A battery fed MQTT weatherstation

(Note: if you consider building this… I am working on a similar station with a bare ESP12 and a BME280 for even better energy efficiency)

Weatherstations are a popular build for DIY-ers and with the ESP8266 WiFi capabilities that has become very easy.
Here I will present a simple weatherstation (DHT sensor BMP sensor for temperature, humidity and atmospheric pressure) that can be battery fed or mains fed.
It will send the info by MQTT and in order to save batteries, go to sleep most of the time.
Components needed are:

  • A wemos mini D1
  • A DHT11 or DHT22
  • A BMP280
  • A 100k resistor for the battery
  • A small switch or jumper
  • A LithiumIon charger module

Just a word for the wise… the Wemos D1 is not the best option to battery feed, as it has some peripheral components (such as the CH340 chip) that draw current. A better choice would be the bare ESP8266, but it is not for everybody to solder that.

The application has OTA. Using OTA together with deep sleep is always a bit of a tricky situation as you cannot do OTA when the ESP8266 is in deep sleep.
basically there are 3 solutions for that:

  1. Include a pause in which you can do OTA. Problem with that is that you would need some indication of when that pause starts (e.g. once the MQTT messages have come in). Another problem is that that pause in which the processor is not in deep sleep, it wil consume more energy.
  2. Make the deepsleep dependent on the state of a pin. With a jumper or Switch you could make one pin high or low and thus allow deepsleep or not.
  3. By having the software check on a seperate website whether or not there is an update for the software and then download that software. Obviously that requires some more organization and if the Wemos will do that check whenever it awakes, it will do that say every 15 minutes for something that may never happen.

Considering the above, Option 2 seemed the best, so I added a small switch to tell the software whether or not to go into deepsleep.


The program once loaded (get it here) has several tabs in the IDE.
The main program “esp8266_weatherstation” is quite straightforward, it sets up all the usual parameters, variables and the sensors.
It reads the sensors and sends these via MQTT to a broker.
Once that is done, it goes into deepsleep, depending on two parameters:

  • the “FORCE_DEEPSLEEP” variable: if present there will be a 15 min deepsleep. If not set, the deepsleep will be determined by the battery voltage.
  • the state of the ‘SLEEPPIN (D6):  if LOW there will be deepsleep, if HIGH, there will be no deepsleep

If there is no deepsleep, a pause of 30 secs is added to avoid the MQTT messages being send in a frenzy.

You may be puzzled by the following code section:

float h = dht.readHumidity();
float t = dht.readTemperature();
float h = dht.readHumidity();
float t = dht.readTemperature();

The reason for this is the response time of the DHT22. If you  start a reading, the results you get are from a reading that was started 2 secs before. Usually that is not an issue, but if you put the processor to sleep for 15 minutes, that first reading will give you 15 min old values. If that is not a problem, by all means remove the delay and second reading. It will not make a difference on your battery life as your  DHT22 is not put to sleep.
technically ofcourse it is possible to feed the DHT22 from one of the I/O pins and switch it off when not needed, but the DHT22 also needs a warm-up time of about 2 secs, so you would need to take that into account. No doubt that warm up time could be catered for with the delay(2000) command mentioned above.

The “OTA” section sets up the OTA function. As explained before, OTA and DEEPSLEEP is a bit of an uneasy combination. In order to use OTA, tou would need to put the switch on D6 in the right position (HIGH) and either reset  the WEMOS or wait till the next wake up. Mind you that when DEEPSLEEPing the OTA port  can disappear, but it should come back once the processor is out of DEEPSLEEP.

the “config.h” file defines some variables

the “webconnect” file connects to WiFi. If for some reason that connection cannot be made after 10 times trying, the processor is going back to sleep for 30 more seconds and then tries again.


The libaries used are quite standard, with the exception of the PubSubClient library. This is not the well known PubSubclient library from Nick O’Leary, but a fork made by Ian Tester. The reason for using that one is that it is easier to send variables in the MQTT payload.
The DHT library is the DHT sensor Library from Adafruit. If you prefer another DHT library (there are many) you will probably only need to make minor changes.
The BMP280 library is also from Adafruit.

The sensors


The DHT22 is the more accurate sibling of the DHT11, but both have a tendency of failure. generally the DHT22 needs a pull up resistor, but one can also use the internal pullup of the I/O pin for that.

There are two specification where the DHT11 is better than the DHT22. That’s the sampling rate which for the DHT11 is 1Hz or one reading every second, while the DHT22 sampling rate is 0,5Hz or one reading every two seconds and also the DHT11 has smaller body size.

If you are using this sensor outside, you can protect it against dirt and  creepy crawlies by covering it with a single layer of teflon tape. Supposedly this is fully permeable for humidity.

The DHT sensors do not have eternal life. they give up working alltogether (and usually permanently) when the humidity is 100%

Properties DHT11 DHT22
Voltage 3-5 Volt 3-5 Volt
Operating current 2.5mA 2.5mA
Temperature range 0 t/m 50 ºC -40 t/m 80 ºC
Temperature accuracy ±2 ºC ±0,5 ºC
Humidity range 20-80% RH 20-95% RH
Humidity accuracy ±5% RH ±2-5% RH
Sampling rate 1Hz 0.5Hz

The BMP280 is the successor of the BMP085, BMP 180 and BMP183

Ofcourse it is possible to use one of the above predecessors, but these older sensors are generally more expensive than the BMP280.
The BMP280 can measure barometric pressure with ±1 hPa absolute accuraccy, and temperature with ±1.0°C accuracy. It can be controlled both by SPI and I2C.

