Sensor Placement: Where Readings Go Wrong
Most control problems blamed on a sensor are placement problems. A probe reads whatever is at its tip, and the tip is usually somewhere convenient rather than somewhere representative. Hydroponic sensor placement starts from a simple question: what should this number describe — the solution at the root, the air at the canopy, or the water leaving the last channel? This guide answers that per sensor, then covers sampling position, immersion depth, fouling and climate sensor height.
What Each Sensor Must Represent
Write down what each reading is supposed to mean before choosing where it goes. The list below is the version used on commissioning drawings, and it resolves most arguments about EC probe location before they reach site.
| Sensor | What it must represent | Where that puts it |
|---|---|---|
| EC | The solution the roots are in | Return line or sample loop, after mixing, before fresh dilution |
| pH | Mixed solution, not the acid slug | Same loop as EC, 10 to 30 seconds downstream of injection |
| Nutrient temperature | Root-zone temperature | Tank, plus one probe at the furthest channel end |
| Dissolved oxygen | Aerated solution in motion | In-line with flow, or tank with circulation past the probe |
| Air temperature and RH | Air at the canopy | Crop height, in an aspirated radiation shield |
| CO2 | Air the leaves are in | Canopy height, away from the injection nozzle |
| PAR | Light reaching the crop | Canopy level, levelled, cleaned on a schedule |
The distinction that matters most is between supply and return: the supply number tells you what you dosed, the return number what the crop did with it. Confusing them is why a controller holds a perfect setpoint while the crop drifts.
Return Line, Tank or Inlet: Where to Sample
Sample the return line for control and the tank for verification. The return carries the integrated signal from the whole zone, while the tank is where a handheld meter can be dipped for a cross-check.
Distance from the dosing point decides whether the reading is of the mix or of the dose. Put the probe at least 10 to 20 pipe diameters downstream of injection, or allow 10 to 30 seconds of travel time so acid and stock have mixed. A probe a metre past the injector reads pH spikes that exist nowhere else.
Avoid the two classic bad positions: a dead leg, where the probe sits in a tee that nothing flows through, and the top of a pipe, where an air pocket forms and the probe dries and wets by turns. A sample loop teed off the main with its own isolation valve and drain is the standard fix, and it makes calibration possible without draining anything.
Probe Depth, Flow and Fouling
Flow past the sensing element is what makes a reading repeatable. Most in-line pH and EC probes want a minimum flow in the region of 0.1 to 0.3 m per second across the tip, and dissolved oxygen probes are more demanding still, since a membrane reads the water moving past it rather than the water it is sitting in.
In a tank, depth is the equivalent question. Keep the tip roughly 150 to 300 mm below minimum water level and at least 100 to 150 mm clear of the floor. A probe that spends part of the day in air drifts; one sitting in silt reads a local environment, not the tank.
Fouling is the other half. Biofilm, salt crust and root debris coat a tip long before the reading looks wrong, so the trend matters more than the instant value. Cleaning intervals, storage solution and replacement timing are set out in the EC, pH and DO probe maintenance guide.
Climate Sensors: Height, Shade and Aspiration
A climate sensor measures the air around its own housing. Put it where that air is the air the crop is in, and protect it from everything that is not the crop.
- Height: at canopy height for the crop being grown, rising with the crop. Add a second sensor at vent or gutter height where stratification is suspected.
- Shade: in a radiation shield, never in direct sun. An unshielded sensor in sun can read several degrees high, and the error moves with the hour of the day.
- Aspiration: a small fan drawing 2 to 3 m per second over the sensor keeps the reading honest in still, humid air. Passive shields lag badly in a closed house.
- Siting: away from heater discharge, fan outlets, pad faces, doors and any wall that stores heat.
- Density: one sensor per zone is common, with zones of roughly 200 to 500 m². Map the house once with a handheld logger before trusting a single point.
Temperature and humidity together drive vapour pressure deficit, and both must come from the same place at the same time. Two sensors in different microclimates produce a plausible VPD that describes neither.
Placement Mistakes That Skew Readings
The same handful of errors accounts for most of the bad data seen on site.
- Measuring the pipe, not the water: strap-on or un-insulated sensors pick up pipe wall temperature, which tracks the building more than the solution.
- Probe above minimum level: reads air whenever the tank drops, then recovers and looks like a process upset.
- Sampling after dilution: a probe downstream of a fresh make-up tee reads the blend, not the crop.
- One probe for a house with gradients: inlet and outlet ends of a long channel run differ; the crop sees both.
- Sensor near a door or a fan: every opening produces a spike that the controller then chases.
- No isolation valve: the probe cannot be removed for calibration without shutting the zone, so it never gets calibrated.
A probe is only as good as the sample reaching its tip. The failure mode we keep meeting is a controller fighting a setpoint that is correct, fed by a probe installed in a dead leg two metres from the injector. Fit a sample loop with flow, an isolation valve and a drain, and half the control problems disappear before any tuning.
Labelling, Access and Calibration Routes
Placement includes getting to the probe later. A sensor installed where nobody can reach it will not be calibrated, and an uncalibrated sensor is worse than none because its numbers still look credible. Design the access at the same time as the position.
| Task | Typical interval | What placement must allow |
|---|---|---|
| Visual and logged check | Daily | Readable or on the network; reachable without tools |
| Rinse and clean the tip | Weekly to monthly | Isolation valve and drain on the loop |
| Handheld cross-check | Weekly | Sample port next to the probe |
| Two-point calibration | Monthly to quarterly | Probe removable without draining the tank |
| Probe replacement | Annually or as drifted | Labelled cable, spare of the same model held |
Label both ends of every cable with the point ID used in the controller, and keep a schedule at the panel. Where the network allows it, feed readings into the monitoring system rather than a notebook; the IoT and remote monitoring overview covers what is worth logging, and the alarm thresholds guide covers how a badly sited probe creates nuisance alarms.
Solution chemistry is the other half: what EC and pH measure, and how temperature compensation moves the number, is set out in the EC, pH and PPM grower’s guide. Sampling taps follow the same isolation logic as the pipework and manifold layout rules.
Frequently Asked Questions
Where should an EC probe be installed?
A: In the return line or a sample loop taken off it, downstream of dosing and upstream of any fresh make-up dilution. It should see mixed solution at the flow the probe needs.
How deep should a tank probe sit?
A: Roughly 150 to 300 mm below the lowest operating level, and at least 100 to 150 mm clear of the floor. That keeps it wet at low level and out of the sediment.
Should pH and EC share the same location?
A: Usually yes, in the same sample loop, provided both sit far enough downstream of injection for the dose to mix. Ten to 30 seconds of travel time is a safer rule than a pipe length.
How high should a climate sensor be mounted?
A: At canopy height, moved up as the crop grows, inside an aspirated radiation shield. Keep it away from heaters, fans, doors and pad faces.
How many sensors does a greenhouse need?
A: One climate point per zone, with zones around 200 to 500 m², plus a probe at the furthest channel end for temperature. Map the house once to confirm the single point is representative.
Request a Monitoring Point Layout Quote
Hydroponic sensor placement belongs on the drawing beside the pipework, not added after commissioning when the probes arrive. Decide what each point must represent, give each one flow and isolation, and leave a route to reach it with a calibration kit. Send the zone layout, dosing arrangement, tank sizes and controller type, and the engineering team will return a monitoring point schedule with positions, sample loops and access notes.
To request a monitoring point layout quote, include the EC probe location you are using today and any reading you do not trust, since that is usually where the survey should start.
Planning a Hydroponic Project?
Need help choosing a growing system or equipment? Tell us about your project requirements.
