A dry-block temperature calibrator and a decade resistance box both sit on the calibration bench, but they measure different physical quantities and resolve different error sources — pick the dry block to verify the temperature sensor's output against a known thermal source, and pick the decade resistance box to verify the readout instrument, transmitter loop, or simulation of an RTD/thermistor input [S1][S4].
For typical 2026-vintage bench work: a portable dry-block like the Digitron Italia CS 650 covers +30 to +650 °C at 0.1 °C resolution with ±0.7 °C accuracy from 30–200 °C and ±1.5 °C above 200 °C, weighs 6 kg, and ships with 11 probe holes plus RS232 output [S1]; a multi-decade resistance box typically covers 0.1 Ω to 111 MΩ in 0.1 Ω steps with ±0.1% accuracy and is used to inject a known resistance into a RTD Pt100 loop or a 4-wire ohmmeter under test [S4].
Core Definitions and What Each Instrument Actually Calibrates
A dry block temperature calibrator is a heat source: an aluminium or bronze block is heated (and sometimes cooled) to a setpoint, the sensor under test is inserted into a drilled well, and the sensor's indicated temperature is compared to the block's reference temperature read by the calibrator's internal sensor [S1][S2]. It is the right tool when the unknown is the sensor itself — a thermocouple junction drifting, an RTD element losing its characteristic curve, or a probe with a damaged sheath. Modern units run from sub-ambient dry blocks such as the DEARTO low-temperature smart series, which extends the operating envelope below ambient for pharmaceutical and cold-chain work [S6], up to 650 °C portable sources like the CS 650 [S1], and into higher ranges (e.g. 150 A-class bench units such as the Kds382-150A from KDS Instrument) for industrial heat-treatment lines [S5].
A decade resistance box, also written as decade resistance box or in Chinese 十进电阻箱 [S4], is a passive switched-resistor network arranged in decade steps (×0.1 Ω, ×1 Ω, ×10 Ω, ×100 Ω, etc.) so the user can dial a precise known resistance from the front panel. It is the right tool when the unknown is the readout or the loop: a panel meter, a temperature transmitter being commissioned, a dry-type transformer winding-resistance tester being checked, or a process simulator emulating an RTD input during a PLC loop test [S4].
Selection Criteria: Temperature Range, Accuracy, Resolution, Stability
For a dry block, the four spec gates are (1) temperature range vs. process, (2) display accuracy with its temperature band, (3) axial/radial uniformity well-to-well, and (4) stability over time at setpoint. The CS 650 lists accuracy of ±0.7 °C from 30 to 200 °C and ±1.5 °C above 200 °C, stability better than 0.1 °C, 25-minute heat-up, and natural cooling [S1]. A general-purpose bench unit such as those listed in the DwyerOmega dry-block catalog targets ambient +5 °C to 450 °C, with high accuracy quoted at ±0.4 °C and well-to-well radial uniformity to 0.020 °C, which is the metric that determines whether two probes inserted in different holes see the same temperature [S3]. For a decade box, the spec gates are (1) total resistance range, (2) step resolution (0.1 Ω is common, 0.01 Ω is precision grade), (3) stated accuracy per decade (typically 0.1% to 0.05% for mid-range boxes, ±1% on the lowest decades), and (4) residual resistance and thermal EMF of the internal switches, which dominate below 10 Ω [S4].
For an RTD loop specifically, both instruments end up in the same SOP — but at different stages. The dry block verifies the sensor (insert probe, read temperature); the decade box verifies the transmitter (inject known resistance equivalent to a temperature per IEC 60751 for a Pt100, e.g. 100.00 Ω at 0 °C, 138.51 Ω at 100 °C) [S4]. Engineers routinely do both, which is why the two appear on the same bench rather than being treated as alternatives.
