Multichannel thermostat on arduino uno. Fan using Arduino, which depends on temperature. Components of a PID controller

The schematic diagram of a homemade four-channel thermostat with temperature display, built on the basis of Arduino UNO, LM325 and with a 1602A display, allows you to control four separate loads.

Here we describe an upgraded version of the device, which, in addition to passively measuring and indicating temperature, can also control four heaters in order to maintain a set temperature in four different places, where in the first version only temperature measurement was possible.

The operation of a four-channel thermometer and its program is described in great detail in the previous article (L.1), so here we will only talk about changes to implement the operation of a four-channel thermostat.

Schematic diagram

The schematic diagram is shown in Fig. 1.

Rice. 1. Schematic diagram of a thermostat on Arduino UNO, LM325 with 1602A display.

Unlike the first circuit (L.1), here there are four transistor switches loaded on the relay windings K1-K4, which control four different heating devices (H1-H4). Control commands for these keys come from four digital ports D9-D12 of the ARDUINO UNO board.

Program

The program is shown in Table 1.

Table 1.

The first difference is that ports are provided to control the heaters. These are ports D9-D12, they are specified here:

pinMode(12, OUTPUT);

pinMode(11, OUTPUT);

pinMode(10, OUTPUT);

pinMode(9, OUTPUT);

You can select other free digital ports; after connecting the indicator, there are still D8 and D13 left. But the author chose these: D9, D10, D11, D12. The second difference is that the if comparator function is used to control the heaters. In lines:

if(temp< -15)digitalWrite(12, HIGH);

if(temp > -15)digitalWrite(12, LOW);

if(tempi< 1)digitalWrite(11, HIGH);

if(tempi > ​​1)digitalWrite(11, LOW);

if(temp2< 20)digitalWrite(10, HIGH);

if(temp2 > 20)digitalWrite(10, LOW);

if(temp3< 10)digitalWrite(9, HIGH);

if(temp3 > 10)digitalWrite(9, LOW);

These lines indicate at what temperature, what logical level should be on the corresponding port. For example, if the temperature of the first sensor (T1) is below -15°C, there will be a logical one at port D12. When the temperature of the second sensor (T2) is below 1°C, there will be a logical one at port D11.

When the temperature of the third sensor (TZ) is below 20°C, there will be a logical one at port D10. When the temperature of the fourth sensor (T4) is below 10°C, there will be a logical one at port D9. Of course, you can set completely different temperatures - any that are needed for the specific application of this device.

Moreover, already in the finished device, they can be easily changed if necessary. To do this, you need to connect a personal computer to the USB port of the ARDUINO UNO board and load a program with other data on temperature thresholds. In principle, for this you can provide a USB connector on the body of the finished device.

Usually, in a “typical” thermostat there is a hysteresis loop, that is, a circuit that creates a difference between the temperature at which the heater turns on and the temperature at which it turns off. This is necessary so that the heater does not turn on/off very often. This is especially important if the heater is controlled by an electromagnetic relay.

Because the relay contacts are not designed for this mode of operation, and can quickly fail from burning due to sparking. But hysteresis introduces an error into the operation of the thermostat. Here it was decided not to create hysteresis, but to ensure that the relay contacts do not switch too often, simply slow down the operation of the device. To do this in the line:

The display time has been increased to three seconds. As a result, the measurements are repeated with a period of three seconds, and in any case, the relay cannot switch more than once every three seconds.

Details

Any electromagnetic relays K1-K4 can be used with 12V windings and contacts powerful enough to control specific heaters.

Output stages can be made using other circuits, for example, using opto-mistors. In this case, the LEDs of optosimitors or so-called “solid-state relays” are connected to the corresponding ports of the ARDUINO UNO board through current-limiting resistors.

Karavkin V. RK-08-17.

Literature: 1. Karavkin V. - Four-channel thermometer on ARDUINO UNO, RK-06-17.

The main task of the refrigerator controller is to maintain the set temperature in the chamber. The temperature controller will do this by changing the electrical power on the Peltier module.

In the previous lesson we developed a power regulator. The connection between power and temperature regulators looks like this.

• The temperature controller receives the measured temperature, compares it with the set temperature and calculates the set power value for the power controller.
• The power regulator generates PWM corresponding to the specified power.

