吃瓜tv

An RTD transmitter PCB should preserve a known measurement circuit, not merely place an ADC close to a terminal block. Separate lead resistance, excitation, reference accuracy and thermal effects before routing. Then use dedicated bench experiments to find out which error source the board actually exhibits.

Define the measurement before placing components

Write down the RTD type, resistance at the reference temperature, wire configuration, temperature range and required uncertainty. Add the allowed cable length, connector arrangement, update rate, supply range and operating ambient. A specification such as 鈥�24-bit temperature measurement鈥� does not establish usable accuracy; ADC resolution is only one property of the measurement chain.

Distinguish a four-wire sensor from a two-wire transmitter powered by a current loop. Those descriptions refer to different parts of the system. A loop-powered transmitter can measure a three-wire or four-wire RTD, depending on its input circuit. Label both the sensor interface and the output interface explicitly in the schematic and quotation package.

Choose the topology from the ADC manufacturer鈥檚 documented configurations, including excitation routing, reference connections and input common-mode limits. TI鈥檚 RTD guide compares two-, three- and four-wire arrangements and explains their different assumptions. Keep the selected configuration next to the schematic review so that changes to protection or switching do not silently alter the measurement model.

Lead resistance: where the sense connection actually joins matters

In a two-wire measurement, cable and contact resistance are included with the sensor resistance. A three-wire circuit can compensate lead resistance under assumptions about the paths and measurement method; it does not make arbitrary unequal leads disappear. Four-wire sensing separates the excitation path from the voltage-sense path at the sensor terminals, subject to real input and leakage effects.

Do not turn a two-wire remote sensor into a claimed four-wire measurement by adding two traces only at the board connector. The shared cable resistance still lies between the takeoff and the sensing element. Draw the actual connector, cable and sensor junctions during review. A logical net name such as SENSE+ is not evidence of a correct physical Kelvin connection.

For an illustrative error screen, assume a local resistance-to-temperature slope of 0.385 惟/掳C. An uncompensated 0.20 惟 series contribution would then represent about 0.52掳C. This is a local approximation using an explicitly assumed slope, not a replacement for the sensor鈥檚 specified resistance curve. It is enough to show why a small contact or lead imbalance can outweigh a finely displayed decimal digit.

Make the reference path part of the precision layout

A ratiometric circuit compares the sensor voltage with a reference voltage produced by the related excitation path. The useful cancellation depends on the actual circuit, not the word 鈥渞atiometric鈥� on the schematic. Route the reference sense connections from the reference-resistor pads, separate from avoidable current-carrying copper. Inspect both takeoff points at close zoom before approving the layout.

Analog Devices discusses matched filtering in the signal and reference paths for RTD measurements. Use the selected ADC鈥檚 requirements to choose and verify those filters. Adding an arbitrary capacitor to make a noisy trace look quiet can change settling behavior or create unequal responses. Keep the component tolerances, leakage and source impedances in the analysis.

Document the reference resistor鈥檚 initial tolerance, temperature coefficient, power rating and actual dissipated power. Also document what calibration removes. An initial gain error measured at one ambient is different from drift during a warm operating condition. Locate heat-generating circuitry with this distinction in mind rather than assuming that the nearest available resistor position is the best precision location.

Check excitation headroom at the worst operating point

Evaluate the maximum sensor resistance, cable drops, protection elements, reference resistor and current-source compliance together. The excitation source needs its specified operating headroom. The ADC input and programmable-gain amplifier must also remain within their allowed differential and common-mode ranges. Check both low and high temperature, minimum supply and any supported sensor configuration.

Use a voltage worksheet rather than a single nominal schematic simulation. List each series drop and the resulting voltage at every relevant pin. Then compare with the limits in the actual ADC data sheet. If the design multiplexes channels or changes excitation routes, repeat the check for each state, including open-sensor detection and startup.

For a current-loop transmitter, maintain a separate power budget for the ADC, processor, reference, communications and output stage. A circuit that works from a bench supply may not meet the available operating budget in its intended loop condition. TI鈥檚 TIDA-00165 is a relevant reference architecture for reviewing this interaction; it is not a drop-in validation of a different board.

