How to read an MCH datasheet, line by line

Rated voltage

Rated voltage is the supply the power and resistance numbers are referenced to, and it is the first number to fix because power scales with its square. For a resistive element P = V² / R, so a ±10% band on the supply becomes roughly ±20% on power — a tolerance that has to appear somewhere in the thermal design, not just in the electrical notes. Datasheets in this family span 3.7–220 V with power from 5–300 W, and custom parts to 3,000 W; the point is not the range but that voltage and power are quoted together. Ask which voltage the cold resistance was measured at and which voltage the rated power is guaranteed at; if they differ, the sheet is quietly describing two different parts. Battery-powered designs should state the cut-off voltage, because that is the condition at which heat-up time is worst.

• Ask: the nominal voltage, the tolerance band, and the voltage at which rated power is guaranteed.
• Ask: for battery parts, the end-of-discharge cut-off voltage used for the heat-up figure.
• Trap: quoting power at nominal voltage only — a ±10% rail moves power by about ±20%.
• Trap: a 110 / 220 V dual-voltage line with no statement of how the two are wired.

Rated power

Rated power is not an independent input; it is the quotient of voltage squared and hot resistance, R = V² / P. A 24 V, 60 W part therefore needs 9.6 Ω at operating temperature, and a 230 V, 100 W part needs 529 Ω — two very different printed circuits for the same catalogue wattage. This is why a request that changes voltage on an existing drawing is never free: the resistance and the trace geometry have to change with it. Power is also the numerator of watt density, so it cannot be judged without the heated area. Ask whether the figure is continuous at a stated mounting condition or a peak at a stated duty cycle; a number quoted with neither is a marketing number. Ask also for the power tolerance at the tolerance extremes of the supply, since the two compound.

• Ask: whether rated power is continuous or peak, and at what mounting and duty cycle.
• Ask: hot resistance at rated power, so R = V² / P can be checked against the cold value.
• Trap: treating power as free to change — it forces a new resistance and a new printed pattern.
• Trap: comparing two parts by watts alone when the heated areas differ.

Cold resistance and its tolerance

Cold resistance is the only electrical parameter that can be verified at goods-in without a thermal rig, and it is the proxy for power: at fixed voltage, 10% low resistance means 10% high power. That is why the ±10% tolerance at 25 °C looks wide but is tight against the 3.6× swing between cold and hot for a tungsten element. Production parts typically show ratios of 2 to 3 at 600 °C because the printed resistor is tungsten paste with glass frit, not pure metal. The measurement defines the number: four-wire Kelvin, 25 ± 1 °C after two hours of stabilisation, meter zeroed against a shorting bar. Two-wire clips add 0.1–0.3 Ω, which can be 20% on a low-voltage part. Ask for the measurement method, not just the value.

• Ask: cold resistance with tolerance, and the test temperature and stabilisation time (25 ± 1 °C, 2 h).
• Ask: four-wire (Kelvin) method and the fixture, meter and date recorded on the certificate.
• Trap: two-wire clips adding 0.1–0.3 Ω — a 20% error on a low-ohm part.
• Trap: measuring straight off the line, before the part has stabilised at 25 °C.

The resistance–temperature curve

An MCH element is a near-linear resistor, which is why it suits a PID loop, but its resistance is far from constant. For tungsten R(T) = R₂₀ [1 + α(T − 20)] with α about 4.5 × 10⁻³/K; the R/R₂₀ table on a production sheet runs 1.00 at 20 °C, 2.26 at 300 °C, 3.16 at 500 °C, 4.06 at 700 °C and 4.96 at 900 °C. Printed tungsten paste with glass frit sits below the pure-metal calculation, so ratios of 2 to 3 at 600 °C are typical. The curve is what turns cold resistance into inrush current and what sets the steady-state current the protection device must carry. Ask for the curve of the actual part, not a generic material curve, and ask over what temperature range it was measured.

