What it is

On-off control is one wire, one decision: energized or not. Modulating control needs a way to say "give me 40%... now 65%... now 100%." That's what a control signal does — it carries a continuously variable command from a controller (thermostat, DDC controller, board) to a device that can run at any output: an ECM blower, a modulating damper actuator, a variable gas valve, a VFD, a modulating boiler.

The three you'll meet constantly are 0-10V (or 2-10V) analog, 4-20mA current loop, and PWM (pulse-width modulation). They all do the same job — carry a 0-to-100% command — but they encode it differently, and that difference is where techs get tripped up.

How it works

0-10V analog (and 2-10V). The command is a DC voltage. 0V = 0% (off/minimum), 10V = 100% (full). Halfway, 5V, is roughly 50%. The variant 2-10V uses 2V as the bottom instead of 0V — so 2V = 0%/minimum and 10V = 100%. The reason 2-10V exists: a true 0V is ambiguous (is it "0% command" or "broken wire / dead controller"?), so starting at 2V lets the device tell the difference between "you asked for minimum" and "the signal died." Easy to read with a meter, but voltage drops over long wire runs and it's sensitive to noise.

4-20mA current loop. The command is a DC current, not a voltage. 4mA = 0%, 20mA = 100%. Current loops are the industrial favorite because current doesn't drop over long wire runs the way voltage does, and the "live zero" at 4mA means 0mA = a fault (broken wire), which is detectable. You read it with a meter in current/mA mode in series, or across a known sense resistor.

PWM (pulse-width modulation). Instead of a steady voltage level, PWM sends a square wave that switches fully on and fully off, and the percentage of time it's on (the duty cycle) is the command. 0% duty = off, 50% duty = half, 100% duty = full. The voltage only ever sits at "on" or "off" — the information is in the timing, not the level. This is extremely common on ECM blower motors and many small actuators because it's cheap, noise-immune, and easy for a microcontroller to generate. The catch for techs: a regular DC voltmeter shows you an average of a PWM signal, not the real picture, so PWM can read confusingly low or jumpy on a basic meter.

In every case the device interprets the signal against its own scaling. The controller and the device have to agree on the type (0-10V vs 2-10V vs 4-20mA vs PWM) and the direction (does max signal mean max output, or is it reversed?). Mismatch the type or the scaling and the device does the wrong thing even though the signal is "present."

In the field

  • Identify the signal type before you test. Check the device's terminals and literature: 0-10V, 2-10V, 4-20mA, or PWM? Each is tested differently, and assuming the wrong one wastes time.
  • 0-10V: read DC volts across the signal and its reference (common). Confirm it swings as the call changes — e.g., force more demand and watch the voltage climb toward 10V. Steady 0V (on a 0-10V device) or steady 2V (on a 2-10V device) with a demand present means the controller isn't driving it or the wire's broken.
  • 4-20mA: read in mA, in series with the loop, or measure the voltage across a known sense resistor and calculate. 4mA at minimum, 20mA at full; a reading of 0mA means an open loop (broken wire), which is the whole point of the live zero.
  • PWM: don't trust a plain DC voltmeter. A scope or a meter with a duty-cycle/frequency function reads it correctly; a basic voltmeter shows an average that looks like a low analog voltage. Many ECM-blower no-airflow calls get misdiagnosed here. If you must use a basic meter, know you're seeing an average and interpret accordingly.
  • Confirm polarity and common reference. Analog signals are referenced to a common; get the common wrong and you'll read nonsense. Many actuators need the signal common tied to the controller common.
  • Check the scaling/direction config. Some devices are field-set for 0-10V vs 2-10V, or for direct vs reverse acting. A damper that drives full open when you command minimum is usually a reversed-action setting, not a broken actuator.

Normal values & targets

  • 0-10V: 0V = 0%/min, 5V ≈ 50%, 10V = 100%. Linear in between.
  • 2-10V: 2V = 0%/min, 6V ≈ 50%, 10V = 100%. The 0–2V band is "dead/fault," not a valid command.
  • 4-20mA: 4mA = 0%, 12mA ≈ 50%, 20mA = 100%. 0mA = open-loop fault.
  • PWM: duty cycle is the command — 0% duty = off, 50% duty ≈ half, 100% duty = full. The carrier voltage and frequency vary by device (ECM motors commonly use a low-voltage PWM at a fixed frequency); the duty cycle is what matters.
  • Signal common: analog signals must share a common reference with the controller — verify it's landed.

Common faults & what they mean

  • Device stuck at minimum despite a demand: no signal arriving (broken wire, controller not driving), or signal common not connected. On 4-20mA a reading of 0mA confirms an open loop.
  • Device stuck at full / runs backwards from command: reversed action setting (direct vs reverse acting), or 0-10V vs 2-10V scaling mismatch. Check the device's configuration, not just the wire.
  • PWM-driven ECM blower "reads ~2-3V, must be dead": classic misread — a basic voltmeter averages the PWM pulses. Use a duty-cycle-capable meter or scope before condemning the signal or the motor.
  • Erratic/hunting output on a long run: voltage drop and noise on a 0-10V signal over a long cable. This is exactly the case 4-20mA was designed to beat — consider the signal type for the distance.
  • Signal present but device ignores it: type mismatch (controller outputs 0-10V, device expects 4-20mA or PWM), or the device isn't enabled. Confirm both ends agree on signal type.
  • Reads correct at the controller but wrong at the device: wiring fault, voltage drop, or a bad common between the two points. Measure at both ends.

Tech tips & gotchas

  • Know the signal type before you grab the meter. 0-10V, 2-10V, 4-20mA, and PWM each test differently. Half the battle is identifying which one you're looking at.
  • 2-10V's bottom is 2V, not 0V. On a 2-10V device, reading 2V means "minimum command," not "dead." Reading 0V means the signal itself failed. That gap is a feature — it distinguishes a valid minimum from a fault.
  • PWM fools a voltmeter. This is the single most common modulating-signal mistake. A basic DC meter averages the pulses and shows a misleading low voltage. Reach for duty-cycle/frequency mode or a scope, especially on ECM blowers.
  • 4-20mA wins over distance. Current doesn't sag over long runs and 0mA is a detectable fault. If a 0-10V signal is hunting on a long cable, the signal type may be the problem.
  • Direction and scaling are configurable. "Backwards" behavior is usually a reverse-acting setting or a 0-10 vs 2-10 mismatch, not a failed actuator. Check the config before you replace parts.
  • Mind the common. Analog signals are meaningless without a shared reference. A floating or wrong common makes good signals read like garbage.

Safety / code notes

  • Modulating control signals are low-voltage Class 2 and must be installed and protected per the applicable Class 2 wiring provisions; keep them separated from line-voltage conductors to limit induced noise and for code separation.
  • The devices these signals drive (VFDs, modulating valves, blowers) sit on line-voltage equipment — de-energize and verify zero energy at the disconnect before working on the power side, even when you're only chasing a low-voltage signal.
  • A failed modulating signal should drive the device to a safe default (typically minimum or a fail-safe position); verify the fail-safe behavior so a broken signal wire doesn't leave a damper or valve in a hazardous position.