Dry Block Temperature Calibrator

What a Dry Block Calibrator Is

A dry block temperature calibrator is a temperature-controlled solid metal block used as a transfer standard for comparison calibration of temperature sensors. The principle is deliberately simple: heat a metal block, and on cooled models also cool it, then hold its temperature stable while a sensor inserted into a drilled well comes to the same temperature. The technician sets a target value, waits for the block and the sensor to equilibrate, then compares the sensor reading against the calibrator display or against a separate reference probe in an adjacent well. Because the heat-transfer medium is solid metal rather than a stirred fluid, the instrument is called dry, distinguishing it from the liquid baths and micro-baths that share the same comparison method.

Functionally a dry block sits between two neighbors. Below it is the simple liquid bath or ice point cell used as a fixed reference. Above it is the primary fixed-point apparatus (triple point of water, melting points of metals) used by national laboratories to realize ITS-90. The dry block is the workhorse middle tier: more accurate and faster than a portable bath, far cheaper and more transportable than a primary standard. In a plant, it is what a maintenance team carries to the field to verify that a temperature transmitter loop reads correctly, and what a calibration laboratory uses to issue routine certificates for process RTDs and thermocouples.

Structurally a dry block has four functional parts. First, the metal block itself, machined from a high-conductivity alloy so heat spreads evenly across the wells. Second, the heating system, normally resistive cartridge heaters, and on cooled units a thermoelectric (Peltier) stack that can both add and remove heat. Third, the interchangeable insert or sleeve, a removable metal cylinder bored with holes sized to the sensor sheath diameter so that the air gap around the sensor is minimized. Fourth, the internal control and reference sensor, typically a precision Pt100 RTD that both regulates the block and provides the displayed temperature. Higher-grade units add a separate channel for an external reference probe.

The reason a dry block exists at all is that comparing a working sensor against a known temperature is the only practical way to assign and verify its error. A bare thermometer cannot certify itself. The dry block provides a uniform, repeatable, traceable temperature environment that a reference probe can measure with low uncertainty, so the difference between the reference and the device under test becomes the calibration result. Everything else in the design, from block geometry to multi-zone control, exists to make that environment as uniform and stable as possible.

Four engineering metrics determine whether a given dry block is fit for a task: stability over time, axial and radial uniformity, reference accuracy, and temperature range including any below-ambient capability. These four, read together against the sensors to be calibrated, decide both the achievable measurement uncertainty and the practical fit of the instrument. A calibrator with an impressive display accuracy but poor axial uniformity will still produce large errors on short sensors, which is why a single headline number on a brochure is never enough.

Types and Classification

Dry block calibrators are classified along two axes that buyers most often confuse: by accuracy grade (industrial, reference, metrology) and by temperature capability (heat-only versus actively cooled). A third axis, the comparison-calibration peer group, places the dry block alongside the micro-bath and the surface or infrared calibrator that share the same comparison principle but suit different sensor shapes. The table below frames these families so the rest of the chapter can address each in turn.

Industrial dry blocks are the portable, rugged field tools that maintenance teams carry between sites. They emphasize fast heating, robustness, and a single permanently fitted or simply swapped insert. Display accuracy is usually in the 0.2 to 0.5 degrees Celsius band, which is sufficient to verify a process loop whose transmitter accuracy is 0.5 to 1 percent of span. The AMETEK JOFRA ATC series and WIKA CTD9100 industrial models occupy this tier, alongside compact Fluke field units.

Reference and metrology wells are laboratory-grade instruments built for the lowest practical uncertainty short of a fluid bath. They use interchangeable inserts, multiple independently controlled heating zones to flatten the axial profile, and a separate channel for an external reference probe. The Fluke metrology well family (models 9170 to 9173) spans -45 to 700 degrees Celsius with stability as tight as plus-or-minus 0.005 degrees Celsius on the cold units, and the AMETEK JOFRA RTC reference series reaches plus-or-minus 0.11 degrees Celsius with an external STS reference. Fluke notes that better stability than these wells can be found only in fluid baths and primary fixed-point devices.