Do not mix it up with the BME280. That sensor can also read humidity, but it is slightly more expensive.
The BME280 ofcourse would be a perfect chip to go to if you want to avoid the DHT22 sensor. It can read temperature, humidity and barometric pressure. I just didnt have it lying around, thus couldnt try.

If you are no fan of the DHT11/22 consider the I2C HTU21D, or theSi7021 (replaced by the HTU21D).
If you are only interested in say temperature and not in humidity, the BPM280 is the best choice and you can omit the DHT22. If you only want temperature, the DS18B20 is probably the best choice.

The battery

Lipo Charge module

It is hard to give recommendations for what battery to use. I am using the battery shield to which I connect a LiPo battery. That makes it all very simple but it is not the most efficient way of using the energy from a battery as the battery shield first boosts it up to 5 Volt after which the Wemos makes 3V3 Volt from it.
Another easy way is to use a Lipo battery with a lowdrop 3V3 regulator and a cheap charging module (pictured top). You could combine the charging module with a small solar cell (with a 5-6 Volt max output). The 100k resistor in the  circuit sketch should go directly to the V+batt, to monitor the battery voltage. The choice for the 100k resistor is explained here.  If for power efficiency you decide to use a bare ESP8266 instead, one has the opportunity to increase the resistors of the voltage divider to come to a smaller current through these resistors


Dimming an AC lamp via MQTT

Years ago I published a TRIAC dimmer that could be controlled by a simple microcontroller such as an arduino. Times have changed and right now it is of more importance to be able to control that lamp from a home automation system such as OpenHab, Homematic or NodeRed.

So, let’s have a look at the circuit first:
This is a rather classic circuit. It detects the zerocrossing on the grid and then a microcontroller ignites the TRIAC with a time delay that determines the brightness of the attached Lamp.  The full cycle is 10mS (at 50Hz)
A circuit like this is easy to make for very low cost (1-2 Euro).

However, if you shy away from soldering  such a circuit, there are ready made modules available as well:
You should not need to pay more than 3-4 USD for such a module. Anything above that is robbery

Software to control this dimmer would look something like this:

#include <Ethernet.h>
#include <PubSubClient.h>
#include  <TimerOne.h>

#define CLIENT_ID       "Dimmer"
#define PUBLISH_DELAY   3000

String ip = "";
bool startsend = HIGH;
uint8_t mac[6] = {0x00, 0x01, 0x02, 0x03, 0x04, 0x07};

volatile int i=0;               // Variable to use as a counter volatile as it is in an interrupt
volatile boolean zero_cross=0;  // Boolean to store a "switch" to tell us if we have crossed zero
int AC_pin1 = 4;// Output to Opto Triac
int dim1 = 0;// Dimming level (0-100)  0 = on, 100 = 0ff

int inc=1;          // counting up or down, 1=up, -1=down
int freqStep = 100; // make this 83 for 60Hz gridfrequency

EthernetClient ethClient;
PubSubClient mqttClient;
long previousMillis;

void zero_cross_detect() {    
  zero_cross = true;               // set the boolean to true to tell our dimming function that a zero cross has occured
  digitalWrite(AC_pin1, LOW);       // turn off TRIAC (and AC)

// Turn on the TRIAC at the appropriate time
void dim_check() {                   
  if(zero_cross == true) {              
    if(i>=dim1) {                     
      digitalWrite(AC_pin1, HIGH); // turn on light       
      i=0;  // reset time step counter                         
      zero_cross = false; //reset zero cross detection
    else {
      i++; // increment time step counter                     

void setup() {
  attachInterrupt(0, zero_cross_detect, RISING);    // Attach an Interupt to Pin 2 (interupt 0) for Zero Cross Detection
  Timer1.initialize(freqStep);                      // Initialize TimerOne library for the freq we need
  Timer1.attachInterrupt(dim_check, freqStep);
  pinMode(4, OUTPUT);

  // setup serial communication

  while (!Serial) {};

  // setup ethernet communication using DHCP
  if (Ethernet.begin(mac) == 0) {
    Serial.println(F("Unable to configure Ethernet using DHCP"));
    for (;;);

  Serial.println(F("Ethernet configured via DHCP"));
  Serial.print("IP address: ");

  ip = String (Ethernet.localIP()[0]);
  ip = ip + ".";
  ip = ip + String (Ethernet.localIP()[1]);
  ip = ip + ".";
  ip = ip + String (Ethernet.localIP()[2]);
  ip = ip + ".";
  ip = ip + String (Ethernet.localIP()[3]);

  // setup mqtt client
  mqttClient.setServer( "", 1883); // <= put here the address of YOUR MQTT server //Serial.println(F("MQTT client configured")); mqttClient.setCallback(callback); Serial.println(); Serial.println(F("Ready to send data")); previousMillis = millis(); mqttClient.publish("home/br/nb/ip", ip.c_str()); } void loop() { // it's time to send new data? if (millis() - previousMillis > PUBLISH_DELAY) {
  previousMillis = millis();