• Ask: the R–T curve for this part number, with the measurement range and method.
• Ask: the hot resistance at the working temperature, to size fuses and steady-state current.
• Trap: sizing a fuse on hot current when the first-cycle inrush can be 3× higher.
• Trap: quoting a pure-tungsten curve for a printed paste resistor — the ratio comes in lower.

Outline dimensions and tolerances

Outline dimensions are a fit requirement, not a description. Planar parts in this family run to L ≤ 100 mm and W ≤ 100 mm, discs to Ø3–60 mm and rods to Ø3–20 mm with L ≤ 90 mm; every one of those has to drop into a machined pocket or a housing, so the tolerance is as important as the nominal. Ceramic is fired, not machined to a final size, so dimensional tolerance is set by the green tape and the firing shrinkage, and it must be allowed for both the ceramic body and the lead exit. Ask for a 2D drawing with tolerances on every mating dimension, including flatness and the position of the termination area. A nominal size with no tolerance column is a common cause of a heater that will not seat.

• Ask: a 2D drawing with tolerances on every mating dimension, plus flatness and lead exit.
• Ask: how dimensional tolerance is held given firing shrinkage — green-tape control and lot data.
• Trap: specifying a nominal outline with no tolerance band for a part that fits a machined pocket.
• Trap: leaving the lead exit and termination position undefined; they are part of the envelope.

Thickness

Thickness is the thermal and mechanical lever hiding in a single number. Fired thickness in this family is typically 0.7–2.0 mm, with low-voltage parts from 0.55 mm, and conduction follows q″ = k · ΔT / t: with k = 20 W/m·K for 96% alumina, a 0.635 mm plate carries 3.15 W/cm² for every kelvin across it. A thicker plate spreads heat better and runs cooler at the surface at the same flux, but it has more thermal mass and responds more slowly; a thinner plate is faster but more fragile and less even. Ask what thickness the power density and heat-up figures were measured at, and ask for the thickness tolerance — a thin plate also sets the dielectric path length between the buried circuit and the surface.

• Ask: fired thickness with tolerance, and the thickness at which heat-up and watt density were measured.
• Ask: whether a thinner option changes the dielectric withstand or the mechanical margin.
• Trap: choosing minimum thickness for speed without re-checking flatness, fragility and dielectric path.
• Trap: comparing heat-up times quoted at different thicknesses.

Power density and the border that does not heat

Power density is P divided by heated area, q″ = P / A, and the area that matters is the printed resistor footprint, not the outline of the plate. Every part carries an unheated margin — the resistor-to-edge distance plus the termination area — and on a small chip that margin can be 25–35% of the outline, so calculating on the outline flatters the number by exactly that much. Worked example: a 40 × 20 mm chip at 24 V, 60 W with the resistor covering 32 × 16 mm has a heated area of 5.12 cm² and a watt density of 11.7 W/cm²; the same part rated 100 W would sit at 19.5 W/cm². 45 W/cm² is treated as the ceiling for continuous duty. Ask for the resistor footprint on the drawing and say which area the number refers to.

• Ask: the resistor footprint drawn on the 2D drawing and the area used for q″ = P / A.
• Ask: the watt density at the maximum supply tolerance, not only at nominal voltage.
• Trap: dividing by the outline — on a small chip the unheated border is 25–35% of it.
• Trap: treating 45 W/cm² as a target rather than a continuous-duty ceiling.

Maximum working temperature

Maximum working temperature is a materials limit, not a control setting. Co-fired MCH parts in this family are quoted at 700–1,000 °C surface, with rods up to 800 °C and 3D-printer hot ends around 500 °C continuous, while PTC elements are limited by their Curie point to roughly 250–300 °C. The tungsten–alumina interface and the glass frit in the paste set the upper bound: the higher the watt density, the hotter the buried element runs for the same surface temperature. Ask whether the stated maximum is a surface temperature or an element temperature, and under what mounting and atmosphere it applies. Ask also for the allowable surface temperature uniformity if the process needs it — a ±5 K requirement needs a graded resistor pattern or a spreader, not an average.

