How to Test a Potentiometer Position Sensor: Readings, Mistakes, and ESP32 ADC Setup

A good potentiometer position sensor reads its exact stamped resistance (e.g., 10.0kΩ ±10%) across the outer pins, and sweeps smoothly from 0Ω to max resistance on the wiper without infinite (OL) dropouts. When powered, the wiper voltage should scale linearly from 0V to your reference voltage (usually 3.3V or 5.0V). If your CNC joystick, 3D printer Z-axis, or robotic throttle is stuttering, the fault usually lies in a worn carbon track or a parallel loading issue in your circuit. Here is exactly how to isolate the fault using a digital multimeter (DMM) and verify the analog output for your microcontroller.

Meter Setup and Safety Categories for Potentiometer Testing

Before you touch the probes to the terminals, you must configure your meter correctly and verify the safety category of the environment. Testing a 5V sensor on a breadboard is vastly different from testing a speed-reference potentiometer on an industrial motor drive.

Meter Setup Block

• Dial Position: Set to Resistance (Ω) for unpowered track testing, or DC Voltage (V⎓) for live signal tracing.
• Lead Jacks: Black lead in COM, Red lead in V/Ω. (Do not use the A or mA current jacks; placing an ammeter across a voltage-fed pot will create a dead short and blow the meter's internal fuse).
• Range: Set to Auto-ranging. If using a manual meter, select the 20kΩ range for a standard 10kΩ pot, or the 20V range for a 5V/12V circuit.
• Zeroing: Touch the probe tips together. Note the lead resistance (usually 0.1Ω to 0.4Ω). Subtract this from your final readings if testing low-resistance sensors (e.g., 100Ω wirewound pots).

Step-by-Step Probe Placement and Expected Readings

A standard 3-terminal rotary or slide potentiometer position sensor has two fixed end terminals (Pin 1 and Pin 3) and a moving wiper terminal (Pin 2). The physical placement of Pin 1 and Pin 3 dictates whether clockwise or counter-clockwise rotation increases the voltage, but electrically, the outer pins form the fixed resistor.

1. Isolate the Component: For accurate resistance testing, desolder or unplug the sensor from the circuit. Measuring in-circuit yields parallel resistance values that will mask an open track.
2. Test the Fixed Track (Pins 1 and 3): Place probes on the two outer pins. Rotate the shaft; the reading should not change.
3. Test the Wiper Sweep (Pin 2 to Pin 1): Place one probe on the wiper (middle pin) and the other on an outer pin. Slowly rotate the shaft through its full mechanical travel.
4. Test for Wiper Noise: While sweeping, watch the meter for sudden jumps to 'OL' (Open Loop) or massive resistance spikes. On an oscilloscope, this looks like vertical noise spikes, indicating a dirty or worn carbon track.

Expected Reading Table: Good vs. Bad Values

Common Mistakes That Give Misleading Readings

Even with a high-quality meter like a Fluke 87V, bench techniques can lead you to misdiagnose a perfectly good potentiometer. Watch out for these specific traps:

1. In-Circuit Parallel Loading

If you measure the resistance of a 10kΩ position sensor while it is still soldered to a PCB, the multimeter's test current will flow through the parallel paths of the microcontroller's ADC input impedance and pull-down resistors. If there is a 10kΩ pull-down resistor on the wiper, your meter will read 5kΩ across the track, leading you to falsely conclude the potentiometer is damaged. Always measure resistance out-of-circuit.

2. Ignoring the Taper (Audio vs. Linear)

If your position sensor reads 8kΩ at the exact mechanical midpoint, it might not be broken—it might just be an Audio taper (logarithmic, usually marked 'A10k') rather than a Linear taper (marked 'B10k'). Audio pots are designed for volume controls, not positional feedback. Using an audio pot for a CNC axis will result in severe non-linear movement mapping in your firmware.

3. Ghost Voltages on High-Impedance Meters

Modern DMMs have an input impedance of 10MΩ. If the wiper is disconnected from VCC but running parallel to a noisy AC line or switching power supply, capacitive coupling can induce a 'ghost voltage' of 1V to 2V on the meter. If you see a voltage reading on a pot that is supposedly unpowered, switch your meter to a Low-Z (Low Impedance) mode if available, or place a 10kΩ load resistor across the probes to bleed off the phantom charge.

Interfacing and ADC Verification on the ESP32

Translating a smooth analog voltage into a digital position value introduces a new layer of complexity, particularly if you are using the ubiquitous ESP32-WROOM-32. The ESP32's internal 12-bit Analog-to-Digital Converter (ADC) is notoriously non-linear and has specific hardware quirks you must account for when debugging a 'jittery' position sensor.

• The 3.1V Ceiling: Even though the ESP32 operates at 3.3V, the ADC cannot read all the way to 3.3V. The maximum readable voltage is typically around 3.1V to 3.15V. If your potentiometer is fed by a 3.3V LDO, the final 5% of your physical rotation will read as a flatlined maximum value (4095). Fix: Use a voltage divider to drop a 5V pot output to 3.0V max, or calibrate your software limits.
• ADC1 vs ADC2 Conflicts: Only use GPIO pins connected to ADC1 (GPIO 32, 33, 34, 35, 36, 39) for position sensors. ADC2 pins (GPIO 0, 2, 4, 12-15, 25-27) are shared with the WiFi radio. If your ESP32 connects to WiFi, ADC2 reads will fail or return garbage data, making your position sensor appear broken.
• Software Oversampling: Carbon track pots generate micro-voltage noise. Implement a software moving average filter or use the ESP-IDF adc_oneshot driver with hardware multi-sampling enabled to smooth out the physical wiper noise before feeding the data to your PID control loop.

For deeper technical specifications on the ESP32's analog peripherals and attenuation settings, refer to the official Espressif ADC Oneshot Driver Documentation.

Frequently Asked Questions

Why does my potentiometer position sensor read jump around on the ESP32?

Jumping readings are usually caused by one of three things: physical wear on the carbon track causing momentary open circuits (wiper noise), a lack of a hardware low-pass RC filter (a 0.1µF capacitor between the wiper and GND works wonders), or reading from an ADC2 pin while the ESP32's WiFi radio is actively transmitting. Verify you are using an ADC1 pin (like GPIO 34) and add a software moving-average filter of at least 16 samples.

Can I test a potentiometer position sensor while it is connected to the circuit?

You can test it for voltage while powered and connected, which is the best way to verify the signal reaching the microcontroller. However, you cannot accurately test it for resistance while connected. The multimeter's test current will flow through the microcontroller's input protection diodes and parallel bias resistors, giving you a falsely low resistance reading that might trick you into replacing a perfectly good sensor.

What is the difference between a linear (B) and audio (A) taper position sensor?

The taper defines the mathematical relationship between the shaft angle and the resistance output. A linear taper (marked with a 'B', e.g., B10K) changes resistance at a constant rate; at 50% rotation, you get exactly 50% of the total resistance. An audio taper (marked with an 'A', e.g., A10K) is logarithmic, changing slowly at first and rapidly at the end, designed to match human hearing perception. For embedded position sensing, joysticks, and robotics, you must always use a linear (B) taper.

How do I fix a scratchy or noisy carbon track potentiometer?

If the resistance sweep shows minor dropouts but the track isn't completely worn through, you can sometimes salvage it by spraying a specialized contact cleaner (like DeoxIT D5) into the casing slot and rotating the shaft back and forth 50 times to scrub the wiper contacts. However, for safety-critical or high-precision applications like CNC Z-axis probing, replace it with a sealed, precision wirewound or Hall-effect (magnetic) position sensor to eliminate mechanical wear entirely.