  Serial.print("dim1 in loop = ");

void sendData() {
  char msgBuffer[20];
  if (mqttClient.connect(CLIENT_ID)) {
    if (startsend) {
      mqttClient.publish("home/br/nb/ip", ip.c_str());
      startsend = LOW;

void callback(char* topic, byte* payload, unsigned int length) {
  char msgBuffer[20];
   payload[length] = '\0';            // terminate string with '0'
  String strPayload = String((char*)payload);  // convert to string
  Serial.print("strPayload =  ");
  Serial.println(strPayload); //can use this if using longer southbound topics
  Serial.print("Message arrived [");
  Serial.print("] ");//MQTT_BROKER
  for (int i = 0; i < length; i++) {



The value freqStep is usually set at 75 or 78 with this type of circuits, which allows for 128 steps of dimming at 50Hz gridfrequency. I have set the value here purposely to 100 allowing for only 100 steps, which is convenient for use in OpenHAB as  the slider goes from 0-100%.
The calculation is as follows:

As we have two zerocrossings per sine wave, the frequency of zerocrossings is 100Hz. Therefore the period between two zerocrossings is 10mSec is 10000uS.
So if you want to have 100 levels of dimming the steps you need to take are 10000/100=100uS

The frequency of zerocrossings is 120Hz, therefore the period between two zerocrossings is 8.3mS is 8300uS. So if you want to have 100 levels of dimming, the steps you need to take are 8300/100=83uS -> freqStep=83


Controlling Neopixel or RGB LED with Openhab

Controlling Neopixels or RGB LEDs from an ESP8266, controlled by OpenHab is fairly simple. I will present here a working system with a Colorpicker, some predefined buttons and sliders. It will be updateable by OTA and provide some  feedback to openhab on the node parameters

Using a Colorpicker
The sitemap and items file are fairly easy, just using a Colorpicker. I will describe some additions, but the basic files are quite simple


Group All
Color RGBLed "NeoPixel Color" (All)
String RGBLedColor (All) {mqtt=">[mosquitto:OpenHab/RGB:command:*:default]"}

There are 2 items here. One item contains the HSB value, coming from the colorpicker, The other item, that will not be visible on the screen, will contain the calculated RGB code that will be sent out via MQTT.

sitemap file is as follows:

sitemap NeoPixel label="NeoPixel"
	Frame label="NeoPixel" {
		Colorpicker item=RGBLed icon="slider"

The colorpicker sends an HSB (Hue, Saturation, Brightness) code, but that isnt anything a regular Neopixel LED or regular RGB LED can use, so I set up a rule that triggers on any change of the RGB item to change the HSB code into a regular RGB code. That RGBcode is joined in 1 string called ’ color’. The rule then assigns that ‘color’ string to a new item called “RGBLedColor”

rules file

import org.eclipse.smarthome.core.library.types.DecimalType
import org.eclipse.smarthome.core.library.types.HSBType

rule "Set HSB value of item RGBLed to RGB color value"
	Item RGBLed changed
	val hsbValue = RGBLed.state as HSBType

	val brightness = hsbValue.brightness.intValue 
	val redValue = ((((hsbValue.red.intValue * 255) / 100) *brightness) /100).toString
	val greenValue = ((((hsbValue.green.intValue * 255) / 100) *brightness) /100).toString
	val blueValue = ((((hsbValue.blue.intValue * 255) / 100) *brightness) /100).toString

	val color = redValue + "," + greenValue + "," + blueValue

	sendCommand( RGBLedColor, color)

//part of Neopixel conf files

On the receiving end, there is a Wemos D1 mini that controls a NeoPix strip. The ‘color’ string is split in its separate RGB values. As such it could also be used for an RGB LED or seperate Red, Green, Blue Led’s.
The code is reacting to one MQTT topic, basically switching the entire strip at one go in the desired color and brightness, but I added a structure catching some other MQTT topics that now go to empty routines but that could be used for patterns like a breathing light or a moving pattern. As illustration I added a simple red-blue pattern that responds to MQTT “OpenHab/RGB/scene” with payload “2”  I am using the broker name ‘mosquitto’, if your broker has a different name you needto modify that.

I have to say though that this is not the most efficient way of adding topics, but for beginners it is easiest to follow.
A much more efficient way is to keep the topics of one specified length with only say the last 2 or 3 characters defining the desired action and then only checking for those last characters. A good example of this is for instance in the gateway or sonoff code on https://github.com/computourist/ESP8266-MQTT-client4. For readability sake however, in this example I have chosen to use a more traditional approch and as long as you just want to control some LEDs it is easier.

The ESP8266 code does send some information back: it returns the software version, the IP number and the RSSI.
With regard to the software version and IP number… if you have several ESP8266’s and or Arduino’s in your system, it is easy to lose track which one does what and with what software.