We built the power regulator according to the integral regulation law. To stabilize the temperature, we will use a more complex control algorithm - a proportional-integral-derivative (PID) controller.

PID controller.

In the previous lesson I talked in detail about . He emphasized its advantages and disadvantages.

A regulator operating on this principle has high accuracy. The remaining criteria for the quality of regulation - speed and stability - are not up to par.

In order to achieve high performance for all criteria, it is necessary to use a regulator that combines different regulatory laws.

The proportional-integral-derivative (PID) controller is just such a device. It generates an output signal that is the sum of three components with different transfer characteristics. Thanks to this, the PID controller provides high quality regulation and allows you to optimize management according to individual criteria.

The following are involved in generating the output signal of the PID controller:

• Proportional component– the value is proportional to the mismatch error (the difference between the specified and real values ​​of the controlled parameter).
• Integrating component– mismatch error integral.
• Differentiating component– derivative of the mismatch error.

The mathematical form of writing the PID controller law is:

o(t) = P + I + D = K p e(t) + K i ∫e(t)dt + K d de(t)/dt

• o(t) – output signal;
• P – proportional component;
• I – integrating component;
• D – differentiating component;
• Kp, Ki, Kd – coefficients of proportional, integrating, differentiating links;
• e(t) – mismatch error.

In schematic form, the PID controller can be represented as follows.

The block diagram of the PID voltage regulator U looks like this.

• The measured voltage Ureal(t) is subtracted from the given Uset.
• The resulting mismatch error e(t) is fed to the proportional, integrating and differentiating links.
• As a result of the sum of the components, a control action o(t) is obtained, which is applied to the control element.

When implementing a PID controller in software, calculations of the output signal occur at regular intervals. Those. The controller is discrete in time. Therefore, further I will use expressions: previous signal state, previous value, etc. We are talking about the state of the system at the previous time sampling point.

Components of a PID controller.

Again. The output signal of the PID controller is the sum of three components:

• proportional;
• integrating;
• differentiating.

Proportional component.

P(t) = K p * e(t)

Has no memory, i.e. the value of the output signal does not depend on the previous state of the system. Simply the following error, multiplied by a factor, is transmitted to the output. The output signal compensates for the deviation of the controlled parameter. The larger the mismatch error, the larger the signal. If the error is 0, the output signal is also 0.

The proportional component is not able to compensate for the error completely. This can be seen from the formula. The output signal is Kp times the error. If the mismatch error is 0, then the output signal of the controller is 0. And then there is nothing to compensate with.

Therefore, in proportional controllers there is always a so-called static error. It can be reduced by increasing the Kp coefficient, but this can lead to a decrease in the stability of the system and even to self-oscillations.

The disadvantages of proportional controllers include:

• presence of static regulation error;
• low stability with increasing coefficient.

There is a significant advantage:

• High speed regulation. The response of a proportional controller to a following error is limited only by the sampling time of the system.

Regulators operating only according to the proportional law are rarely used.

The main task of the proportional component in the PID controller is to increase performance.

Integrating component.

I(t) = K i ∫e(t)dt

Proportional to the integral of the mismatch error. Taking into account the time discreteness of the controller, we can write this:

I(t) = I(t -1) + K i * e(t)

• I(t-1) – value of I at the previous time sampling point.

The mismatch error is multiplied by a coefficient and added to the previous value of the integrating link. Those. the output signal accumulates all the time and increases its impact on the object over time. Thus, the mismatch error is fully compensated even for small values ​​of the error and coefficient Ki. In a steady state, the output signal of the controller is completely provided by the integrating component.

The disadvantages of the integral regulator include:

• low performance;
• mediocre stability.

Dignity:

• Ability to fully compensate for mismatch error at any gain.

In practice, integrating controllers (only integrating component) and proportional-integrating controllers (integrating and proportional components) are often used.

The main task of the integrating link in the PID controller is to compensate for the static error and ensure high control accuracy.

Differentiating component.

D(t) = K d de(t)/dt

Proportional to the rate of change of the following error. A kind of indicator of acceleration of mismatch error. The differentiating component predicts deviations of the controlled parameter in the future and counteracts this deviation. As a rule, it compensates for delays in the influence of the regulator on the object and increases the stability of the system.

