Six numbers and what each one is for.
Exhaust gas temperature. Not a structural limit and not a health figure — a mixture instrument. Its absolute value depends on where the probe sits in the pipe; only the shape of its curve means anything.
Cylinder head temperature. This one IS a limit, and the number that actually decides whether the engine reaches its overhaul life. Watch the hottest cylinder, not the average.
Too low is a fault too: oil must get hot enough to boil off the water that condensed in it, or it turns acidic. A gauge that never leaves the bottom of the green is telling you about your cowling, not your engine.
The fastest-acting warning on the panel. A pressure that falls while temperature rises is the classic signature of an engine about to end the flight.
The control input for leaning and, integrated over time, the most accurate fuel gauge in the aircraft — as long as somebody tells it what went into the tanks.
Together they set power. With a constant-speed propeller they are two levers for one number, which the propeller article takes apart in detail.
Why EGT has no red line.
Pilots meet exhaust gas temperature as a number with a scale and reasonably assume that lower is safer. It is not that kind of instrument. The exhaust leaving a cylinder is hottest when the mixture is close to chemically correct, and cooler both when there is excess fuel to evaporate and when there is excess air to heat. So the EGT curve has a peak, and both sides of that peak are cooler — one side because the engine is running rich, the other because it is running lean. Two very different conditions, one identical reading.
That is why the absolute value tells you almost nothing. It also depends on how far down the pipe the probe was welded — move the probe two centimetres and every number changes. What EGT is superb at is telling you where you are on the curve, which is the only thing you need in order to lean deliberately instead of by folklore.
Lean it and watch both curves.
Start rich and lean slowly. EGT climbs to its peak and then falls again — but notice what CHT does: it peaks slightly before EGT does, on the rich side, and then drops away steadily as you go lean. That offset is the whole of mixture management. Running well rich of peak is cool and thirsty; running properly lean of peak is cool and economical; the expensive place to live is the narrow band just rich of peak, where the head is hottest and the fuel saved is negligible.
Which side of the peak, and why anyone argues about it.
Two legitimate techniques exist for a carburetted or injected piston engine with a mixture control. Rich of peak keeps extra fuel in the cylinder as a coolant, gives the most power for the amount of air, and is what the manuals of the 1950s specified. Lean of peak puts excess air in instead, burns less fuel for a given true airspeed and runs the head cooler — but it demands that every cylinder receives a very similar mixture, or the leanest one will be starving while the richest is still rich.
That is where balanced injectors and a per-cylinder monitor earn their keep: with one probe you are guessing about five other cylinders, and with six you can see the spread. The rules that survive every argument are short. Do not lean at climb power. Do not sit just rich of peak at high power. And whatever the internet says, the engine manual for your engine is the authority — a modern light-aircraft engine with automatic altitude compensation and no mixture lever is not asking for your opinion at all.
The trend is the diagnosis.
Single readings catch dramatic failures. Recorded trends catch everything else, and everything else is what actually grounds aircraft. A cylinder whose EGT has drifted forty degrees away from its five neighbours over ten flights has an injector, a plug, a valve or an induction leak — and it said so long before it produced a rough-running magneto check. An oil temperature that has crept up over a season is a cooling problem, a baffle problem or an oil problem, and none of those improve by themselves.
This is the real argument for an engine monitor that logs. The panel display is for the flight; the log is for the maintenance conversation, where a graph is worth an hour of describing a noise. It is also the honest way to settle old hangar debates — shock cooling being the classic — with data from your engine rather than opinion about somebody else's.
Alarms that mean something.
An alarm set exactly at the manufacturer limit tells you that damage is happening now. An alarm set a little below it tells you that damage is about to happen, which is the only version that leaves room for a decision. The same reasoning applies to how alarms present themselves: a discreet colour change for a parameter drifting, a genuinely intrusive alert for oil pressure, and nothing at all for the things that are merely interesting.
Do that, and the engine page stops being twenty numbers to monitor and becomes what it should be — a quiet instrument that says nothing for two hundred hours and then says one specific, early, actionable thing.
The engine monitoring unit — temperatures, pressures and fuel on one bus, visualised on every display and logged with the flight.
Full-HD, 2500 nit screens where the engine page is readable in the same sunlight that washes out an analog gauge.
Logging, cloud sync and alerts — the flight and its engine record arrive on the ground with you.