The ESP8266 code is OTA updateable
I have used a fork of the O’Leary PubSub library that makes it easier to send variables as the MQTT payload without having to resort to setting up buffer space and using c_string. Library is here4: https://github.com/Imroy/pubsubclient4.
If you already have the O’Leary PubSub library installed, you will need to put the library files in your sketch folder and change #include <PubSubClient.h> into #include “PubSubClient.h”

Sending predefined colors
If next to the Colorpicker you also want to be able to send predefined colors via a button, then you only need to add some lines in the items and sitemap file:

add to itemsfile:

    Switch NEO_RED	"Red"	<red> {mqtt=">[mosquitto:OpenHab/RGB:command:ON:255,0,0]"}
    Switch NEO_YELLOW	"Yellow"	<yellow>	{mqtt=">[mosquitto:OpenHab/RGB:command:ON:100,96,0]"}
    Switch NEO_BLUE	"Blue"	<darkblue>{mqtt=">[mosquitto:OpenHab/RGB:command:ON:0,0,255]"}

add to the sitemap:

Switch item=NEO_RED mappings=[ON="ON"]
Switch item=NEO_YELLOW mappings=[ON="ON"]
Switch item=NEO_BLUE mappings=[ON="ON"]

As it can be a bit of a bother to find and type all the codes for the colors you may want, I added for download some 160 predefined color settings to pick and add to your sitemap and itemsfile, including some simple icons. As these items already send an R,G,B, code they do not need the rulesfile, only the colorpicker does.

How about Sliders
Perhaps you do not want to use a colorpicker but just individual sliders for Red, Green and Blue.
That is quite easy too.
add the following to your itemsfile:

Dimmer NEO_SLIDERRED "Neopixel Red [%d %%]"  <red> {mqtt=">[mosquitto:OpenHab/RGB/RED:command:*:default]"}
Dimmer NEO_SLIDERGREEN "Neopixel Green [%d %%]" <green> {mqtt=">[mosquitto:OpenHab/RGB/GREEN:command:*:default]"}
Dimmer NEO_SLIDERBLUE "Neopixel Blue [%d %%]" <blue> {mqtt=">[mosquitto:OpenHab/RGB/BLUE:command:*:default]"}

and this to your sitemap

Slider item=NEO_SLIDERRED 

Normally that would be enough already, but our ESP8266 program is expecting one input rather than seperate RGB inputs. That is because the colorpicker we used before sends one string of info.
We could add topics to the ESp8266 program to respond to individual RG and B values (and that is what the MQTT topics OpenHab/RGB/RED, OpenHab/RGB/GREEN, and OpenHab/RGB/Blue would be for), but it might be easier to combine those and send as one. For that we need a new rule.

add this to your rules file:

rule "Send  RGB items from slider"
Item NEO_SLIDERRED changed or
Item NEO_SLIDERGREEN changed or
val redValue=  Math::round(2.55* (NEO_SLIDERRED.state as DecimalType).intValue)
val greenValue=  Math::round(2.55* (NEO_SLIDERGREEN.state as DecimalType).intValue) //
val blueValue=  Math::round(2.55* (NEO_SLIDERBLUE.state as Number).intValue)
val color = redValue + "," + greenValue + "," + blueValue
	sendCommand( RGBLedColor, color)

While we are at it
If you want the Filename, IP number and RSSI to show up, add the following to your items file:

Number W_RSSI  "RSSI [%d dbm]" <signal>  {mqtt="<[mosquitto:home/nb/weer/RSSI:state:default]"}
String W_IP "IP [%s]" <network>   {mqtt="<[mosquitto:home/nb/weer/IP:state:default]"}
String W_version "Software versie [%s]" <version>  {mqtt="<[mosquitto:home/nb/weer/version:state:default]"}
String W_MAC "MAC [%s]" <mac> (Wweather) {mqtt="<[mosquitto:home/nb/weer/mac:state:default]"}
Number W_IP_Uptime "Uptime [%d min]" <clock> (Wweather) {mqtt="<[mosquitto:home/nb/weer/uptime:state:default]"}

and the following lines to your sitemap:

   Text item=W_RSSI
   Text item=W_IP
   Text item=W_version
   Text item=W_MAC
   Text item=W_IP_Uptime


Although I tried on a Wemos D1 mini, I presume this will also work perfectly on the ESP8266-01. Perhaps a great way to use those if you still have on your scrapheap as this is typically a minimal pin project.

The ESP8266 program
The Esp8266 program is rather self explanatory and going into it too deep might fall outside the openhab scope a bit, but in brief, you will find several ino files and a userconfig.h file. These all should be in the same directory and it is sufficient to open the NeopixelOpenhab_vx.ino file in your IDE. The other files will then open up too.

You need to make a few modifications in the userdata.h file.

The essential changes are your password, your SSID and the address of your MQTT server. The other options are indeed ‘optional’, can leave them as is.
the Filename is sent back to openhab as an MQTT message, so in the future you will still know what file it was you put in your ESP8266. obviously you would need to fill out the correct filename. Maybe there is a possibility to automatically pull the filename from the system but if it is, then I haven’t found out about it yet

IFTTT Webhooks on Arduino

IFTTT is a real handy service that allows you to make applets doing a variety of things, like email you the weather update every day.
Also, most home automation services like ‘Openhab’, ‘NodeRed’ or ‘HomeAssistent’ have a link with IFTTT so you can use your own data to trigger events or have have IFTTT take action in your home automation application, like switch on lights.

But, what if you dont have a homeautomation system and you just want your individual lonely arduino with ethernetshield to trigger something in IFTTT?

Fortunately IFTTT has a service for that called ‘Webhooks’ (it used to be called ‘Maker’). I have used that service in the past out of curiosity, but recently I needed it again and sadly had to scratch my head on ‘how did I do that again?’
So after getting myself re-aquainted with it, it might be good to share that with people who are interested in using it.

