Working principle of electronic blood pressure monitor based on MN101EF32D microcontroller

Features of MN101EF32D

MN101EF32D is a product launched by Panasonic in early 2008. The MN101Exx series 8-bit microcontroller combines multi-functional peripheral functions, has a flexible and optimized hardware structure, a simple and efficient instruction system, and fully realizes economy and high speed.

The MN101E32D microcontroller has built-in 64KB Flash, 4KB RAM, 6 external interrupts, 20 internal interrupts (including NMI), 9 timer counters, 3 serial interfaces, 8-channel A/D converter, 32×4 segment LCD driver, watchdog timer, single system data automatic transmission function, synchronous output function, buzzer output and other peripheral functions. The minimum instruction execution time can reach 50ns, and the package is 64-pin LQFP. The functions of the MN101EF32D used in this blood pressure monitor are as follows:

a. 10-bit A/D sampling, used for static pressure and pulse wave measurement.

b. LCD display controller directly drives a 23*4 segment LCD to display the measurement process and results.

c. Timer function, used to time A/D sampling data and calculate automatic shutdown time.

d. Use digital signal processing technology to process the A/D sampled signal, mainly including digital low-pass filtering and related calculations.

e. The power is turned on by hardware control, and the power is turned off by software control. When shutting down, except for the voltage regulator module, other chips are in a power-off state, and the power consumption is extremely low.

f. When measuring, you can choose mmHg and Kpa as the main display mode, with high measurement accuracy, reaching 1mmHg for static and 3mmHg for dynamic. Since ferroelectric memory is used as the storage medium, the data can be stored for a long time.

Hardware connection between MN101EF32D and external serial ferroelectric memory

When selecting external memory, considering the need to repeatedly erase and write the set working parameters and important information measured for a long time, and save a large amount of historical data, a large capacity static memory must be used to write as much data information as possible and ensure that the data is not lost after power failure. Due to the design process of EEPROM itself, the life is limited and the writing time is long, so it is not suitable for battery-powered systems. The data that the sphygmomanometer needs to save are designed to be systolic pressure (2 bytes), diastolic pressure (2 bytes), average pressure (2 bytes), pulse (2 bytes), time of each record (5 bytes), etc. Each measurement requires 13 bytes to store data. Assuming that 4 measurements are taken per day, 13×4＝52 bytes are required. The sphygmomanometer needs 364 bytes to store data for 7 days, so the "Ferroelectric" 24cL04 is selected. When the sphygmomanometer is turned on, the microcontroller simulates the SCL of the IIC bus at its PA0 port and inputs it to the SCL pin of the external memory 24cL04. At the same time, the PA1 port exchanges data with the SDA port of the 24cL04 to display useful data on the LCD.

Power processing module and its related circuit design

This blood pressure monitor uses two No. 7 batteries as the power input. In order to achieve better power supply quality, the DC/DC boost chip RN5RK331A is selected in this circuit to increase the voltage of about 3V formed by two 1.5V No. 7 batteries connected in series to 3.3V, which is used to supply power to the analog circuits in the system and also as the positive power supply for the digital circuits to the MCU (as shown in Figure 3). Considering that if the air pump and air valve directly share a power supply with the analog circuit and digital circuit, it will introduce relatively large interference, thereby affecting the normal operation of the pressure sensor, operational amplifier and MCU, the air pump and air valve are designed not to be connected to other devices and are directly powered by batteries. [page]

In addition, the important data collected by the sphygmomanometer are the cuff air pressure amplified by the operational amplifier and the pulse wave after DC isolation. Since they are obtained by amplifying tiny signals, the design of A/D conversion is also extremely important. The system adopts intelligent inflation measurement and automatic pressure reduction, and measures during the pressure reduction process. Since the power supply fluctuates when the air valve works to reduce the pressure, if the system power supply is directly used as the reference voltage of A/D, it will inevitably cause errors in the measurement. The LM385 of National Semiconductor is used as the voltage reference of A/D conversion and connected to the VREF+ pin of the chip to ensure the accurate conversion of the collected data.

Design of LCD display module

As shown in Figures 4 and 5, in order to make it more convenient and simple for users to use this system, LCD display is used.

Panasonic's MN101EF32D chip has a built-in LCD driver module that can directly drive the LCD. First initialize the LCD mode control register 1 (LCDMD), which is an 8-bit register used to specify the LCD clock, LCD display ON/OFF, display duty cycle, etc.