Calculate self-heating, then measure the real probe

Electrical dissipation in the RTD is I虏R. For a hypothetical 100 惟 sensor, 0.25 mA produces 6.25 碌W, whereas 1 mA produces 100 碌W. Increasing current by four increases electrical heating by sixteen. This calculation does not tell you the temperature rise without the probe鈥檚 thermal behavior in its actual medium and mounting.

If, solely for an example, the installed probe had a dissipation constant of 0.4 mW/掳C under the test conditions, those powers would correspond to approximately 0.016掳C and 0.25掳C steady-state rise. Do not use that assumed constant for another probe or for a different airflow or immersion condition. Obtain the applicable sensor data and verify the assembled measurement.

Reducing excitation can reduce self-heating while also reducing signal amplitude. Pulsed excitation introduces its own settling and timing requirements. Compare alternatives using both the thermal experiment and the ADC noise performance at the intended conversion settings. Record the time between enabling excitation and accepting a sample so that firmware changes cannot undermine the tested behavior.

Separate board temperature effects from sensor temperature

The RTD measures its own location; the electronics can drift with the transmitter board鈥檚 temperature. A regulator, output driver or power resistor near the reference network may create a changing local environment. Place precision components so that their operating conditions can be understood, and keep high-current return paths from sharing avoidable impedance with the measurement network.

Do not split a ground plane automatically because the circuit contains analog and digital functions. Follow the converter鈥檚 layout guidance and inspect return paths. A split beneath a signal can create a different problem from the one it was meant to solve. The relevant questions are where current flows, which voltage is sensed, and how interference reaches that measurement.

A board thermal-resistance calculation cannot by itself predict the error of a remote RTD channel. The difference between thermal conductivity and thermal resistance helps frame the heat-flow problem, while this layout review focuses on precision measurement. Keep measured board temperature, applied sensor resistance and reported temperature as separate columns in the test log.

Review protection, leakage and settling as a complete input network

Industrial input protection must satisfy the specified electrical environment, but protection components also affect the measurement. Review leakage over temperature, series resistance, capacitance, clamping behavior and recovery. A protection part chosen only by its headline transient rating may introduce an unacceptable error or disturb normal input operation.

Keep the input path understandable: connector, protection, filtering and converter. Provide the intended return path for disturbances without routing them through the precision reference takeoff. Confirm the chosen arrangement against the product鈥檚 safety, EMC and isolation requirements; this article does not prescribe clearances or establish compliance for a particular installation.

When switching channels, changing gain or enabling a diagnostic current, define which samples are discarded and when a valid reading is available. Inspect the ADC鈥檚 digital filter behavior as well as the external RC network. A stable-looking average can conceal a repeatable settling bias if firmware accepts readings too early after every transition.

Use a validation sequence that exposes the cause

Start with a precision resistor at the connector and record raw ADC codes, configured gain, reference value and calculated resistance. Test more than one resistance across the required range. Confirm the software conversion independently before connecting a temperature probe. This separates a units or coefficient error from a physical sensor problem.

Next introduce known lead resistance and deliberate imbalance. For a three-wire implementation, test the actual compensation sequence rather than a simplified spreadsheet alone. For a four-wire implementation, verify that sense and force leads are connected as released, including the harness. Keep the cable topology in the record so that an accidental bench shortcut does not become a misleading acceptance result.

Then hold the input resistance fixed while changing board ambient, output load and digital activity. Finally use the real probe in the intended medium and compare excitation settings after adequate settling. Assign separate pass criteria to these tests. A single room-temperature comparison cannot establish uncertainty over temperature, cable length and operating conditions.

Release the board with a reproducible test definition

The manufacturing package should identify the sensor interface, connector pinout, reference-resistor specification, populated variants and any cleanliness or handling requirements justified by the design. Include the firmware version, calibration method, test resistances, acceptable limits and required saved results. If fixtures or calibration services are needed, specify them as part of the build scope.

Use the industrial PCB planning page for the wider operating environment and the PCB assembly service for component sourcing, assembly and agreed inspection. Send the RTD board files and test requirements for review when you need a build proposal. State which accuracy claims are design targets and which have already been demonstrated; manufacturing an accurate layout does not by itself validate the complete measurement system.

Technical references: ; ; . Use the selected ADC and sensor data sheets for actual limits.