Field Calibration for the OHTS1022 Temperature Sensor: Zero Lab Equipment N

OrangeHorse Engineering Team
August 10, 2026
© OrangeHorse Electronic Technology Co., Ltd.

Field Calibration for the OHTS1022 Temperature Sensor: Zero Lab Equipment Needed

You’ve buried a dozen OHTS1022 sensors across a research plot. The moisture data looks solid, but you’re seeing a consistent +0.4°C offset in the temperature readings compared to your handheld reference probe. Is the sensor drifting, or is your reference wrong? Before you pull the sensor out of the ground and ship it back, there’s a faster path: field calibration using the sensor’s own RS485 interface and a simple two-point check.

The OHTS1022 uses a Class A PT1000 platinum resistance element with integrated signal conditioning. That gives you a baseline accuracy of ±0.2°C across the full -30 to 70°C operating range. But “baseline” assumes factory calibration conditions. In the field, you’re dealing with thermal gradients, soil moisture variability, and long cable runs. A quick offset check—and if needed, a zero-point adjustment—can restore confidence in your data without a lab bench.

What You’re Actually Calibrating

The PT1000 element itself doesn’t drift much. What drifts is the signal conditioning circuit, the ADC reference, and the mechanical coupling between the probe and the soil. The sensor’s measurement zone is a cylinder 7 cm in diameter and 7 cm high, centered on the central probe. That’s your thermal mass. When you calibrate, you’re calibrating the whole assembly—not just the element.

Here’s the practical reality: you’re not going to adjust the PT1000’s resistance curve in the field. You’re going to verify the sensor’s output against a known reference and, if the offset is consistent, apply a correction factor in your data logger or SCADA system. The OHTS1022’s Modbus registers give you raw temperature values; your correction happens downstream.

The Field Calibration Procedure

You’ll need:

  • A reference thermometer with a stated accuracy of at least ±0.1°C (traceable calibration preferred)
  • A container of water at a known temperature (ice bath for 0°C, or a stable ambient water bath)
  • A USB-to-RS485 adapter (if your sensor is wired for RS485)
  • A laptop with a Modbus polling tool (e.g., QModMaster, Modbus Poll, or a simple Python script)

Step 1: Prepare the sensor. Remove the OHTS1022 from the soil. Clean the probe shaft and the housing with a soft cloth. Do not submerge the entire sensor—the IP68 rating allows it, but you want the probe tip and the measurement zone fully immersed in your calibration bath, not the cable gland.

Step 2: Create a stable thermal environment. Fill an insulated container with water at a known temperature. For a zero-point check, use a well-stirred ice bath (0°C). For a single-point offset check at ambient, let the water equilibrate for at least 15 minutes. The sensor’s settling time is under 1 second, but the water bath needs time to reach thermal equilibrium.

Step 3: Read the sensor via Modbus. The OHTS1022 defaults to address 01, Modbus-RTU, 9600 baud (refer to the datasheet for exact register mapping). Poll the temperature register. Record the value. Simultaneously, record the reference thermometer reading.

Step 4: Calculate the offset. Offset = Sensor Reading − Reference Reading. If the offset is within your acceptable tolerance (say, ±0.2°C), no correction is needed. If it’s consistently off by +0.4°C, you have two options:

  • Apply a software offset in your data acquisition system (e.g., subtract 0.4°C from all readings)
  • Perform a two-point calibration if your logger supports linear correction

Step 5: Document and repeat. Log the date, sensor serial number, reference thermometer ID, and the offset. Repeat the check at a second temperature point (e.g., 30°C water bath) to verify the offset is linear. If the offset changes with temperature, you’re dealing with a gain error, not a zero error—that requires factory recalibration.

Wiring and Interface Notes

If your sensor is configured for RS485, the wiring is straightforward:

Wire Color (typical)Function
RedV+ (5–24V DC)
BlackGND
Yellow/WhiteRS485 A (D+)
Green/WhiteRS485 B (D−)

For 4–20 mA current loop output, the supply voltage must be 12–24V DC. The current output is linear with the measurement range—so 4 mA corresponds to the lower range limit and 20 mA to the upper limit. If you’re using the 0–2V, 0–5V, or 0–10V voltage outputs, note that the supply voltage requirement differs: 5–24V DC for voltage/RS485 outputs, but 12–24V DC for the 0–5V, 0–10V, and 4–20 mA configurations.

When Field Calibration Isn’t Enough

Field calibration handles zero and linear offset errors. It won’t fix:

  • Probe damage (bent or corroded 316L stainless steel probes)
  • Cable degradation (long runs over 1200 m may introduce signal loss—refer to the datasheet for cable length limits)
  • Housing failure (cracks in the ABS housing or epoxy seal compromise the IP68 rating)

If your offset is large (more than ±1°C) or unstable across readings, stop calibrating and inspect the physical integrity of the sensor. The epoxy resin sealing and IP68 rating protect against moisture ingress, but they don’t protect against mechanical damage from rocks or frost heave.

A Practical Takeaway

Field calibration isn’t about making the sensor “better”—it’s about quantifying the uncertainty in your measurement chain. The OHTS1022’s ±0.2°C accuracy is a factory spec. Your field accuracy is a function of installation, cable length, and soil contact. A simple two-point check with a reference thermometer gives you the data to state your actual measurement uncertainty with confidence.

Keep a log. Calibrate at the start of each season. If you’re deploying a new batch of sensors, spot-check 10% of them before burial. And if you need the full register map or wiring diagrams for your specific output configuration, the OHTS1022 product page has the datasheet. For custom ranges—say, a 0–50°C temperature range instead of the standard -30 to 70°C—the factory can configure that before shipping.

One more thing: if you’re running a multi-sensor network, don’t assume all sensors share the same offset. Each unit has its own signal conditioning path. Calibrate each one individually. The few minutes per sensor will save you weeks of questionable data later.

If you’re unsure about your specific output configuration or need help interpreting your Modbus register values, contact our support team with your sensor’s serial number and output type. We can walk you through the register map for your exact firmware version.

Contact our engineering team for application-specific deployment guidance.