Tech OVN

August 2026 · Pump Control & Protection

Pressure Sensor vs Float Switch for Pump Control

Float switches have controlled pumps for decades. They are cheap, simple, and well-understood — but they fail silently, offer no diagnostics, and cannot protect against dry-run when it matters most. Pressure sensors offer a different approach: continuous measurement, software-adjustable setpoints, and the ability to infer level, detect dry-run, and enable constant-pressure control from a single device. Here is how the two compare, and when each is the right choice.

How a float switch works

A float switch is a mechanical device. A buoyant float rides the liquid surface on a cable or pivot arm; as the level rises or falls to a setpoint, the float triggers a reed switch that closes or opens a contact. The controller sees a binary signal: on or off. High level and low level.

This simplicity is the float switch's main virtue. It requires no power supply to the sensor, no analogue input on the controller, and no calibration. Install it at the correct height, wire the two contacts, done.

The limitation is also in the simplicity. A float switch knows nothing between its two setpoints. It cannot tell you the actual level, the rate of change, or whether the pump is running on air. When the float fails — and it will, eventually — nothing tells you it has failed.

How a pressure sensor works for level and pump control

A pressure transducer measures the pressure of the fluid at its installation point and outputs a proportional analogue signal — typically 4–20 mA or 0–10 V. Installed at the bottom of a tank, it measures hydrostatic head: the weight of fluid above it. Since fluid density is known and constant for most clean-water applications, the pressure reading translates directly to level.

On the suction side of a pump, a pressure transducer measures inlet pressure. When the pump runs dry — drawing air instead of water — inlet pressure drops sharply below atmospheric or below a minimum threshold. The controller detects this within seconds and stops the motor before mechanical damage occurs.

On the outlet side, a pressure transducer provides live feedback for constant-pressure control. The controller compares actual outlet pressure to a target setpoint and adjusts pump speed (via a variable-frequency drive) or switches pump stages to maintain that setpoint continuously, regardless of flow demand fluctuations.

Unlike a float switch, the sensor output is a continuous analogue measurement — not a binary event. Setpoints are software parameters, adjustable without entering the tank or repositioning hardware.

Float switch failure modes — and why they are silent

Float switch failures are common in the operational environments where pumps work hardest.

Fouled or stuck float

In sumps, raw-water tanks, or any application with suspended solids, the float collects debris, scale, or biological growth. This changes its buoyancy — making it ride higher or lower than intended — or jams it at one end of travel. A float stuck in the “high” position prevents the pump from starting. A float stuck “low” runs the pump continuously until the tank empties and the motor overheats.

Tangled or kinked cable

In deep sumps or tanks with turbulence from the pump, the float cable wraps around pipework, the pump body, or other floats. The float is mechanically prevented from tracking the water surface. The controller never knows this has happened.

Contact wear and corrosion

The reed switch inside the float corrodes in humid or aggressive environments, or wears out after many thousands of switching cycles. A failed contact in the open position prevents the pump from starting; a failed contact in the closed position runs the pump regardless of level.

No diagnostics

The defining problem: there is no signal that says “float switch failed.” The controller sees the same binary contact it always saw — it cannot distinguish a float stuck at mid-level from a tank actually at mid-level. Discovery comes when someone investigates an overflow, a dry pump, or an unexplained trip — not from the control system.

Where pressure sensing wins

Dry-run detection

A suction-side pressure transducer detects a dry-run condition — the pump pulling air — within seconds of onset. The controller can stop the motor immediately, preventing impeller damage, seal failure, and overheating. A float switch on the tank cannot do this: by the time the level has dropped to the low-level float, the pump may already have been running partially dry.

Constant-pressure control

Pressure-boosting applications — domestic hot water, fire suppression headers, irrigation zones — need stable outlet pressure regardless of how many outlets are open simultaneously. A pressure sensor on the outlet combined with a capable pump controller enables PID-based pressure regulation: the controller continuously adjusts pump speed to hit the setpoint, eliminating the pressure swings a float-controlled on/off pump produces.

Level inference from a single sensor

One pressure transducer at the tank bottom gives a continuous level reading — not just two binary events. You can configure as many start/stop setpoints as needed, see the rate of level change, and log trends over time. Comparing daily level trends reveals patterns — abnormal drawdown, unexpected inflow — that a float switch cannot surface.

Trending, alerting, and remote visibility

With a sensor-based controller connected to a monitoring platform, you can see tank level and pump operating state remotely, receive alerts on abnormal conditions (low suction pressure, high outlet pressure, motor overcurrent), and review historical data. This is the foundation that AI-based condition monitoring builds on — pattern recognition over electrical and process signals to detect developing faults before they cause a breakdown.