Use-Case Decision Matrix: Which One Fits the Job
Use the comparison below to pick or to justify buying both. Three criteria scored from research-grounded facts: typical accuracy, primary physical quantity, and the sensor class each one covers. [S1]
1. Sensor suspect — thermocouple or RTD reads wrong: dry-block. Heat to 100 °C and 0 °C ice-point, compare reading to known block temperature. CS 650 covers 30–650 °C; DwyerOmega ambient +5–450 °C range covers most HVAC and process probes; DEARTO low-temperature smart units cover the –20 to ambient cold-chain segment [S1][S3][S6].
2. Transmitter or panel-meter suspect — loop is offset or non-linear: decade resistance box. Inject 100.00 Ω, 138.51 Ω, and 200.00 Ω to simulate a Pt100 at 0 °C, 100 °C, and ~266 °C. ±0.1% box accuracy translates to ~0.26 °C equivalent at 0 °C, well inside most transmitter specs [S4].
3. Field loop check on a 4–20 mA temperature transmitter: decade box with HART handheld. This is the single most common overlap scenario, and it is why both tools ship in the same calibration kit rather than either replacing the other.
4. Insulation resistance or winding resistance test on a dry-type transformer or motor: use an insulation resistance tester at high voltage, not a decade box — a decade box is a low-voltage precision source and cannot substitute for a 500 V/1000 V megger.
Failure Modes and Spec Pitfalls to Avoid
Dry block: insert depth must match the well — a probe inserted only into the upper third of the well sees stem conduction error and reads low. Well-to-well radial uniformity (e.g. 0.020 °C) is the real-world repeatability, not the display accuracy (e.g. ±0.4 °C) [S3]. Heat-up time of 25 minutes on the CS 650 [S1] means the operator should plan a duty cycle, not expect instant setpoint. Stability better than 0.1 °C is a 10-minute number on most datasheets; verify the test interval, not just the headline figure.
Decade box: residual resistance (zero-resistance offset of the closed switches) dominates below 10 Ω and is the usual reason a 0.1 Ω dialed value actually reads 0.3 Ω — short the output terminals with a copper link and record the residual before trusting the low decades [S4]. Switch contact thermal EMF (Cu-to-manganin junctions) adds a microvolt-level error that becomes visible on micro-ohm ranges and on RTD simulation when the working current is high. Per-decade accuracy is usually worst on the highest decade; a 0.1% spec on ×1 MΩ is 1 kΩ, which is fine for loop checks but wrong for laboratory Pt100 work — always cross-check against a precision multimeter at the working resistance. The accuracy of decade resistance boxes is typically expressed in percent of reading, with common commercial grades at ±0.1% to ±0.05% [S4].
Standards, Traceability, and Where the Two Converge
For temperature metrology, the relevant family of standards governs sensor characteristics rather than the calibrator itself: IEC 60751 defines the Pt100 resistance-versus-temperature curve (100.00 Ω at 0 °C, 138.51 Ω at 100 °C, 200.00 Ω at ~266 °C) and is the reference table a decade box simulates against when checking a temperature transmitter [S4]. Dry-block calibrators are typically specified with traceability to a national metrology institute via a reference standard platinum resistance thermometer (SPRT) and a fixed-point cell, and the user's SOP should record the uniformity and stability data shown on the OEM datasheet [S1][S3].
For RTD/thermistor work the bench setup converges: a dry block verifies the sensor at temperature, and a decade box verifies the transmitter or readout against a simulated resistance — they are sequential, not substitutable. The Pt100 RTD itself, with its 38.5 Ω/100 °C nominal slope and 0.385 Ω/°C sensitivity, is the bridge between the two instruments and the reason a complete temperature-calibration bench needs both. Two trackable signals to watch in the second half of 2026: portable dry blocks continue to push higher maximum temperatures (the CS 650's 650 °C ceiling is now common in mid-range portable units) [S1], and resistance-box datasheets are starting to publish residual-resistance and thermal-EMF figures that were previously hidden in the small print [S4]. A useful adjacent reference when sourcing bench instruments is the platform-trolley vs order-picker spec map approach to certificate traceability — the same documentation discipline that applies to calibrated instruments applies to the handling equipment that moves them between the calibration lab and the line.