Therefore, let’s make a program that uses IFTTT to Tweet data every time the garagedoor is opened. Yes I know, you can tweet directly from an Arduinoprogram as well, but that’s not the point.

First ofcourse you need to register for IFTTT if you haven’t yet done so already.
Then you create a new applet and for the “IF” service, you choose “webhook”. If you never have used webmaker before, it will ask you to connect that service and provide you with an API key. Store that one, you will need it.
In Step 2 it will ask you for an “event name”.
this is a free to choose name that we will use later to trigger the event. Let’s make that ‘garage_door’.




In step 3 and 4 we chose the Twitter action. If you never used twitter before with IFTTT, you will now be asked for authorization from Twitter.

In Step 5 we format the message that we want to Tweet. The picture shows how I have done that.
Save the applet

We now have finished the IFTTT part, but before we go on to the Arduino part, we will test it:
Go to your browser and type:


in which “d56e34gf756” should be replaced by YOUR apikey.

Within moments you should see your tweet appear in your timeline.

So all we need to do now is to have an Arduino make a POST or GET webrequest akin to the one above.
However, in this case I will not add the values as arguments but as a JSON, but it gives the same result

You will find the code belonging to this action on github.
Now obviously this code needs a trigger as well in the shape of a doorcontact (NC) on digital Pin 4, pulling that down to ground

Adding a popular 5Volt 4 channel relay board to a 3V3 processor (beginners)

Chinese webshops sell a 5Volt 4 channel relay board that is quite popular amongst hobbyist. Sometimes I get the question if it can be used with a 3V3 microprocessor. This ofcourse is very basic Ohms law, so just skip this article if you already know, but for those who still have questions: read on.

Well in short: any relay can be made to work with any microprocessor, but lets look at how this can be done in an efficient way.

Normally, if you have a bare 5Volt relay, it would only require a transistor or a FET to drive the relay with any voltage sufficient to open the transistor. The relay board in question though already has some electronics around it, so lets have a look at the circuit:

It is quite clear that this is not the standard one transistor relay board: the Transistor driving the relay in itself is driven by an optocoupler, that also inverts the signal: It is a LOW on the INx input that will  bring the optocoupler to open, thus feeding the base of the transistor, that will subsequently open and activate the relay. The circuit also shows that the optocoupler can be fed from the same  voltage source as the  relay, or from a different voltage source.

First question we have to ask ourselves: is a 3v3 circuit enough to drive the board?
Well, suppose we feed the board entirely with 5Volt. If we then would connect a 3V3 processor pin such as from a raspberry or ESP8266 or  a modern Arduino, a LOW would indeed activate the relay, as described above, but what happens if we do a HIGH?
Well, with a HIGH there would be a voltage of (5-3.3=) 1.7 volt over the series resistor, the optocoupler and the LED. Given the fact that the  forward voltage of the  optocoupler is about 1.5 Volt and the forward voltage over the red LED is on average 2.2 Volt, that basically absorbs the 1.7 Volt making it safe to say there will be no current flowing through the optocoupler when the IO pin is made HIGH (3V3).
Factually, even if the ‘HIGH’  would only be 2.5 Volt, chances are slim that there would be any current flowing through the optocoupler.

Now ofcourse there are people who shrug at the idea of  having any 5Volt source connected to their 3V3 pins, so what happens if we  take away the Voltage jumper, feed the relay with 5Volt and the optocoupler with 3V3.

Well, suppose we again make the IO pin LOW, then there will be 3V3 over the optocoupler, the LED and the resistor.
The LED and the optocoupler will  have a voltage drop of 1.5+2.2=3.7 Volts. Ergo, there will not be any current flowing through the optocoupler and the relay will not be activated.

Therefore: yes, the relay board can be used with a 3V3 processor, but  you will have to feed the entire board with 5Volt.

But, what if I do not want the signal to be inverted, if I want a LOW to deactivate the relay and a HIGH to activate it?

Well, that is fairly easy: do that with an inverter. One can either use an inverter IC such as a 7404, 7414 or a 4069, but ofcourse then you’d have to wonder whether 3V3 would be enough input for those IC’s  (it is).
Another possibility is just a transistor with a base resistor 470-1000 Ohm (1x for every channel).


MQTT for beginners

Although there are ample ready made programs for the ESP8266 to connect to
MQTT, I still get questions of people on how to do it. So I will explain it step by step here. If you already know exactly on how to do this, you may just as well skip this post, although there might be some things of a more general interest towards the end.

For those who really have no idea what MQTT actually is, it is a lightweight protocol for machine to machine communication. It consists of a topic and a message (also called payload) that is used often in iot and homeautomation projects. In short, and pulling it a bit in absurdum, with MQTT your refrigerator can talk to your oven and your coffee-maker, provided these are all ‘smart devices. However, they do not talk directly to eachother, they do that via a ‘broker’. A broker is a machine that keeps track of all the MQTT traffic and knows what messages need to go to what device. That means that when your fridge
wants to talk to the coffee maker, there is no need for the oven to listen to that message.

