{"id":13356,"date":"2022-01-04T16:15:44","date_gmt":"2022-01-04T16:15:44","guid":{"rendered":"https:\/\/wolles-elektronikkiste.de\/wireless-bbq-thermometer-433-mhz"},"modified":"2024-10-29T15:21:28","modified_gmt":"2024-10-29T15:21:28","slug":"wireless-bbq-thermometer-433-mhz","status":"publish","type":"post","link":"https:\/\/wolles-elektronikkiste.de\/en\/wireless-bbq-thermometer-433-mhz","title":{"rendered":"Wireless BBQ thermometer (433 MHz)"},"content":{"rendered":"\n

About this Post<\/h2>\n\n

Normally, my posts are of a more basic nature, but here I would like to present a very concrete project of mine, namely a wireless BBQ thermometer with graphical display of the temperature curve.<\/p>\r\n

In winter, I find it especially nice to have a good piece of meat on the grill while sitting inside, drinking a wine and watching the temperature of the food. And if you can do the latter with your self-made device, it’s even more fun. Do you need a graphical display? No, of course not, but it’s exciting, and you can show off!<\/p>\r\n\n

Probe for the wireless BBQ thermometer<\/h2>\n\n

To build a BBQ thermometer, you first need a suitable temperature probe. You can get these under the search term “replacement probe” for 5 to 15 euros in online shops such as Amazon.<\/p>\r\n\n

\"Various<\/a>
Various temperature probes<\/figcaption><\/figure>\n
\n
\"\"<\/a><\/figure>\n<\/div>\n

In most cases, the probes have a 2.5 or 3.5 mm jack plug as connection. Whether the plug has two or three poles does not matter because usually only two of them are active. There are corresponding sockets that you can install in the housing of the transmitter unit of your thermometer.<\/p>\r\n\n

The probes are based on a temperature-dependent resistor. Unfortunately, the resistance value does not change linearly with temperature and, moreover, each model is different. Calibration is therefore unavoidable. To do this, record a temperature-resistance curve with a known and reliable thermometer. A first test with my multimeter showed that the resistance at 0 \u00b0C was 330 k\u03a9. At room temperature it was about 120 k\u03a9, at 100 \u00b0C it dropped to 6.3 k\u03a9. This is far from any linearity.<\/p>\r\n\n

Measurement of the thermometer resistance<\/h3>\n\n

Theory<\/h4>\n\n

Theoretically, you can measure the resistor values for the temperature-resistance curve directly with a multimeter. In practice, it is better to measure the resistor values as you do later in your circuit. The advantage is that this way you can compensate for errors such as possible non-linearity of your A\/D converter.<\/p>\r\n

The basis for determining resistance is Ohm’s law:<\/p>\r\n\n

<\/p>\r\n

  <\/span>   <\/span>\"\[<\/p><\/p>\r\n\n

Voltages can easily be measured with ADCs, currents are more complicated, especially as they are extremely low for these resistance dimensions. Therefore, I use a second, known resistor (reference), which I connect in series with the thermometer. Since the current is the same for both resistors, the following applies:<\/p>\r\n\n

  <\/span>   <\/span>\"\[<\/p><\/p>\r\n\n

In order to achieve a good resolution over the entire range in the temperature measurements, the size of the reference resistor should correspond to the thermometer resistance somewhere in the middle of the temperature measurement range. I selected 33 k\u03a9.<\/p>\r\n\n

Practice<\/h4>\n\n

I wanted to measure as accurately as possible and therefore chose a reliable, external ADC, namely the ADS1115<\/a>. A few years ago, I built a version 1.0 of this BBQ thermometer using the internal ADC of an Arduino Nano. With this, I was able to achieve an accuracy of +\/- 1 degree. So this is also possible and there is less soldering to do later – your choice.<\/p>\r\n

I used the following circuit to record the calibration curve:<\/p>\r\n\n

\"\"<\/a><\/figure>\n\n

So, here I measure the total voltage and the voltage drop across the reference resistor. Uthermometer<\/sub> is the difference.<\/p>\r\n\n

The temperature sensor is represented in the circuit diagram by the connection socket. Add it to your thoughts. And here is the sketch:<\/p>\r\n\n

