MCH vs PTC: how to choose

The structural difference

A PTC element is a barium-titanate semiconductor whose resistance rises sharply as it approaches its Curie temperature, so the power it draws falls automatically as it heats. An MCH element takes the opposite route: a tungsten or Mo-Mn resistor is printed into alumina green tape and co-fired at about 1,650 °C in a hydrogen atmosphere, leaving the circuit sealed inside a dense 96% Al₂O₃ body. The result behaves as a near-linear metal resistor rather than a switching semiconductor. That one material decision propagates into every downstream specification: temperature ceiling, control architecture, inrush behaviour, long-term drift and the safety case. Before comparing datasheet numbers, it is worth fixing which physical mechanism you are buying, because many apparent performance differences are simply consequences of this split.

• PTC: barium-titanate semiconductor, resistance rises sharply at the Curie point and power self-limits
• MCH: tungsten / Mo-Mn printed into 96% alumina, co-fired at ~1,650 °C in hydrogen, circuit sealed inside the ceramic
• PTC: usable ceiling typically ≤ 250–300 °C; above it the self-limiting effect collapses output
• MCH: near-linear R–T curve, industrial range 700–1,000 °C, requires external control or protection

Temperature ceiling and heat-up: read both numbers together

The published figures are a useful starting pair. An MCH chip reaches 200 °C in about 10 s and continues to a 700–1,000 °C industrial surface range, whereas a PTC part is typically quoted at 60–120 s to 200 °C with a practical ceiling of 250–300 °C. The reason is not that one ceramic is a better conductor; it is that PTC is engineered to stop delivering power near its Curie point while MCH is engineered to keep delivering it. Use the pair as a screening test: if the process temperature must sit above 300 °C, the self-regulating mechanism has already left the design space. If the target is a constant low-temperature surface, the fast ramp and high ceiling of MCH buy you nothing and the PTC behaviour becomes the cheaper system choice. Estimate the ramp from your own thermal mass rather than assuming the catalogue value transfers unchanged.

• MCH: ~10 s to 200 °C, 20–30 s to 600 °C (site specification values)
• PTC: 60–120 s to 200 °C, typically ≤ 250–300 °C usable
• MCH industrial surface range: 700–1,000 °C, e.g. ignition duty
• Treat the quoted heat-up as a starting point; the installed thermal mass sets the real ramp

Specify MCH when

Choose MCH when the product needs fast heat-up, temperatures above 300 °C, high power density in a small envelope, stable power over the product lifetime, or lead-free RoHS / REACH compliance for EU and US markets. These conditions tend to arrive together: a compact tool that must reach temperature in seconds also needs a dense circuit and a control loop that can be tuned once and left alone. Because the tungsten circuit is buried in dense ceramic with no exposed wire to oxidise, the resistance stays stable across repeated thermal cycles, which is why a PID loop tuned at release does not need constant re-tuning. Typical fits include hair tools, instant water heating, soldering irons, pellet stove and industrial gas igniters, 3D printer hot ends and sensor light-off heating — in each case there is already a controller in the system.

• Fast heat-up (~10 s to 200 °C) or any target above 300 °C
• High power density in a small space, up to the 45 W/cm² ceiling
• Stable power over lifetime; no exposed resistance wire to oxidise
• Existing PID / thermostat control loop, and lead-free RoHS / REACH compliance

Specify PTC when

Choose PTC when inherent overheat protection matters more than peak temperature — low-temperature constant-temperature appliances, defrost and air heating, battery preheating, or designs where the control circuit can be removed entirely. Self-limiting behaviour is a safety feature rather than a limitation: as the element approaches its Curie temperature its resistance climbs and the drawn power falls, so the surface settles without a thermostat in the loop. The design trade is a narrow usable band, typically ≤ 250–300 °C, and a cold-start inrush that must be checked against the fuse and the supply. If RoHS compliance is required, confirm the lead content of the specific PTC formulation, since it is not automatically lead-free the way a co-fired alumina MCH element is.