• Ask: whether the maximum is surface or element temperature, and the mounting it assumes.
• Ask: the predicted surface temperature profile, not a single average value.
• Trap: designing to the surface number while the buried resistor runs hotter at high watt density.
• Trap: assuming ±5 K uniformity comes free — it needs a graded pattern or a spreader.

Heat-up time

Heat-up time is a system result, not a part property. The familiar figure — 200 °C in about 10 s, 600 °C in 20–30 s, or ≤ 30 s to ignition temperature — is only meaningful with the supply, the thermal mass of the load and the mounting stated, because the heater is heating its own plate plus whatever it touches. Low thermal mass is why co-fired plates are fast, and the same mass is why adding a spreader or a clamp slows the response. Ask what was attached to the cold face during the measurement, at what voltage, and from which starting temperature; a number measured on a free part in still air will not survive contact with a metal block. Check the figure at the minimum supply voltage, which is the worst case.

• Ask: the load, mounting, starting temperature and supply voltage used for the heat-up figure.
• Ask: the heat-up time at minimum supply voltage, the worst case for a resistive element.
• Trap: quoting a free-air figure for a part that will be clamped to a metal block.
• Trap: comparing heat-up times without stating the thermal mass of the attached load.

Dielectric withstand and insulation resistance

These two fields are the safety case for a surface that is not live. Typical lines read 4,500 V for 1 s with no breakdown, insulation resistance > 100 MΩ, and leakage < 0.5 mA on assembled parts — the buried circuit is sealed inside dense alumina, so the surface can be touched without a live conductor. Ask for the test voltage, duration and the electrodes used, because a hi-pot result depends on where the voltage is applied; ask whether the test is 100% at end of line or sampled from a lot. Ask also that the requirement be confirmed at system level, not only at component level — the isolation that matters is between mains and the user, after the leads, connector and housing are added.

• Ask: test voltage, duration, electrode arrangement, and whether hi-pot is 100% or sampled.
• Ask: insulation resistance measured after the leads and any connector are attached.
• Trap: accepting a component-level hi-pot figure as proof of system-level isolation.
• Trap: a dielectric line with no duration — 4,500 V / 1 s is not the same as a continuous rating.

Leads and terminations

The lead is where a ceramic part becomes an assembly, and it is a mechanical specification as much as an electrical one. Nickel leads of Ø0.25 / 0.3 / 0.5 mm, Ni tabs, wire, FPC and connectors are all options, and the datasheet should fix gauge, length, colour, exit position and terminal type. Lead pull strength is quoted as ≥ 2 kgf axial and ≥ 3 kgf lateral on rod parts, which is the criterion to apply when the harness will be tugged during assembly. Ask for the metallisation and brazing or plating method at the junction, since a lead joint that oxidises or creeps is a field failure that no electrical test at goods-in will catch. Ask whether the lead is rated for the working temperature at the exit, not just at ambient.

• Ask: gauge, length, colour, exit position and terminal or connector, all on the drawing.
• Ask: lead pull strength and the metallisation / brazing method at the ceramic junction.
• Trap: specifying the electrical termination but leaving lead exit and strain relief undefined.
• Trap: rating the lead at ambient when the exit sits near the working temperature.

Certification and material declarations

The last field is paperwork, and it is the one that decides whether a part can be shipped into a given market. Declarations in this family cover RoHS / REACH and lead-free construction, with no Pb / Cd / Hg / Cr⁶⁺, plus material declarations, test reports and lot traceability with test records on request. Ask for a material declaration tied to the part number rather than a generic company statement, and ask what is kept on file per lot — a lot number plus a test record is what makes a later drift provable. Ask whether third-party test reports can be arranged at the buyer's request. Do not treat a compliance logo on the sheet as evidence; ask what the declaration covers, which part revision it applies to, and whether a test report accompanies the lot.

• Ask: a part-number-specific material declaration and the RoHS / REACH / lead-free status.
• Ask: what is retained per lot — lot number, dielectric and power test records, traceability.
• Trap: a generic company compliance statement with no link to the supplied part number.
• Trap: a logo taken as evidence, with no test report and no revision attached.

Specifications

Questions buyers ask