Where a float switch is still fine

Not every pump application needs continuous measurement. A float switch remains an adequate choice when:

  • The application is a simple single-tank fill or sump empty with on/off pump control and no remote monitoring requirement.
  • The liquid is clean and the environment is benign — no solids, no aggressive chemicals, low ambient humidity.
  • The tank is accessible and regularly inspected, so a stuck float is caught during routine rounds rather than after a failure.
  • There is no need for dry-run protection at the pump inlet — for example, the tank is always kept above a safe level by a separate upstream system.

The decision point is monitoring. If you need remote visibility, level trending, or confident dry-run protection, the float switch cannot give you these. If you only need two setpoints and manual checks are reliable, a float switch costs less and installs faster.

Three-signal dry-run protection — why one signal is not enough

Suction pressure is a reliable dry-run indicator in most conditions — but not all. A partial blockage on the suction strainer can reduce flow without fully dropping pressure. A high-inertia pump takes longer to cavitate than expected, giving a misleading pressure reading in the first seconds of a dry-run event. And a pressure transducer can itself fail, giving a stable reading that masks a true dry condition.

This is why the Tech OVN pump controller correlates three independent signals simultaneously: suction pressure, motor current, and motor power signature. When a pump runs dry, the hydraulic load on the motor drops — and this appears as a characteristic change in both current draw and power consumption. Three signals agreeing that something is wrong is significantly more reliable than any single signal acting alone.

What three-signal dry-run protection catches that pressure alone misses:

  • 1. Partial dry-run — pump intermittently pulling air, pressure fluctuating but not fully dropping
  • 2. Sensor fault — pressure transducer reading stable while actual condition is dry
  • 3. Suction blockage — strainer fouled, reducing flow without a dramatic pressure drop

Frequently Asked Questions

Common questions about pressure sensors, float switches, and pump dry-run protection.

In most industrial and commercial pump applications, yes. A pressure transducer at the bottom of a tank measures hydrostatic head continuously, giving you a real-time level reading that lets the controller switch the pump at any level setpoint — not just the two hard points a float allows. The practical advantage is that you can adjust setpoints in software without entering the tank or repositioning a mechanical float. For a simple sump or single-tank fill application with no monitoring requirement, a float is still a perfectly adequate choice.
On the suction side, a dry-run condition causes pump inlet pressure to drop sharply — the pump pulls air instead of water, and the pressure transducer sees it within seconds. The controller compares suction pressure to a low-pressure threshold and stops the motor before it overheats. However, pressure alone is one signal. Tech OVN's pump controller correlates pressure readings with motor current and power signatures simultaneously — three independent signals — so a sensor fault or a partial blockage that doesn't fully drop pressure is also caught.
The most common failures are: (1) fouled or stuck float — debris, scale, or slime coats the float, changing its effective buoyancy or jamming it at one end of travel; (2) tangled or kinked cable — in a deep sump, the float cable wraps around the pump or pipework and prevents the float from rising or falling freely; (3) contact wear or failure — the reed switch inside the float corrodes or wears out after repeated switching; and (4) single-point failure with no diagnostics — when the float fails, nothing signals it, so the pump runs dry or runs continuously until someone investigates.
Constant-pressure control means the pump's speed or staging is adjusted continuously to maintain a target outlet pressure regardless of flow demand. A float switch cannot do this — it only knows 'tank high' and 'tank low'. A pressure sensor on the outlet gives the controller a live feedback signal, allowing it to modulate pump speed via a VFD or switch multiple pump stages to keep system pressure within a tight band. This matters in booster and pressure-boosting applications where pressure swings cause noise, valve wear, or inconsistent flow at the point of use.
Typically, no. A pressure transducer uses a 4–20 mA or 0–10 V analogue signal connected to a controller with an analogue input — most modern pump controllers and PLCs have these inputs. The mechanical installation is a threaded port on the pipe or tank wall. The main change is in the controller logic: instead of a binary level input, the controller now reads a proportional analogue signal and acts on it according to your setpoints. For new installations, specifying a pressure-sensor-ready controller from the outset avoids retrofitting entirely.
For a simple single-tank fill or sump-emptying application where all you need is 'start when low, stop when high' and you have no need for remote monitoring, trending, or diagnostics, a float switch is adequate and inexpensive. It requires no analogue wiring, no programming, and no calibration. The decision point is whether you need measurement — level data, dry-run confidence, pressure trending, or remote visibility. If you do, a sensor-based approach is the right one.

Replace your float switch with multi-signal pump protection

Tech OVN's pump controller fuses pressure, current, and power signatures for dry-run detection, constant-pressure control, and remote monitoring — no float switch required.