System software design

The main process of the software is as follows:

After power-on, the system is initialized first. The microcontroller starts to power the air pump, allowing the cuff to quickly inflate to about 30 mmHg above the systolic pressure of the subject. After that, the microcontroller starts to collect the air pressure of the cuff through one A/D channel, and controls the exhaust valve to exhaust according to the speed of the air pressure drop in the cuff, so that the pressure in the cuff drops at a uniform rate (3~5 mmHg/s). At the same time, another A/D channel starts to collect the pulse wave that passes through the isolation. When the amplitude of the pulse wave is the largest, the pressure of the cuff is the average pressure of the artery. The systolic pressure of the artery corresponds to the first inflection point of the amplitude envelope, and the diastolic pressure corresponds to the second inflection point of the envelope.

The software is mainly divided into the following three important modules:

1) Constant speed voltage reduction control module

Although the air valve has the feature of automatic slow deflation, in order to quickly inflate the cuff to about 30mmHg above the systolic pressure of the subject and then reduce the pressure at a uniform rate (3~5mmHg/s), the ordinary processing method cannot be used, because the entire measurement process is easily affected by external vibrations, such as artificial vibration of the cuff, vibration of the trachea, human body movement, etc. In addition, the rigidity of the trachea will also affect the slight change of the air pressure in the cuff. Therefore, the speed of pressure reduction in the cuff is nonlinearly related to the frequency of the air valve switch.

This design uses the PID algorithm to control the opening and closing time of the air valve to ensure that the cuff drops the pressure at a constant speed of 3~5mmHg/s. Due to the limitations of the processing speed and RAM resources of the single-chip microcomputer, floating-point operations are not used here. Instead, all parameters are integers, and finally divided by 2N (equivalent to shifting) to perform similar fixed-point operations, which can greatly improve the operation speed. Finally, the value is assigned to the timer to control the opening time of the air valve, thereby ensuring a constant speed of pressure reduction.

The setting and adjustment of the three basic parameters Kp, Ki, and Kd in the PID algorithm are relatively difficult. According to the working principles of these parameters, the adjustment methods are summarized as follows:

1. The pressure drops to the target value quickly, but the pressure drops too much:

a) The proportionality factor is too large;

b) The differential coefficient is too small;

2. The pressure drop does not reach the target value:

a) The proportionality coefficient is too small;

b) The integral coefficient is too small;

3. It can basically be controlled on the target, but the deviation is large and fluctuates frequently.

a) The differential coefficient is too small;

b) The integral coefficient is too large;

2) Signal processing module

The blood pressure meter has two measurement signals. The signal of the MPS-3100-006G pressure sensor is first low-pass filtered to eliminate the error of the signal reading caused by external interference, and then amplified and sent to AD1 as a static blood pressure signal; after being isolated, it is amplified again and sent to AD2 as a pulse wave signal. Since the A/D of MN101EF32D is 10 bits, the highest accuracy can reach 1/1024. In order to maximize the sampling speed of A/D conversion, interrupts are used to realize data processing after A/D conversion. When the A/D conversion is completed, in the interrupt program, the anti-pulse interference moving average method is used to realize simple and effective digital filtering to make the measurement more accurate. The specific method is to perform 5 A/D conversions continuously in a timed interrupt, remove the maximum and minimum values, and calculate the arithmetic mean of the remaining 3 data, which is used as the A/D conversion result this time.

3) Blood pressure calculation module

After the cuff pressure and pulse wave are processed by the signal processing module, the data shown in Figure 6 are obtained. The lower part of the figure is the pulse wave of the subject, and the upper part is the cuff pressure during the pressure increase and pressure decrease process of the sphygmomanometer. On this basis, the signal is analyzed for the calculation of systolic pressure, diastolic pressure, mean pressure and heart rate. The microcontroller has stored the peak value of each pulse wave and the interval time of each pulse wave during the measurement process.

The systolic pressure criterion is determined by the maximum amplitude method, that is, during the deflation process, when the ratio of the amplitude Ui of a pulse wave to the maximum amplitude Um (mean pressure) is just greater than Ks, the corresponding cuff pressure is considered to be the systolic pressure.

Ps=P/Ui=Ks*Um

The diastolic pressure criterion is also determined by the maximum amplitude method, but in the descending section of the pulse wave amplitude envelope, when the ratio of the amplitude Ui of a pulse wave to the maximum amplitude Um (mean pressure) is just less than Kd, the corresponding cuff pressure is considered to be the diastolic pressure.

Pd=P/Ui=Kd*Um

First use the empirical parameters Ks = 0.54 and Kd = 0.72 to calculate, and then make corrections after testing.

Heart rate is the period of pulse wave, which is also the arithmetic mean.

in conclusion

The blood pressure monitor based on the MN101EF32D single-chip microcomputer makes full use of the functions of the chip itself, and has the characteristics of simple circuit, low power consumption, single power supply requirement, high precision and strong practicality, and has broad market prospects.