Micro-baths are small stirred liquid baths included here because procurement teams cross-shop them against dry blocks. By immersing the sensor in moving fluid rather than a drilled hole, the micro-bath sidesteps the fit and immersion problems of a dry well and suits short, square, or irregular sensors. A representative Fluke micro-bath holds stability to plus-or-minus 0.03 degrees Celsius or better with uniformity of plus-or-minus 0.02 degrees Celsius. The trade-off is fluid handling, spill risk, a narrower range, and slower temperature changes.

Infrared and blackbody calibrators present a temperature-controlled target of known emissivity for non-contact infrared thermometers, which cannot be inserted into a well at all. They are a distinct product but belong to the same calibration-room toolkit, so a buyer scoping a temperature laboratory should plan for them separately. The remainder of this guide concentrates on the contact dry block and its reference-well variants.

Heating, Cooling, and Uniformity

The performance of a dry block is governed by how it adds and removes heat and how evenly that heat reaches every point of the well that the sensor occupies. Three engineering choices shape this: the heating and cooling technology, the number of control zones, and the resulting axial and radial uniformity. The table below contrasts representative real units across the temperature span so the numbers are concrete rather than abstract.

Heating on every dry block uses resistive elements that drive cartridge heaters embedded in or around the block. A cooling fan and the block's own surface area shed heat to bring the temperature back down. Because metal conducts and the heaters can be powerful, dry blocks reach setpoint fast: the Fluke 9143 heats from 33 to 350 degrees Celsius in about 5 minutes, though cooling the same span back down takes roughly 32 minutes since it relies on passive heat loss. This heat-only architecture is why the lower limit of an uncooled block sits at or just above ambient, commonly 33 degrees Celsius.

Active cooling below ambient is achieved with thermoelectric Peltier modules, which pump heat across a semiconductor junction in either direction with no refrigerant or compressor. Peltier-cooled units such as the Fluke 9170 (-45 degrees Celsius) and the WIKA CTD9100-COOL (-55 degrees Celsius) bring the same uniformity discipline to sub-ambient calibration of cold-chain, pharmaceutical, and laboratory sensors. Below roughly -50 degrees Celsius, frost and condensation on the open well become the limiting nuisance, handled with a dry gas purge or sealed insert. For points colder than about -90 degrees Celsius a stirred low-temperature bath is normally required instead.

Multi-zone control is the defining feature that separates a reference well from a basic block. The AMETEK JOFRA RTC-700 uses three independently controlled heating zones: the lower and middle zones manage the main heat input while the upper zone compensates for heat lost up the sensor stem, flattening the vertical temperature profile so the homogeneous zone is as long and as level as possible. AMETEK adds Dynamic Load Compensation (DLC), which measures the insert temperature and feeds it back to the controller to correct for the cooling effect of loading sensors into the wells.

Axial uniformity (axial homogeneity) is the temperature variation along the depth of one well, specified over a homogeneous zone at the well bottom, conventionally 40 mm and on better units 60 mm. It is the single most important uniformity number because a short sensor that does not reach the bottom of this zone reads a different temperature than a long reference probe. The Fluke 9143 shows this clearly: axial uniformity tightens to plus-or-minus 0.04 degrees Celsius at 33 degrees but widens to plus-or-minus 0.2 degrees at 350 degrees, since larger temperature differences with the surroundings drive larger gradients.

Radial uniformity (well-to-well or hole-to-hole uniformity) is the temperature difference between separate bores in the same insert. It matters whenever the reference probe occupies one well and the device under test occupies another, because any difference between those bores is a direct calibration error. Typical hole-to-hole uniformity is around plus-or-minus 0.05 degrees Celsius on field units and far tighter on reference wells: the Fluke 9143 holds radial uniformity to plus-or-minus 0.01 to 0.02 degrees Celsius, and the cold metrology wells reach plus-or-minus 0.01 degrees Celsius.