Taking into account the time discreteness of the controller, the differentiating component can be calculated as follows:

D(t) = K d * (e(t) - e(t -1))

It shows how much the mismatch error value has changed over one time unit of the controller discreteness.

There are no regulators consisting of a single differentiating link.

The main task of the differentiating link in the PID controller is to increase stability.

Setting up the PID controller.

The quality of control of PID controllers largely depends on how optimally the coefficients are selected. The coefficients of the PID controller are determined in practice in a system with a real object by selection. There are different setup methods. I will only talk about general principles.

The quality of regulation is judged by the transient response of the regulator. Those. according to the graph of changes in the controlled parameter over time.

To the traditional points of the PID controller tuning sequence, I would add that, first of all, it is necessary to decide which control quality criteria are preferable.

In the previous lesson, when designing a power regulator, we were primarily interested in accuracy and stability. And we even artificially reduced the performance. Some regulators operate under conditions of significant noise and stability is more important to them, while others require high performance even at the expense of accuracy. Optimization criteria may vary. In general, PID controllers are configured to ensure all control quality criteria are at a high level.

The components of the PID controller are configured separately.

• The integrating and differentiating links are turned off and the coefficient of the proportional link is selected. If the controller is proportional-integrating (there is no differentiating link), then a complete absence of oscillations in the transient response is achieved. When setting the controller to high speed, oscillations may remain. The differentiating link will try to compensate for them.
• The differentiating link is connected. Its coefficient aims to eliminate fluctuations in the control parameter. If this fails, then reduce the proportional coefficient.
• Due to the integrating link, the residual mismatch error is removed.

Tuning the PID controller is iterative in nature. Those. points for selecting coefficients can be repeated many times until an acceptable result is achieved.

Due to their high performance and versatility, PID controllers are widely used in industrial automation systems.

In the next lesson we will develop a PID temperature controller.

He will share with you his experience of creating a smart fan rotation controller, using a thermal sensor, an LCD display and, of course, Arduino.

A few months ago I read a number of articles about Arduino and became very interested in this device, and soon decided to purchase it. It should be noted that I am far from microelectronics, so I liked the board primarily because of its relative ease of use. Having dabbled with LEDs and “Hello world” systems, I wanted to do something practical, and at the same time become more familiar with the capabilities of Arduino. Keeping in mind the abnormally hot summer of 2010, the idea arose to assemble a cooler speed controller depending on the temperature with all related characteristics displayed on the LCD. I hope that this scheme or its variations will be useful to someone, so I decided to post my sketches.

For this scheme we will need:

• Actually the board itself Arduino or similar;
• Bread board for assembling circuit components;
• Display WH1601A-NGG-CT with trim resistor on 20 kOhm or similar;
• Resistors – 220 Ohm, 10 kOhm, 4.7 kOhm;
• Bipolar transistor SS8050D or similar;
• Digital temperature sensor DS18B20;
• Diode 1N4148 or equivalent;
• Fan axial three-wire (12V), for example - computer;
• Jack connector nutrition 2.1/5.5 mm.

The computer cooler has three wires, two of which - red (+12V) and black (GND) are used for power supply, and the third (yellow) is connected to a tachometer built on a Hall element. Unfortunately, 5V from the board is clearly not enough for us, but 6 Arduino digital outputs can operate in PWM mode (they are marked on the board itself with white squares or the letters PWM), so we can regulate the signal supply from the board to the relay, which will be responsible for the change voltage supplied to the fan.

We will receive information about revolutions from the third wire from the tachometer, using a modified method based on the implementation interrupts, which for most Arduinos can come to digital pins 2 (interrupt 0) and 3 (interrupt 1). By the way, the Arduino Mega has 4 additional pins with the ability to receive interrupts.

Now you need to position digital temperature sensor, the data of which we will use to regulate the voltage supplied to the digital output with PWM, and therefore to “open” the fan voltage channel. Dallas sensors have their own Arduino library - DallasTemperature, which we will later connect in the sketch. The library must be unpacked into the arduino-0018/libraries/ directory.

The last thing left is connect LCD, where we will display all the current information about the temperature and fan speed. Since I used the WH1601A screen for the build, there may be known issues with displaying lines. To eliminate them, we will use the LiquidCrystalRus library, which must also be unpacked into the arduino-0018/libraries/ directory.