A decade resistance box is a switched precision resistor network offering selectable values in decade steps, with current production units spanning roughly 0.001 Ω to 111 MΩ and stated accuracies from 0.005% to 1% depending on grade [S1][S2][S3][S4].
Selection is governed by four binding criteria: total resistance range, smallest step (resolution), declared accuracy, and per-decade current/voltage limit, with secondary considerations including switch contact resistance, temperature coefficient, and whether the unit is suited for Pt100/RTD simulation or insulation resistance tester calibration [S1][S2][S4].
Definition, Topology, and How Dialed Resistance Is Built
A decade resistance box stacks N rotary decades in series, each decade adding weighted resistors in 1-2-3…10 steps; six-dial instruments therefore achieve one part-per-million resolution at the lowest decade when a 0.1 Ω or 1 Ω step is used as the LSB [S1][S2]. The Yokogawa 2793 family typifies this: 279301 covers 0.1 Ω to 1,111.210 Ω in 1 mΩ steps across 6 dials, while sibling 279303 covers 0 Ω to 111.1110 MΩ in 100 Ω steps for high-ohmage calibration of bridges and insulation testers [S1].
Switch contact resistance is non-trivial on low-ohm decades: Guildline specifies approximately 6 mΩ per switch on the RD Series, with the on-dial value at terminals being direct reading and not requiring the user to add a switch zero [S2]. IET Labs offers decades from 0.001 Ω to 100 MΩ with stability ≤7 ppm below 100 MΩ and ≤10 ppm above, plus optional IEEE/RS-232 control for programmable units [S4].
Accuracy Tiers, Stability, and What the Spec Sheet Really Means
Decade box accuracy is published per decade, not as a single number, because low-ohm decades are dominated by switch contact resistance while high-ohm decades are dominated by resistor temperature coefficient and leakage [S2][S4]. The IET/GenRad/esi product line states five accuracy tiers: 0.005%, 0.01%, 0.05%, 0.1%, and 1%, with stability of 7 ppm (≤100 MΩ) and 10 ppm (above 100 MΩ) over time [S4].
Guildline's RD Series publishes 0.1% adjustment accuracy at DC and 20 °C across all decades, with per-decade current ceilings of 1 A on the 0.01 Ω and 0.1 Ω decades, dropping to 7 mA on the 10 kΩ decade and 500 V maximum on the 100 kΩ–1 MΩ decades [S2]. A laboratory unit should be specified with both an initial accuracy and a stability figure, since the latter determines recalibration interval and ownership cost over 5–10 years.
Main Product Types: Manual, Programmable, and Specialty
The market splits into four practical categories, and the right pick depends on whether the box is used at a bench, in a calibration rig, on a production line, or in a classroom [S1][S3][S4][S6].
Manual bench units (IET 1433, Yokogawa 2793, Time Electronics 1061, Guildline RD Series) cover 0.1 Ω to 111 MΩ with 4–8 dials and 0.01%–0.1% accuracy, and dominate calibration of resistance thermometers and bridges [S1][S2][S4][S6]. Educational units such as the Arbor Scientific four-dial box deliver 1 Ω to 9999 Ω in 1 Ω steps at 1% tolerance with 4 mm safety sockets, intended for student DC-circuit experiments rather than metrology [S3]. Programmable decades (IET PRS-330, PRS-370, HPRS-F, OS-270 OhmSOURCE) add IEEE-488 or RS-232 control and are used in ATE, RTD simulation, and automated sensor calibration [S4][S6]. Specialty high-voltage or high-power units such as the HRRS-5-100M-5kV (5 kV insulation) and HPRS-F-7-0.001 are designed for insulation resistance tester verification and load testing, where standard 0.25 W resistors would fail [S4][S6].
