circuit processor. When P1.2 is "1", V2 conducts, relay K closes, and the telephone is connected; when P1.2 is "0", V2 is cut off, and the telephone is hung up. The P0 port of the 89C51 main control circuit processor is connected to the D0-D7 ports of the 8255 and the 74HC373 respectively. The Q0 and Q1 pins of the 74HC373 are connected to the A0 and A1 pins of the 8255 respectively. The P2.5, P2.6, and P2.7 pins of the 89C51 are connected to the A, B, and C ports of the 74HC138 respectively. The YO pin of the 74HC138 is connected to the CS pin of the 8255. Software Design 1: Signal Tone Recognition Method. After detecting an alarm signal, the system simulates off-hook. To identify whether the telephone system is in a dialable state after the simulated off-hook, whether the call is connected after dialing, and whether the other party has picked up the phone, the system must perform signal tone recognition. To recognize signal tones, the characteristics of various signal tones must be known. The characteristics of various signal tones are as follows: ● Dialing tone: 450±25Hz continuous buzzing tone. ● Busy tone: A 450 ± 25Hz buzzing tone that lasts 0.35s and then 0.35s, with a tone period of 0.7s. ● Ringback tone: A 450 ± 25Hz buzzing tone that lasts 4s and then 1s, with a tone period of 5s. These telephone signals are all analog signals. However, microcontrollers cannot recognize analog signals, so they must first be converted into pulse signals, and then identified based on the number of pulses. The formula for calculating the number of pulses for these telephone audio signals is N = tm/T. Where N is the number of pulses per tone period; T is the audio period of the telephone audio signal, in seconds; and tm is the conduction time of the signal tone period, in seconds. In practical use, it is mainly necessary to identify the dial tone, busy tone, and ringback tone. Analyzing the characteristics of these three signals, it can be seen that the number of pulses is different within a certain counting time. In this design, a 2s count is used to determine the dial tone, and 2.8s (i.e., 4 busy tone cycles) is used to determine whether it is a busy tone. Then, using ls as a counting unit, the cumulative pulse count after five counts is used to determine whether the other party has answered the phone. If yes, the corresponding alarm prompt voice is played; otherwise, ls is counted again, and the pulse count in the last 5 seconds is calculated to determine whether the other party has picked up the phone again. This process is repeated until no one answers the phone after the waiting time, at which point the call is hung up. Due to interference and other factors, there may be false alarms. Therefore, a method is adopted where at least one of the pre-set phone numbers is dialed only once. If all are dialed or no one answers, all pre-set phone numbers are redialed, thus making the probability of false alarms very low and negligible. 2. Software Flowchart and Dialing Program The flowchart of the automatic dialing program is shown in Figure 4. 3. Several points to note during programming First, the DTMF generator of the MT8880 is the main body of the transmitting section. It generates all 16 standard dual-tone signals with low distortion and high accuracy. These frequencies are all generated by frequency division of a 3.579545MHz crystal oscillator. The circuit consists of a digital frequency synthesizer, a row/column programmable frequency divider, and a switched-capacitor D/A converter. Row and column single-tone sine waves are mixed and filtered to generate dual-tone signals. Encoded data is written to the MT8880 transmit register via a DTMF codec table, generating separate fLOW and fHIGH signals. Encoding errors will cause dialing failure; therefore, extreme care must be taken during programming. Secondly, a delay should be made after the phone is picked up before checking for the off-hook tone, as this system uses a mechanical relay for automatic off-hook, and the relay's response time must be considered. Finally, after dialing one phone number, the next phone number should not be dialed immediately. The minimum effective hang-up time must be ensured to guarantee that the previous phone number has indeed been hung up; otherwise, there will be no dial tone when dialing the next number. Figure 4: Dialing Subroutine Flowchart Conclusion The system uses software for both encoding and decoding, and bidirectional transmission is employed. In case of an alarm, a data string is sent to the host, consisting of: a 4-bit address code, an 8-bit data code, a 1-bit start bit, and a 1-bit parity bit. After receiving data, the host first performs verification. If the data is incorrect, it requests a retransmission until the data is correct. Once correct data is received, the next step is to check the address code. If the address code matches the address set by the host, it indicates the data may have originated from a device outside the system, and decoding is not performed. Decoding only occurs when data matching the address set by the host is received.