Every analogue channel on a logger or ECU is a straight line: voltage in, engineering units out. Give it two points from the datasheet or a bench calibration and this works out the gain and offset, checks the fault limits, and shows what a sensor supply that is not exactly 5 V does to the reading.
Calibration points
Sensor supply check
Real Sensors, Real Numbers
Five common cases. Each button loads the figures into the calculator above so you can see the line, the fault limits and the check table for that sensor.
AiM 0 to 100 bar brake sensor
AiM publishes 0.5 V at 0 bar and 4.5 V at 100 bar for its 0 to 100 bar brake pressure sensor. It is powered from 8 to 16 V and regulates its own output, so it is not ratiometric to the logger’s 5 V and the supply check above does not apply to it.
The channel should read zero with the pedal off. If it reads a bar or two, that is usually the sensor’s own offset: re-zero point 1 with the brakes off rather than trusting the 0.5 V on the datasheet.
5 V ratiometric pressure sensor
A three-wire sensor powered from the logger’s 5 V, with a 0.5 to 4.5 V output that is a fixed fraction of its supply. This is an output option on Variohm’s EPT pressure sensors and many others. Take the range from the part number or the label: here 0 to 160 bar as an example.
Because the output follows the supply, a low 5 V rail moves the reading. The example loads a 4.90 V supply so you can see the error, and what happens when the logger measures against the same 5 V.
Linear damper pot
A linear potentiometer across the damper gives a fraction of its supply along its stroke. Datasheet end points are rarely exact once it is fitted, so calibrate it on the car: set the damper to two measured lengths and read the voltage at each.
Example readings: 0.48 V at 10 mm and 4.21 V at 110 mm of pot travel. This is damper travel; wheel travel needs the motion ratio on top. A pot can legitimately read close to 0 V or 5 V at its ends, so set the fault limits just outside your own end stops.
Throttle position, 0 to 100%
Throttle is logged as a percentage, so the two points are the stops, not a datasheet: read the voltage with the pedal released and again held flat to the floor, engine off. Those become 0% and 100%.
Example readings: 0.62 V closed and 4.31 V at full throttle. If the data never quite reaches 100% on the straight, the pedal is not opening the throttle fully, or the calibration was taken with the pedal and not the throttle at its stop.
Steering angle, through zero
A rotary pot on the column reads about mid-range with the wheels straight. Calibrate straight ahead as 0° and one known angle, such as 90° to the right, and the scaling runs through zero: right positive, left negative.
Example readings: 2.48 V straight ahead and 3.05 V at 90° right, so 1.91 V reads 90° left. Check straight ahead on a level floor with the car on its wheels, not on stands.
The AiM figures are AiM’s published characterisation for that sensor. The other ranges and readings are illustrations, not the figures for any particular part. Use your own sensor’s range, or better, your own bench calibration.
Check Table
Voltage against value across the full 0 to 5 V input range. Red rows are outside your fault limits.
| Voltage (V) | Value | Status |
|---|
Check it against something you know. A pressure sensor open to air should read zero gauge or about 1 bar absolute. A position sensor should read the same at the same mechanical stop every time. If the reading is off at a known point, fix the calibration before you start reading anything into the data.
Common Questions
Either the sensor datasheet or your own bench test. A datasheet usually gives the output range and the measuring range, for example 0.5 V at 0 bar and 4.5 V at 10 bar. Those are your two points.
A bench calibration is better if you can do it: apply two known values with a reference gauge or a known temperature, read the voltage at each, and enter those. It takes out the sensor’s own offset and gain error, and any error in the wiring and the logger input.
Because it is an absolute pressure sensor, reading atmosphere. Absolute sensors (bar a) read about 1 bar at sea level with nothing applied. Gauge sensors (bar g) read zero. Oil and fuel pressures are normally quoted as gauge pressure, MAP and baro as absolute.
If you have an absolute sensor on a gauge pressure channel, subtract atmospheric pressure in the logger, or use a calibration point taken with the sensor open to air. Check which type you have on the datasheet or the part number, not from memory.
Most three-wire sensors output a fraction of their 5 V supply rather than a fixed voltage. At 4.9 V supply, a reading that should be 4.5 V comes out at 4.41 V.
If the logger measures against the same 5 V it supplies to the sensor, the error cancels. If it measures against its own internal reference, or the sensor is powered from somewhere else, the error goes straight into the channel. The supply check on this page shows how big it is.
A healthy sensor never outputs outside its rated range, so a reading outside it is almost always the wiring. Depending on how the input is biased, a broken signal wire, a short to ground or a short to 5 V will pull the reading to one end.
Setting the logger’s alarm or diagnostic limits just outside the range lets you tell “no oil pressure” from “broken wire” at a glance. That matters when the alarm is an oil pressure warning.
Only for sensors with a linear voltage output, such as amplified thermocouple modules or sensors with built-in electronics. A plain two-wire NTC thermistor is not linear, so two points will be badly wrong between and beyond them. That needs a curve, not a straight line.