Exhaust Gaz Temperature (EGT) measurement

(Exhaust Gaz Temperature (EGT) measurement: Last updated by Benjamin on October 19,2022)

The Exhaust Gas Temperature (EGT) is one of the many parameters to consider when monitoring an aircraft engine’s operation. This temperature correlates to the air-fuel ratio (AFR) of the mixture introduced into the cylinders. The ideal air-fuel ratio is the one that ensures complete combustion. It is also the one that results in the maximum temperature of exhaust gases. A decrease in exhaust gas temperature is observed with a leaner or richer mixture compared to this ideal ratio. It is not our purpose here to discuss the interest, or otherwise, of EGT monitoring; the experts themselves may have divergent opinions on this subject. We will only present the measurement technique.


Given the temperatures to be evaluated, which can approach 900°C, the sensors used are thermocouples. The very simplified working principle of thermocouples is as follows: by heating the junction of two different metals, a voltage appears between their unjoined ends (fig. 1). This is called the thermoelectric effect. All pairs of dissimilar metals exhibit this effect.

Figure 1: Simplified working principle of thermocouples

There are many types of thermocouples, depending on the metals used. The ones that interest us here are type K thermocouples (fig. 2), where the metals used are Chromel and Alumel. They can assess temperatures up to more than 1300°C. The voltage generated is very low, around 40µV/degree C, with an almost linear response between 0°C and 1000°C, therefore almost proportional to the temperature. Such a low output voltage makes thermocouples particularly susceptible to electromagnetic interference, hence the importance of shielding in the form of a metal braid protecting the wires. The thermocouples we use are available here.

Figure 2: Type K thermocouple used in aeronautics for EGT measurement

Thermocouple amplifiers

Analog-to-digital converters cannot directly measure the very low voltages generated by thermocouples. Therefore, amplification is necessary. Thermocouple amplifiers also correct linearity errors as well. In addition, the amplifier circuit must also consider the junctions between the connectors (generally made of copper) and the Chromel and Alumel wires. These junctions (called cold junctions) also behave like thermocouples and generate parasitic voltages. This is called cold junction compensation. To make this compensation, the thermocouple amplifier integrates a temperature sensor that measures the temperature of the cold junctions. These must therefore be placed close to the amplifier. Chromel-copper and Alumel copper cold junctions require specific connectors (fig. 3).

Figure 3: Thermocouple connectors. On the left, a male connector for the thermocouple wires, and on the right, a female connector to be soldered on a printed circuit board.

Thermocouple amplifiers are either analog or digital. After cold compensation and linearization, an analog amplifier provides an output voltage (generally between 0 and 5 volts) that is strictly proportional to the temperature measured. The microcontroller must then convert this voltage to a digital value using an analog-to-digital converter. Then it computes this value to obtain the temperature. The AD8495 is a straightforward and efficient analog amplifier, and this is the one we use.

Digital amplifiers like the MAX31850, MAX31855, and MCP9600, to cite the most popular, directly output measurement in degrees, transmitted to the microcontroller via SPI, I2C, or 1-Wire bus, which seems simpler than with an analog amplifier. But some digital amplifiers, unfortunately, have a significant drawback, as they require ground-insulated thermocouples that are rare, hard to find, and more expensive. The thermocouples usually used on airplanes are seldom ground-insulated and, therefore, may be incompatible with these digital amplifiers. Indeed, to function correctly, they should neither share the same ground nor the general ground of the aircraft, which is impossible. For this reason, we have chosen to use the AD8495 analog amplifier on an Adafruit breakout board.

To monitor EGTs of the two rear cylinders (cylinders 3 and 4) of the Rotax 912, we have associated two AD8495 boards on a small printed circuit board, and two female thermocouple connectors, available here (fig. 4).

Figure 4: On the right, a blank printed circuit board; on the left, the same PCB is equipped with connectors and Adafruit AD8495 boards.

With a 5 volts power supply, the temperature range of the Adafruit board extends from -250°C to +750°C, which is not enough since the maximum temperature allowed for the Rotax 912 EGT is 880°C. The temperature range of the AD8495 circuit can extend from 0°C to 1000°C. A modification of the Adafruit board is therefore necessary. The goal is to remove the offset applied to the AD8495 REF pin via a TLVH431 shunt regulator and a 1k resistor. The purpose of this offset is to allow the measurement of negative temperatures, which is unnecessary here. The regulator and the resistor must be removed, and the REF pin must be grounded (Fig. 5 below). Given the small size of the surface-mounted components, a means of optical magnification is necessary, binocular loupes (mini x3.5) or stereo-microscope (max x10).

Figure 5: Modification of the AD8495 board. The 1k resistor and the regulator must be unsoldered, and the regulator must be replaced by a wire connecting the AD8495 REF pin to GND.

The voltage on the OUT pin is proportional to the temperature of the thermocouple. The AD8595 generates an output voltage of 5 mV/°C. With a 5 volts power supply, this output voltage ranges between 0 and 5 volts. With the above modification, the measured temperature of the Adafruit board ranges between 0 and 1000°C. The output voltage range should be reduced from 0-5V to 0-3.3V before connecting to an analog pin on the Teensy 4.1 board used in our EMS. The EGT input stage of the EMS can be seen in figure 6 below.

Figure 6: EGT input stage of the EMS.

This input stage includes two TLV271 operational amplifiers (Op-Amp). The first one(U5), supplied with 5 volts (like the maximum output signal of the AD8495), is configured as a follower (unity gain non-inverting op amp circuit or voltage buffer). It copies exactly the input voltage (pin 3) on its output pin (pin 6). Its extremely high input impedance prevents from loading the output impedance of the AD8495. Its almost zero output impedance makes it possible to supply the voltage divider R5-R6 without any risk of voltage drop linked to the resistors. The values ​​of R5 and R6 are chosen to convert 5 volts to 3.3 volts:

5 volts x R6/(R5+R6)=3.3 volts.

The second Op-Amp (U3) is supplied with 3.3 volts. It is also configured as a follower on the same principle. Its almost infinite input impedance prevents from loading the output of the R5-R6 voltage divider. Its almost zero output impedance is connected to an analog input pin of the microcontroller, protecting it from any risk of voltage greater than 3.3 volts.

The code that converts a voltage to a temperature in degrees C is as follows. It is an extract from our EMS software.

#define pinEGT3 A17
float EGT3;

void setup() {
 pinMode(pinEGT3, INPUT_DISABLE);

void loop() {
  // .......
 int digitalValue = analogRead(pinEGT3);
 EGT3 = (digitalValue *5/1023.0)/0.005;
  // ........

In our EMS, the output of the U3 Op-Amp is connected to the analog pin A17 of a Teensy 4.1. In the setup, we explicitly indicate that this pin is not intended to receive a digital signal. The instruction analogRead() on A17 results in a 10-bit value between 0 and 1023. In the following line, this value is first converted to a voltage between 0 and 5 volts (like the output of the AD8495), then this voltage is divided by 5 mV to obtain the temperature value. Remember that 5 mV on the output of AD8495 correspond to 1°C.

Leave a Reply

Your email address will not be published. Required fields are marked *

The maximum upload file size: 5 MB. You can upload: image, document, text, archive. Drop files here