\r\n
#include<ADS1115_WE.h> \r\n#include<Wire.h>\r\n#define I2C_ADDRESS 0x48\r\nconst float constResistorVal = 32.8; \/\/ 32.8 kOhm\r\n\r\nADS1115_WE adc = ADS1115_WE(I2C_ADDRESS);\r\n\r\nvoid setup() {\r\n  Wire.begin();\r\n  Serial.begin(115200);\r\n  if(!adc.init()){\r\n    Serial.println(\"ADS1115 not connected!\");\r\n  }\r\n  adc.setMeasureMode(ADS1115_CONTINUOUS); \/\/ Continuous Mode\r\n  adc.setPermanentAutoRangeMode(true);  \/\/ Autorange Mode\r\n}\r\n\r\nvoid loop(){\r\n  float supplyVoltage = 0.0;\r\n  float resistorVoltage = 0.0;\r\n  float thermometerVoltage = 0.0;\r\n  float thermometerResistance = 0.0;\r\n\r\n  Serial.print(\"Supply Voltage [V]: \");\r\n  supplyVoltage = readChannel(ADS1115_COMP_0_GND);\r\n  Serial.println(supplyVoltage);\r\n  \r\n  Serial.print(\"Resistor Voltage [V]: \");\r\n  resistorVoltage = readChannel(ADS1115_COMP_1_GND);\r\n  Serial.println(resistorVoltage);\r\n\r\n  Serial.print(\"ThermometerVoltage [V]: \");\r\n  thermometerVoltage = supplyVoltage - resistorVoltage;\r\n  Serial.println(thermometerVoltage);\r\n\r\n  Serial.print(\"Thermometer Resistance [kOhm]: \");\r\n  thermometerResistance = thermometerVoltage\/resistorVoltage * constResistorVal;\r\n  Serial.println(thermometerResistance);\r\n  \r\n  Serial.println();\r\n  Serial.println(\"-------------------------------\");\r\n  Serial.println();\r\n  delay(500);\r\n}\r\n\r\nfloat readChannel(ADS1115_MUX channel) {\r\n  float volts = 0.0;\r\n  adc.setCompareChannels(channel);\r\n  for(int i=0; i<10; i++){\r\n    volts += adc.getResult_V();\r\n  } \r\n  return volts\/10;\r\n}<\/pre>\r\n
 <\/p>\r\n<\/div>\r\n\n
If you want to use the internal ADC of your board, you have to modify the sketch accordingly.<\/p>\r\n
For the measurement, I simply brought a pot of water to the boil and then measured the resistance and temperature as it cooled. This is trickier than you might think – you quickly have temperature zones in the pot and should therefore stir. Also, the tips of the thermometers should not touch the bottom of the pot, as shown in the photo. I just didn’t have enough hands while taking pictures.<\/p>\r\n\n
\"Recording<\/a>
Recording of the calibration cures for the wireless BBQ thermometer<\/figcaption><\/figure>\n\n
In the beginning, the temperature drops quickly, then slower and slower. Therefore, after a while I added cold water in portions. Ich habe Eisw\u00fcrfel f\u00fcr den Bereich von Raumtemperatur bis 0 Grad verwendet.<\/p>\r\n
The quality of the calibration can only be as good as the quality of your reference thermometer. As can be seen in the photo above, the two barbecue thermometers differ by one degree. Don’t trust all devices just because they are digital!<\/p>\r\n\n
Evaluation of the calibration curve for the wireless BBQ thermometer<\/h2>\n\n
I entered the measured pairs of temperature and resistance into Excel and created a graph. Then I tried to put a regression curve through the values. In Excel, these are called “trendlines”. To display trend lines, left-click a data point, then right-click and select “add trendline”. Then you can choose which type of function you want to use. Check the box “Display Equation on chart”.<\/p>\r\n
Neither a logarithmic function nor a polynomial was a really good fit:<\/p>\r\n\n
\"\"<\/a>
Calibration curve 0 -100 \u00b0C with fourth order polynomial as trend line <\/figcaption><\/figure>\n\n
It was only when I divided the curve into three parts that I was satisfied with the trend lines. The diagrams are a bit small. Click on the images to enlarge them:<\/p>\r\n\n
\n
\n
\"\"<\/a>
0 – 40\u00b0C<\/figcaption><\/figure>\n<\/div>\n\n\n\n
\n
\"\"<\/a>
40 – 80\u00b0C<\/figcaption><\/figure>\n<\/div>\n\n\n\n
\n
\"\"<\/a>
80 -100\u00b0C<\/figcaption><\/figure>\n<\/div>\n<\/div>\n\n
Another hint: Do not blindly adopt the trend lines in your sketches. Especially with higher order polynomials, often not enough decimal places of the coefficients are displayed (as in the calibration curve 0-100\u00b0C above).  Right-click the formula in the chart and select the bars (Format Trendline Label) on the right side. There, you select the category “Scientific” and choose an appropriate number of decimal places.<\/p>\r\n
You should check the reliability and enter some resistance values in the trendline functions and compare the results with your readings. I tested the trendline functions this way:<\/p>\r\n\n
\r\n
float r = 331.0; \/\/ resistance\r\nfloat temperature = 0.0;\r\n\r\nvoid setup() {\r\n  Serial.begin(115200);\r\n  for(int i=0; i<140; i++){\r\n    r = r - 2.0;\r\n    temperature =  8.731*pow(r,4)*pow(10,-9) - 9.025*pow(r,3)*pow(10,-6)+ 0.00354*pow(r,2)- 0.709*r + 69.094; \r\n    Serial.print(\"R: \");\r\n    Serial.print(r);\r\n    Serial.print(\",   T:\");\r\n    Serial.println(temperature);  \r\n  }\r\n  Serial.println(\"********************\");\r\n  r = 60.0;\r\n  for(int i=0; i<49; i++){\r\n    r = r - 1.0;\r\n    temperature = -26.74*log(r) + 146.08;\r\n    Serial.print(\"R: \");\r\n    Serial.print(r);\r\n    Serial.print(\",   T:\");\r\n    Serial.println(temperature);  \r\n  }\r\n  Serial.println(\"********************\");\r\n  r = 13.0;\r\n  for(int i=0; i<15; i++){\r\n    r = r - 0.5;\r\n    temperature = -0.041*pow(r,3) + 1.3525*pow(r,2) - 17.51*r + 166.68;\r\n    Serial.print(\"R: \");\r\n    Serial.print(r);\r\n    Serial.print(\",   T:\");\r\n    Serial.println(temperature);  \r\n  }\r\n  Serial.println(\"********************\");\r\n}\r\n\r\nvoid loop() {\r\n}<\/pre>\r\n
 <\/p>\r\n<\/div>\r\n\n
If everything is in order, you have completed the main part of the wireless barbecue thermometer.<\/p>\r\n\n
The transmitter unit<\/h2>\n\n
Concept of the transmitter unit<\/h3>\n\n
In addition to temperature measurement, the transmitter unit shall:<\/p>\r\n