Reference Configuration and Standards

How the calibrator measures its own temperature, and which document governs the uncertainty budget, determine whether a result is defensible. The choice between an internal and an external reference is the most consequential decision a user makes after picking the range, and it is the point where reference accuracy and display accuracy diverge.

The internal reference is the control sensor built into the block, usually a Pt100 RTD positioned in the metal body rather than in the removable insert. It regulates the block and drives the display. Because it sits in the block, it reaches a stable temperature faster than a sensor placed in the insert, but its reading does not perfectly represent the well that the device under test occupies, especially with short sensors or rapid temperature changes. The accuracy quoted against this sensor is the display accuracy, for example plus-or-minus 0.2 degrees Celsius on the Fluke 9143 or plus-or-minus 0.29 degrees Celsius on the JOFRA RTC-700 internal sensor.

The external reference is a separate precision thermometer (an STS or PRT probe) inserted into a well alongside the device under test and read through a dedicated reference channel on the calibrator. Because it sits in the same insert and at the same depth as the unit being calibrated, it measures essentially the same temperature, eliminating the block-to-insert offset that limits the internal sensor. This is why the RTC-700 reaches plus-or-minus 0.11 degrees Celsius with an external STS reference versus plus-or-minus 0.29 degrees on the internal sensor. The trade is added cost, an extra probe to maintain and calibrate, and slightly longer dwell.

The governing procedural document is EURAMET Calibration Guide No. 13 (EURAMET cg-13, historically EA-10/13), which defines the uncertainty budget for calibrating temperature sensors in a dry block. It enumerates the components every user must account for: axial homogeneity, the temperature difference between borings (radial), the effect of loading, stability over time, hysteresis, and the reference sensor uncertainty. EURAMET stresses that the procedure, particularly insufficient stabilization time, is frequently a larger error source than the instrument specifications themselves.

Several standards frame the surrounding measurement. The temperature scale realized by the reference probe is the International Temperature Scale of 1990 (ITS-90). The sensors being calibrated conform to IEC 60751 for Pt100 platinum resistance thermometers and IEC 60584 for thermocouples, which fix their resistance and voltage versus temperature relationships. Calibration laboratories that issue accredited certificates operate under ISO/IEC 17025. None of these prescribe one accuracy figure, so the datasheet must always be read against the actual sensors to be checked. The table below maps each uncertainty component to its source and how to manage it.

Key Specification Parameters

Reading a dry block datasheet is a discipline of its own because vendors mix display accuracy, reference accuracy, stability, and several uniformity figures that are easy to conflate. Eight parameters truly drive selection: temperature range, display accuracy, reference accuracy, stability, axial uniformity, radial uniformity, loading and stabilization behavior, and well geometry. Each is explained below.

Temperature range is the span of setpoints the unit can reach and control. Note where the lower limit falls: heat-only blocks bottom out near ambient (33 degrees Celsius on the Fluke 9143 and JOFRA RTC-700), while Peltier units go sub-ambient (-45 degrees on the Fluke 9170, -55 degrees on the WIKA CTD9100-COOL). Always confirm the unit covers every setpoint in your calibration procedure, including the lowest and highest, with margin, since accuracy and uniformity both degrade toward the extremes of the range.

Display accuracy and reference accuracy are two different numbers and must never be compared across the line. Display accuracy uses the internal control sensor (for example plus-or-minus 0.2 degrees Celsius on the 9143); reference accuracy uses an external STS or PRT probe (for example plus-or-minus 0.11 degrees Celsius on the RTC-700). A buyer who quotes a reference figure for one brand against a display figure for another will reach a false conclusion. Decide which mode you will actually use, then compare that figure only.

Stability is how tightly the block holds a setpoint over a defined window once equilibrated, expressed in degrees Celsius. Reference wells reach remarkable values, plus-or-minus 0.005 degrees Celsius on the cold Fluke 9170 and 9171 and plus-or-minus 0.008 degrees Celsius on the JOFRA RTC-700 near ambient, loosening to about plus-or-minus 0.03 degrees Celsius at 700 degrees. Stability sets the floor on what the comparison can resolve and is the figure you watch on the display before recording.