So first of all you need an MQTT broker. That could be a public broker or your own broker that you run locally (for instance on a Raspberry Pi). Mosquitto is a popular broker to run yourself (check here how). If you do not want to run your own broker you have options like test.mosquitto.org, CloudMQTT, HiveMQ, MQTT Dashboard (though that uses Hive), RabbitMQ and io.adafruit . To get up and running fast, test.mosquitto.org is the easiest. However, it is a completely open test broker. There is no security on it whatsoever so it is adviseable to change to a private or password protected broker as soon as you have your mqtt up and running.

For the ESP8266 program that we will be making, there will be 4 important credentials that we need. Those are the server, the user, the password and the port. Mind you that “test.mosquitto.org” does not require a user log-in and also when you are running your own broker you do not need (but can set) a log-in, but I will introduce these two variables in the program for when you decide you need/want a log in.

The ESP8266 using the MQTT protocol is considered a ‘client’ of the ‘broker’. To let the ESP8266 talk to the broker there is very popular library called PubSubClient. It is called ‘PubSub’ because it allows the ESP to ‘publish’ MQTT messages to the broker and to ‘Subscribe’ to MQTT messages of the broker. Do you remember I said the oven des not need to listen to the fridge talking to the coffeemaker? Well that is what the ‘subscription’ is for: The specific ESP8266 is only getting messages that it ‘subscribed’ to. The library can be installed via Arduino IDE library manager.

The code

Like in all programs, we begin by including the libraries that are needed for all the functionality we want/need. We need the ESP8266WiFi library, for the ESP8266 to connect to a WiFi network, and the PubSubClient library, for the ESP8266 to connect to a MQTT broker and publish/subscribe messages in topics.

#include <ESP8266WiFi.h>
#include <PubSubClient.h>

Next, we will declare the global variables for our connections. We need the WiFi credentials, to connect to the WiFi network.

constchar* ssid = "YourWiFiSSID"; constchar* password = "YourWiFiPassword";

Next we need the information of the MQTT server. As explained earlier, you will need the server address, the port, the username and the password, but for our test we can leave the user and password empty. In most MQTT brokers, the port is 1883

constchar* mqttServer = “test.mosquitto.com”; constintmqttPort = 1883; constchar* mqttUser = “YourMQTTUsername”; constchar* mqttPassword = “YourMQTTPassword”;

Next we will declare two ‘objects’ (like with all libraries): the WiFiClient object establishes a connection to a specific IP and port. The PubSubClient receives input of the WiFiClient that we just created

We will keep the PubSubclient object simple with just one argument, but it can receive more arguments. For that you may want to check the the API documentation.

WiFiClient espClient;
PubSubClient client(espClient);

In the setup function we connect to the WiFi network and for convenience we print the progress to the serial monitor

WiFi.begin(ssid, password);

Serial.print("Connecting to WiFi..");
Serial.println("Connected to the WiFi network");

Then we set the address and the port of the MQTT server. We do this with the setServer method of the PubSubClient object.

client.setServer(mqttServer, mqttPort);

We are almost there, but we still need to tell the program what to do with incoming MQTT messages. For that we use the setCallback method of the PubSub object to set a function that is executed when a MQTT message is received. Traditionally this function is called ‘callback’, but you can give it any other name


Now, we need the Setup function to connect to the MQTT server. Preferably we do this in a loop till we get succes.

(!client.connected()) {
Serial.println("Connecting to MQTT...");
if (client.connect("ESP8266Client", mqttUser, mqttPassword )) {
Serial.print("failed with state ");

The setup function is a good place to publish our first MQTT message to tell the world we are there. Kind of an “Hello World” to let the broker know we are on line. We do this with the publish method, that takes a topic and a message as arguments. As a topic we choose “test/esp8266”, and for a message “Hello ESP world”.

client.publish("test/esp8266","Hello ESP world");

Leaves us to subscribe to a topic to let the broker know what messages from other publishers we like top receive. We do that with the subscribe method, with the topic name as argument. We can take any topic we desire, but for now lets subscribe to the same topic as we publish.


The full setup function now looks like this:

  while(WiFi.status() != WL_CONNECTED) {
    Serial.println("Connecting to WiFi..");
  Serial.println("Connected to the WiFi network");
  client.setServer(mqttServer, mqttPort);
  while(!client.connected()) {
    Serial.println("Connecting to MQTT...");
    if(client.connect("ESP8266Client", mqttUser, mqttPassword )) {
      Serial.print("failed with state ");
  client.publish("test/esp8266", "Hello ESP world");

The Topic

This would be a good moment to say some words about the format of the topic. In its smallest form, the topic only needs to be one character long.
Traditionally though, topics have a hierarchy that is defined by backslashes. A topic for instance could be “home\garden\lawn\sprinkler”. This hierarchie allows
subscription on several levels. You could for instance subscribe to the topic “home\#” to get all the messages that deal with ‘home’. You could subscribe to the topic “home\garden\#” to get all the messages that have to do with your garden etc. etc. Often you will see a topic starting with a backslash, like
“\home\garden\lawn\sprinkler” There is absolutely no gain in starting a topic with a backslash.

The callback function

The callback function receives the MQTT messages (topic + payload) that our ESP PubSub client has subscribed*) to. The arguments of the callback function are the name of the topic, the payload and the length of that payload.

