BR1560/D DSP56F803/5 56F803 DSP56803/5 DSP56F803 DSP56F80X DSP56F805 RS-232 - Datasheet Archive
Preliminary Information . APPLICATIONS: Steppers and Encoders Home Appliance Controls Integrated with Voice Control Smart
BR1560/D BR1560/D Preliminary Information . APPLICATIONS: Steppers and Encoders Home Appliance Controls Integrated with Voice Control Smart Appliances Home Security Digital Telephone Answering Machine Engine Management Power Line Modem Servo Drives Automotive Control Application Brief Personal ECG monitor using DSP56F803/5 DSP56F803/5 Personal electrocardiogram (ECG) monitors are devices capable to rescue human life. However, their high price obstructs expansion. A low cost, single chip solution with DSP Motorola 56F803 56F803 can change that obstruction. The proposed solution, shown in Figure1, describes control and digital processing aspects. The ECG signal is collected by probes located on skin near the heart. Then the signal is amplified and read by analog to digital converter ADC, an internal DSP peripheral. The quality of collected ECG signal depends on probe location. This influence can be compensated using multiple probes in different positions. A maximal probe number is limited by the number of ADC channels. ADC on DSP 56803 has two channels, each multiplexed to four pins. Eight sources of analog signals can be connected in all. One ADC pin is connected to a phone line via subscriber line interface circuit (SLIC). The remaining seven pins can be connected to probes. Three probes are considered a common set. Two samplings will be taken one at a time, a third will be sampled consecutively. Maximal sampling frequency of used ADC is 800 kHz, common sampling frequency of an ECG signal is 500 Hz, so there is enough reserve to perform time multiplexed sampling. ADC resolution is 12-bit, also enough for this purpose. Electric Lawn Equipment Noise Cancellation Phone line Probe 1 Internet Appliances IP Phone SLIC Memory 16K AB Probe 2 (optional) Flash/RAM Ring (optional) Modems Magnetic Card Readers Tip Ext. Data DSP56803/5 DSP56803/5 ADC Probe 3 DAC (opt.) Security Acoustic signaling SPI Digital Speakers Keyboard Voice Recognition Systems GPIO O LED1 LED2 "Hands-free" Kits Digital Cameras Telecom Test Equipment Fuel Management Systems and more. . . . . . Figure 1. Personal EKG monitor using DSP56F803/5 DSP56F803/5 A sampled ECG signal is stored, using a double buffering technique-one buffer for storing, a second buffer for data processing. Following signal processing are filtering, abnormalities detection, and optional data compression. Usual ECG monitors use three filters: 1) notch 50/60Hz, 2) band stop 35Hz- the muscle interference, 3) high pass 0.5 Hz -drift compensation. Instruction set of DSP56F803 DSP56F803 is adapted to digital filtering so the above mentioned filters can be implemented very effectively. The ECG can be simply described as periodic pulse signal, but the wide range of its abnormalities was described in literature.1 To simplify ECG evaluation, determine the difference between measured ECG and 1. Online Journal of Cardiology, EKG World Encyclopedia, http://www.mmip.mcgill.ca/heart/egcyhome.html, 2000. reference ECG signals scaled to current beat frequency. This diagnostic is not able to classify type of disorder. The output can be two state- healthy / diseased, or continual- degree of membership to diseased state. To classify concrete disorder-infarct or arrhythmia, additional criteria must be used based on recognition of pulse shape and detection of parasitic pulses.1 Inasmuch as the theoretical number of instruction per ECG sample is 40MIPS/500Hz/3chanel ~ 26000 and instruction set is adapted to digital signal processing we have sufficient reserve to perform additional disorder classification. Input Core Algorithms Output ADC interrupt Signal processing SPI interrupt 1. Probe sample -> buffer 2. SLIC sample -> buffer 1. Filtering 2. Abnormality detection 3. Compression 4. Arbitrator 5. Set LED state 6. Prepare modem data Sample from buffer -> DAC GPIO interrupt DSP56F80X DSP56F80X CUSTOMER SUPPORT: Read key Technical Support: Figure 2. Software model The result of classification can be optionally verified by physician via a modem. The DSP is connected to a phone line via SLIC, the internal ADC is used at the input side and the external digital to the analog converter (DAC) is used at the output. DAC and DSP are connected via serial peripheral interface (SPI). Modem functions are provided by software. All the software of ECG monitor should be programmed in C. Metrowerks C compiler and integrated debugging environment is the only one supported software tool at present. Proposed layering of the software and interrupt usage are displayed in Figure 2. Individual tasks communicate using semaphores, global status variables. Modem functions and basic DSP algorithms are components of the Software Development Kit (SDK) provided by Motorola. Assuming maximal ECG frequency is three beats per second and three-channel processing with a sampling frequency of 500 Hz, 1500 samples per second and detect 9 pulses per second process are required. Using an IIR filter with six biquades, about 0.18 MIPS (in SDK implementation) to preprocess ECG signal is required. Detecting shapes of pulses by correlation with 128-point reference vectors repeating this operation for 10 pulse types, we would perform three beats three channels, 10 references = 90 computations of correlation of 128 elements/ 128 elements vectors, representing about 1.6 MIPS under SDK. Summarily, we can expect about three MIPS necessary for basic ECG evaluation, keeping reserve for time scaling of reference vectors. An additional software modem will require about 6 MIPS in V.22bis implementation (SDK) or about 15 MIPS in V.32bis implementation. www.motorola.com/semiconductors/ dsp/support email@example.com 1-800-521-6274 Website: www.motorola.com/ semiconductors/dsp Literature Distribution Center for Motorola: 1-800-441-2447 Other Inquiries: Contact your Motorola sales representative or authorized distributor Disclaimer: This sheet may not include all the details necessary to completely develop this design. It is provided as a reference only and is intended to demonstrate the variety of applications for the device. Whole ECG monitors can be designed in several variants starting with low cost variant without modem and external memory, real-time evaluation only, ending with full variant with keypad, display, modem and external memory for ECG storage. Variant with external memory must be based on DSP56F805 DSP56F805 because general purpose input/output (GPIO) pins in DSP56F803 DSP56F803 are shared with external address bus. GPIO pins are used by a keyboard, LED signaling, and optionally by display. Keyboard or display can be connected also via Serial Communication Interface (SCI). To provide connectivity with personal computer the SCI can be also used for RS-232 RS-232 interface. The functionality of proposed application design can be demonstrated on evaluation modules DSP56F805EVM DSP56F805EVM or DSP56F803EVM DSP56F803EVM, without DAC. This board contains all required hardware except input analog part, probes and amplifiers and output SLIC. DSP56F80X DSP56F80X evaluation module is fully supported by Metrowerks IDE. ©2000 Motorola, Inc. Motorola is a registered trademark and OnCE, DigitalDNA and the DigitalDNA logo are trademarks of Motorola, Inc.