Axial and radial uniformity were defined in Chapter 3 and are repeated on every serious datasheet at several setpoints. Read them at your actual working temperatures, not just at the most flattering point. The 9143 family illustrates the spread: axial uniformity moves from plus-or-minus 0.04 degrees Celsius at 33 degrees to plus-or-minus 0.2 degrees at 350 degrees, while radial stays within plus-or-minus 0.01 to 0.02 degrees. A unit excellent near ambient can be ordinary at high temperature.

Loading and stabilization behavior describe how the block reacts to real sensors and how long it must dwell. Loading effect is the cooling caused by heat conducted out through the inserted sensor stems; load-compensated designs (AMETEK DLC, Fluke and WIKA gradient control) correct for it. Stabilization is the dwell after the display reaches setpoint, and reading too early is, per EURAMET, often the single largest error source. The relevant practical specs are heating and cooling time and the time to reach a given stability band.

Well geometry covers the insert outer diameter, the well depth, and the bore sizes that fit your sensor sheaths. Depth matters because immersion drives uncertainty: the 9170 well is 160 mm deep, while the 9171 to 9173 wells are 203 mm to allow deep immersion of long probes. The Fluke 9143 uses a 25.3 mm (1.00 in) insert with 150 mm (5.9 in) immersion. Specify inserts bored close to the sensor diameter so the surrounding air gap, and the error it causes, stays small.

Selection Decision Factors

To turn the preceding chapters into a specific model, follow the decision sequence below. Most selection mistakes are not single wrong steps but premature decisions at the wrong level, for example fixing on a brand before the required range and accuracy are settled. These eight steps double as a fixed RFQ template.

1. Temperature range and below-ambient need: List every setpoint in your procedures, find the coldest and hottest, and decide whether you need a Peltier-cooled unit (sub-ambient, down to -45 or -55 degrees Celsius) or a heat-only block (ambient and up). Add margin at both ends.
2. Target accuracy versus the test sensors: The calibrator should be at least two to four times more accurate than the sensors it checks (the test accuracy ratio). Define which mode you will use, display or external reference, and compare only that figure across brands.
3. Reference channel: Decide whether the internal sensor is sufficient (process loops) or an external STS/PRT reference channel is required (laboratory certificates, lowest uncertainty). If external, budget for the reference probe and its own calibration. Note that the dry block checks the sensor end of a loop, while the electrical signal end is handled separately by a loop calibrator.
4. Uniformity at working temperatures: Check axial uniformity over the 40 or 60 mm zone and radial uniformity at your real setpoints, not just near ambient. Short or numerous sensors raise the importance of these numbers.
5. Insert and well geometry: Match well depth and insert bore size to your sensor sheath diameters and lengths so immersion is adequate and the air gap is minimal. Confirm custom-bored inserts are available for unusual probes.
6. Loading and speed: Estimate how many and how thick the sensors are per batch, then favor load-compensated multi-zone designs for heavy loads. Weigh heating and cooling times against throughput, remembering that passive cooling is slow.
7. Portability, power, and protection: Field work needs light weight, robust housing, and a suitable supply voltage; bench work tolerates larger, more capable units. Confirm mains compatibility for the destination country.
8. Traceability and total cost of ownership: Require a calibration certificate traceable to ITS-90 through a national institute, then add recalibration intervals, reference probe upkeep, inserts, and downtime to the purchase price. A cheaper unit that needs frequent recalibration or fails a TAR check costs more over a five-year horizon.

One last commonly overlooked dimension is serviceability and traceability support: availability of accredited recalibration (ISO/IEC 17025) for both the block and its reference probe, lead time for custom inserts, firmware and documentation continuity, and local service presence. Fluke Calibration (formerly Hart Scientific), AMETEK JOFRA, WIKA, Beamex, and Isotech all maintain calibration laboratories and spare-parts support across major markets, which is what keeps a calibrator defensible and usable years after purchase.