As illustration, we will setup the function to print
the topic to the serial port, and then print the received payload.
Since we also have the length of the payload as argument of the
function, this can be easily done in a loop.

voidcallback(char* topic, byte* payload, unsigned intlength)

  Serial.print("Message arrived in topic: ");
  for(inti = 0; i < length; i++) {

*) To be precise: the subscription is to a topic, the client then receives the topic + the payload coming with that topic. The payload is sometimes called ‘the message’ but confusingly the topic+payload is sometimes also called “the message”

In the above example we have only subscribed to one specific topic:

The topic “test/esp8266/node1” for instance, will not arrive. For that we would need to specify a wildcard “#”

The main loop

In the main loop function, we use the loop method of the PubSubClient. This function is called to allow the client to process incoming messages and maintain its connection to the server.


The full code

The full code now is as follows:

#include <ESP8266WiFi.h>
#include <PubSubClient.h>
constchar* ssid = "YourWiFiSSID ";
constchar* password =  "YourWiFiPassword";
constchar* mqttServer = "test.mosquitto.com";
constintmqttPort =1883;
constchar* mqttUser = "YourMQTTUsername";
constchar* mqttPassword = "YourMQTTPassword";
WiFiClient espClient;
PubSubClient client(espClient);
  WiFi.begin(ssid, password);
  while (WiFi.status() != WL_CONNECTED) {
    Serial.println("Connecting to WiFi..");
  Serial.println("Connected to the WiFi network");
  client.setServer(mqttServer, mqttPort);
  while (!client.connected()) {
    Serial.println("Connecting to MQTT...");
    if(client.connect("ESP8266Client", mqttUser, mqttPassword )) 
      Serial.print("failed with state ");
  client.publish("test/esp8266","Hello ESP World");
  client.subscribe("test/esp8266");// here is where you later add a wildcard
voidcallback(char* topic, byte* payload, unsigned intlength)
  Serial.print("Message arrived in topic: ");
  for(int i = 0; i < length; i++) {


To test the code, upload and run it on your ESP8266. Open the Arduino IDE serial monitor to print the output.

The ESP8266 will send the “Hello ESP World” message, however, that will not be printed on the serial monitor, but as the program also subscribed to that same topic, it will receive that same messages, coming back from the broker and then print it on the serial monitor.

If any other client would send message to the “test/esp8266” topic, then that also will be printed in the serial monitor

Now if it does not work and you are sure the code is OK, is very wise to have MQTT-Spy installed on your computer. MQTT-spy monitors the MQTT traffic on your network and it can publish and subscribe to any desired MQTT topic.

OK, so now what? How do I send a variable?

After your initial joy that everything is working you immediately would want to send something useful for instance that reading of a DHT22 or a DS18B20 sensor and suddenly you need to make the step of the string “Hello World” to a temperature variable. Ofcourse you will try and just insert the variable in place of the “Hello World” string and hope for the best, but that will fail, as the PubSub library (and the MQTT protocol) will only accept strings.
You will see it will not even accept a string variable, so what to do? Well there are several ways to put a variable in the MQTT payload:

1- take another library.
There is a fork of the PubSubClient library by Ian Tester. It will accept variables if defined like this:
client.publish("topic", String (variable));
However, it is often not a drop in replacement, you will need to make minor adjustments in setting up the various calls.

2-use some C++language
Suppose you want to publish some interger variable  called ‘temperature’

sprintf(buff_msg, "%d", temperature);
client.publish("topic", buff_msg);

ofcourse you will need to set up the buff_msg buffer in your global variable declarations like this:

buff_msg[32]; // mqtt message

A very frequently used  payload to send will be OFF or ON. Although often it  might be just as easy to define an IF statement for your payload like this:

if (status==1){client.publish("topic","ON");}
if (status==0){client.publish("topic","OFF");}

it is also possible to do that with  the sprintf command:

sprintf(buff_msg, "%d",  status);

So, can I send a JSON with MQTT?

Sure you can. I am pretty sure it works with the above method, after all a JSON is just some type of string. However, I prefer another method for a JSON.

Suppose I want to send the following JSON:
do that as follows:

String payload="{\"garden\":temperature\",\"data\":["+String(temperature_outside)+","+String(temperature_greenhouse)+","+String(temperature_compost)+","+String(temperature_coldframe)+","+String(temperature_soil)+"]}";

payload.toCharArray(data, (payload.length() + 1));

I like to send my IP number as MQTT, is that possible?

Ofcourse. There are several methods and in the end all roads lead to Rome. Using c_str is sometimes advised but I prefer the following.

in your general variable declaration declare:
String IP;

Then after  you connect to WiFi  state:
IP = WiFi.localIP().toString();

Then where you handle your publishing, do the following:

for (i = 0; i < 16; i++) {
buff_msg[i] = IP[i];
buff_msg[i] = '\0';
client.publish("topic", buff_msg);

Can I send variables in the topic?

Yes you can. This is in fact not so different from sending variables in the payload:
In your variable declaraion  declare a buffer:
buff_topic[30]; // mqtt topic

Now suppose you want to send the ID number or the name of your  ESP8266 as part of the topic and an error as the payload:

in your  global declarations you have defined the number or name of your ESP:
#define nodeId 13 // node ID

and then where you send your  messages do:

sprintf(buff_topic, "home/nb/node%02d/error", nodeId);
sprintf(buff_msg, "syntax error %d", error);
client.publish(buff_topic, buff_msg);

What about receiving MQTT messages?