Decision Criteria and Variant Comparison
For a buyer comparing 3–4 realistic options, the relevant decision criteria are resistance range, smallest step, accuracy, per-decade current, and intended use; the table below maps the four product families against these axes using only data published in current OEM literature. [S2]
Manual mid-range (e.g., Yokogawa 279301, IET 1433): 0.1 Ω to 1.111 MΩ, 1 mΩ step, 0.01%–0.05% accuracy, 0.25–1 W per decade, best for RTD PT100 simulation and bridge calibration. Manual high-range (Yokogawa 279303, Guildline RD11): 0 to 111 MΩ, 100 Ω step, 0.05%–0.1% accuracy, 500 V ceiling, best for insulation tester calibration. Programmable bench (IET PRS-330 series): 0.001 Ω to 100 MΩ, IEEE/RS-232, 0.005%–0.01% accuracy, designed for ATE and automated RTD simulation. Educational/field (Arbor 4-dial, Time Electronics 1061): 1 Ω to 9 999 Ω, 1 Ω step, 0.1%–1% accuracy, 4 mm sockets, low cost for teaching or quick field substitution [S1][S2][S3][S4][S6][S8].
Power Dissipation, Current Limits, and Common Failure Modes
Power dissipation is the single most common cause of decade box failure, and SparkFun's hookup guide states an overall ceiling of 0.25 W with 300 mA maximum current for its 999,990 Ω unit, with explicit guidance to verify current with an ammeter when dialling below 500 Ω [S7]. Guildline quantifies per-decade current explicitly: 1 A on 0.01 Ω and 0.1 Ω decades, 700 mA on 1 Ω, 200 mA on 10 Ω, 70 mA on 100 Ω, 20 mA on 1 kΩ, 7 mA on 10 kΩ, and 500 V maximum on 100 kΩ and 1 MΩ decades, with 50% short-term overload permitted [S2].
Other documented failure modes include switch contact resistance drift on low-ohm decades, which the user sees as a non-zero floor, and insulation leakage on high-ohm decades, which the user sees as a value lower than dialed in humid environments. For metrology work, request a traceable calibration certificate with measured data at the decade points actually used, not just a conformance statement; for related reference on resistance-based standards, the decade resistance box encyclopedia page lists accepted traceability practice. IET Labs explicitly markets Cal Lab, Field, and Transfer Standards grades, with the Transfer Standard tier intended for inter-laboratory comparison rather than routine work [S4][S6].
Who Should Not Buy a Mainstream 1% Educational Unit
An Arbor-class 1 Ω to 9999 Ω, 1% box is the wrong tool for any of the following jobs, even though it costs a fraction of a metrology unit: calibrating a Pt100 sensor to better than 0.1 °C, verifying an insulation resistance tester at 1 kV, sourcing a known 0.01 Ω in a current-shunt test, or replacing fixed resistors in a production ATE rack [S1][S2][S3][S4].
Likewise, a manual 6-dial 0.1 Ω to 1 MΩ unit is the wrong tool for high-voltage insulation tester calibration, where the 500 V insulation between decade and case becomes the limit; specialist units such as the HRRS-5-100M-5kV exist precisely because the standard switch and resistor insulation will flash over at the specified test voltage [S4][S6]. For metrology buyers, shortlist logic is: define the lowest and highest resistance actually used, the smallest step that matters, the worst-case current at the lowest decade, and whether the box must source RTD values; this collapses the catalogue to two or three candidates from the four families above.
Vendors, Replacement Cycles, and 2026 Sourcing Signals
Active manufacturers with published 2026 product lines are IET Labs (Westbury, NY; absorbs legacy GenRad and esi ranges), Yokogawa Test & Measurement (Japan; 2793 series), Guildline Instruments (Canada; RD Series), and Time Electronics (UK; 1061 series), with esi/Tegam DB62 still supported as a service base for installed units [S1][S2][S4][S6][S8]. The IET PRS-300 has been replaced by the PRS-330, and the HPRS-F-7-0.001 is the current high-power programmable platform, both of which are shortlist items for ATE buyers [S4][S6].
Trackable signals for the next procurement cycle: ISO/IEC 17025 recalibration interval drift on units older than 10 years, lead times for IET PRS-330 series versus legacy GenRad 1433, and pricing pressure from Chinese metrology-grade decade boxes entering the 0.05% tier; for context on how calibration instrument sourcing is moving in 2026, see this measuring instruments sourcing map and the related ISO GPS overhaul write-up.