Three different control parameters that get talked about like they're one dial marked "sensitivity." They aren't.
Ask someone to "tighten up" a zone that's swinging too far from setpoint, and they'll often reach for whichever of these three terms they heard most recently — without stopping to check which one actually applies to the control scheme in front of them. That mix-up is easy to make and genuinely consequential: a setpoint is simply the target value, a deadband is a no-action zone that only makes sense for on/off (two-position) control, and a throttling range is a proportional response curve that only makes sense for modulating control. Widening the wrong one doesn't just fail to help — it can make the loop behave nothing like what was intended.
The setpoint is the single value a control loop is trying to hold — a zone temperature setpoint of 72°F, a duct static pressure setpoint of 1.5 in. w.c., a supply air temperature setpoint of 55°F. On its own, a setpoint carries no information about how the system reacts when the measured value drifts away from it. That behavior is defined separately, by one of two entirely different mechanisms depending on what kind of control is actually driving the equipment: a deadband for on/off (two-position) control, or a throttling range for modulating (proportional) control.
"Where is control aimed?" is the setpoint's job, full stop — it carries no information about response behavior at all. "When does on/off equipment do nothing?" is the deadband's job, and it exists specifically to stop the equipment from short-cycling on every trivial fluctuation right at the setpoint — widen it and the equipment cycles less but tolerates more drift; narrow it and control tightens but wear and cycling increase. "How far does a modulating output travel as the measured value moves away from setpoint?" is the throttling range's job, and it describes a continuously proportional relationship that a deadband never has — narrow it and the loop gets sharper but more overshoot-prone; widen it and the loop gets smoother but slower to respond. A system using on/off control tuned via deadband behaves nothing like a system using modulating control tuned via throttling range, even sitting at the exact same setpoint value, because the two mechanisms govern fundamentally different kinds of equipment response.
False, and mixing them up leads to genuinely wrong tuning decisions. A deadband is specifically a no-action zone used with on/off (two-position) control: inside it, the equipment does absolutely nothing — stays off, or holds its last position — until the measured value exits the zone entirely. A throttling range is specifically the proportional response curve used with modulating control: across it, the output continuously changes in proportion to the deviation, with no zone at all where the loop simply stops responding. These describe fundamentally different kinds of control behavior — binary on/off versus continuous proportional — not two labels for one "sensitivity" knob. "Widening the deadband" on a system that is actually modulating via a throttling range isn't a smaller version of the right adjustment — it's adjusting a parameter the control scheme doesn't even use.
Explains why setpoint, deadband, and throttling range are three genuinely separate control parameters rather than interchangeable settings on one 'sensitivity' dial — a setpoint is simply the target value, a deadband is a no-action zone used with on/off (two-position) control to prevent short-cycling, and a throttling range is the proportional response curve used with modulating control, describing how far the output travels as the measured value moves away from setpoint.
A setpoint is the single value a control loop is trying to maintain — a zone temperature setpoint of 72°F, a duct static pressure setpoint of 1.5 in. w.c., a discharge air temperature setpoint of 55°F. By itself, the setpoint says nothing about how the system responds when the measured value drifts away from it. That response behavior is defined by a separate parameter, and which one applies depends entirely on whether the equipment is controlled on/off or modulated proportionally.
A deadband is a range around the setpoint within which no control action happens at all — the equipment simply stays off (or holds its last position) as long as the measured value is inside that range. Deadband exists specifically to prevent short-cycling: without it, on/off equipment would try to respond to every insignificant fluctuation right at the setpoint, starting and stopping constantly. A wider deadband means less equipment cycling and wear, at the cost of allowing the controlled variable to drift further before any correction begins. A narrower deadband tightens control but increases equipment starts and stops.
A throttling range is the span over which a modulating (proportional) control output — a valve position, a damper position, a variable-speed drive output — moves from fully closed/off to fully open/on as the measured value moves away from setpoint. Unlike a deadband, nothing is inactive across a throttling range: as the sensed value moves further from setpoint, the output moves proportionally further from its resting position. A narrow throttling range produces an aggressive, sensitive response that's prone to overshoot and hunting; a wide throttling range produces a smoother but potentially sluggish response. Because it describes a continuous proportional relationship rather than a no-action zone, a throttling range is not a substitute term for deadband, and the two should never be tuned as if they were the same knob.
Not for the same control action. A given output (say, a heating valve) is either controlled on/off, in which case it uses a deadband, or modulated proportionally, in which case it uses a throttling range — not both. A sequence with multiple stages, however, can combine them: for example, a modulating first stage with its own throttling range followed by an on/off second stage that engages with its own deadband once the modulating stage saturates.
Not automatically. A wider deadband reduces equipment cycling and the wear/energy cost associated with frequent starts and stops, but it also allows the controlled variable to drift further from setpoint before any correction begins, which can increase comfort complaints or, in some cases, increase runtime once a correction does start. The right deadband width is a tradeoff between cycling frequency and control tightness, not a one-directional energy lever.
A narrow throttling range makes the modulating output swing from 0% to 100% over a very small deviation from setpoint. That produces an aggressive, highly sensitive response that is more prone to overshoot and hunting — the output can overcorrect, drive the measured value past setpoint in the other direction, then overcorrect again, producing oscillation instead of stable control.
They describe closely related ideas but are not always identical terms. Throttling range (common in pneumatic and simpler proportional-only controllers) is typically expressed directly in engineering units of deviation from setpoint. Proportional band and proportional gain (common in PID controllers) express essentially the same proportional relationship, but as a percentage of the sensor range or as a numeric gain constant respectively. The underlying concept — output changes in proportion to error — is the same; the units and exact definition vary by manufacturer and controller type.
Try our Smart Buildings Studio
More calculators, simulators, and guides for this discipline.