Well, I spoke about the callback function that handles the incoming messages, but I admitt that just waiting for one standard solitary “Hello World” message isn’t very useful.
To subscribe to more than one message we need to use a wildcard, so
this will catch  messages such as:
The most obvious way to use these would be to check the incoming topics is it Light1 or Light2, then check the incoming payload (is it “ON” or “OFF”) and then act upon the content of the payload.

Especially if you have MQTT  topics with varying lengths this requires quite some checking of possibly long topics.
Say you have:

Although it is very well possible to check for all these messages and then determine what the payload is, it is memory consuming.
A more elegant way is to keep the MQTT messages all the same length and only say vary the last 2 characters of the topic to indicate what it is about. That gives room for 100 different topics, while uou only have to check the last 2 characters, rather than a long string.
Lets give an example:
Suppose your MQTT messages all have the format:  “home/nodeID/devID” or -in practice:

Your subscription then would be to: “home/node13/#”
and you would only have to check on the last two digits.

An example of  your callback then would be

void callBack(char* topic, byte* payload, unsigned int length) { // receive and handle MQTT messages
int i;
for (int i = 0; i < length; i++) {
if (strlen(topic) == 17) { // correct topic length ?
DID = (topic[15] - '0') * 10 + topic[16] - '0'; // extract device ID from MQTT topic
payload[length] = '\0'; // terminate string with '0'
String strPayload = String((char*)payload); // convert to string
if (length == 0) {
error = 2; // no payload sent
}else {
if (DID == 0) {
device0 = true;
if (DID == 1) {
device1 = true;
if (DID == 2) {
device2 = true;
if (DID == 99) {
device99 = true;

What the code does here is to just look at the last 2 digits (DID = Device ID) and then set a flag (device0 -device 99) that can be checked in the main loop upon which action can be taken, depending on the content of  the ‘strPayload’  string.

For an extensive example of this check the gateway projects of Computourist on github.

OpenHab GPIO Example

I know this might not directly be Arduino or ESP8266 related, but I thought I’d publish it anyway as it may come in handy with openhab

First of all, install the GPIO binding in the PaperUI. After one has done that, there isnt really much else to do on that binding. There is no services/gpio.cfg although some of the documentation mentions that file.

Before I go further it is good to note that I used the openhabian distro and I presume a lot of things are already OK in that, but if you did a manual install, it might be good to check if the user ‘openhab’ is member of ‘gpio’ and if not, set that with

sudo adduser openhab gpio

That’s basically all one needs to do.
Let’s setup an ‘input’ which in openhab is a ‘Contact’ and an ‘output’, which in openhab is a Switch. In fact, those are the only states the binding allows.

There is no pressing need to define anything in your sitemap file if e.g. you add the items to a group in your itemsfile, but if you prefer a sitemap then add this to your sitemap file:

Switch item=GPIO_LAMP

Let’s pick some pins to use. It is easiest to use pins on the Raspberry header that are close to a ground pin.

Looking at the pin out of the raspi, we see that for instance GPIO pins 21,13,5, 7, 10, 27,22,24,25,12,16,26, 3, 4, 8 and 17 are next to a ground pin, but some of those have reserved functions so i stayed away from those.
I picked GPIO23 and GPIO24 as that one is also close to a 3V3 pin and when using that as input one can always decide later to make that have a HIGH or a LOW input.The configuration string for the binding is defined as follows:

gpio="pin:PIN_NUMBER [debounce:DEBOUNCE_INTERVAL] [activelow:yes|no] [force:yes|no] [initialValue:high|low]"

for those who consider this as double dutch, it will become clear in the items definition:

Add this to your items file:

Switch GPIO_LAMP "Lamp" (LivingRoom) { gpio="pin:23 force:yes" }
Contact GPIO_BUTTON "Button [%s]" (LivingRoom) { gpio="pin:24 activelow:yes" }

The group ‘LivingRoom’ is not mandatory if you have defined your sitemap and of course you may use another group that suits you more.

The Switch that I set up is quite simple. In fact it already would be enough if it would have been:

Switch GPIO_LAMP "Lamp" {gpio="pin:23"}.

But with ‘force:yes’ we force the use of the pin even if it would have been in use by another process. Obviously that might not be the most elegant way to do it, but as i am pretty sure I am not using it for another process, I just wanted to make sure Openhab would have the pin available.

The current of a GPIO pin is far to small to drive any significant load as it can deliver some 60mA I think.
But the output can drive a small LED to test it… but use a 330 ohm series resistor, no matter what other websites tell you: you dont want to fry your GPIO pin. Eventually it would be more useful to drive a small relay module as long as it has a separate transistor to drive the relay (I know, I am cautious)

The configuration allows for some other parameters.
i already mentioned ‘force:yes|no

activelow yes|no  this means that when the switch is off, there is no voltage on the pin
initialValue:high:low determines the state the pin will take during initialisation.
debounce:interval is used to set the debounce value for the switch in milisecs

If we look at the output, I have not added an activelow because I thought that was the default value (but correct me if I am wrong)
If we look at our input, I set that to activelow, because the way the button works is it applies voltage to trigger the pin and I added it because I was not sure about the default
I did not set a debounce for the contact, but you could state:

Contact GPIO_BUTTON "Button [%s]" (LivingRoom) { gpio="pin:24 debounce:10 activelow:yes" }