• Inherent overheat protection is a safety feature, not a limitation
• Usable band typically ≤ 250–300 °C; power falls as resistance rises at the Curie point
• Check cold-start inrush against the fuse, relay and supply
• Confirm lead content explicitly if RoHS / REACH is required

Power density and what the alumina can actually conduct

Watt density is where the two technologies separate most clearly. For MCH, compute the power from P = V² / R, then divide by the heated resistor footprint — not the outer outline. 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 outline at 100 W sits at 19.5 W/cm². The ceramic is rarely the bottleneck. Heat crosses the plate by conduction, q″ = k · ΔT / t; with k = 20 W/m·K for 96% alumina and a fired thickness of 0.635 mm, each kelvin across the plate carries about 3.15 W/cm². At the 45 W/cm² ceiling the plate needs only about 14 K of gradient. The limits that arrive first are the interface to the load and the tungsten–alumina interface temperature, not the alumina.

• Compute P = V² / R, then watt density = P / A_heated; use the resistor footprint, not the outline
• 0.635 mm alumina conducts ~3.15 W/cm² per K via q″ = k · ΔT / t
• 45 W/cm² is treated as the upper limit; it implies only ~14 K across the plate
• The load interface usually dominates total thermal resistance, not the ceramic

Resistance over temperature, inrush and long-term drift

The two parts also differ in how resistance behaves once they are hot. A printed tungsten circuit follows R(T) = R₂₀ [1 + α(T − 20)] with α ≈ 4.5 × 10⁻³/K for tungsten, so the hot-to-cold ratio is large — the site datasheet example quotes 147 Ω cold against 529 Ω hot, and printed paste typically shows a 2–3× ratio at 600 °C rather than the full metal value. That temperature coefficient is exactly why cold-start current must be sized: at 230 V / 100 W the hot resistance is 529 Ω and the hot current 0.43 A, but a 3× cold factor gives 176 Ω and a 1.3 A first-cycle current against a 1 A fast fuse. A PTC element instead shows a steep resistance rise near its Curie point, so its power drifts with temperature by design and cold-start inrush is pronounced. MCH resistance stays stable over life because the circuit is sealed against oxygen.

• MCH: R(T) = R₂₀ [1 + α(T − 20)], tungsten α ≈ 4.5 × 10⁻³/K
• Printed paste shows ~2–3× hot/cold at 600 °C; size the fuse for the cold-start current
• PTC: steep resistance rise at the Curie point; power drift with temperature is intrinsic
• MCH: no drift over life, because the tungsten circuit is sealed inside dense ceramic

Safety, dielectric and the control loop

One caveat is easy to miss: MCH does not self-limit. Any MCH design therefore needs a thermostat, thermal fuse or closed-loop control, and the protection device is normally supplied as part of the heater assembly so the safety case is closed at component level rather than left to the appliance. Electrical isolation is structural, not a coating — the tungsten circuit is buried in dense alumina, giving a 4,500 V / 1 s dielectric withstand with leakage below 0.5 mA and insulation resistance above 100 MΩ. PTC offers self-regulation as an intrinsic overheat limit but still needs its cold-start inrush checked. Whichever technology is chosen, confirm the dielectric requirement at system level, specify the protection scheme on the drawing, and treat the control loop as part of the heater specification rather than a downstream firmware task.

• MCH is not self-limiting — a thermostat, thermal fuse or PID loop is mandatory
• Dielectric withstand 4,500 V / 1 s; leakage < 0.5 mA; insulation > 100 MΩ
• Isolation is structural (buried tungsten in dense alumina), not an added coating
• Confirm dielectric and protection at system level, not only at component level

A worked selection procedure

A repeatable sequence keeps the decision auditable. First, fix the required surface temperature: above 300 °C the choice is MCH; below it, compare the value of self-regulation. Second, compute the electrical starting point with R = V² / P at your nominal voltage and tolerance band (a ±10% voltage band moves power by roughly ±20%). Third, check watt density against the 45 W/cm² ceiling using the resistor footprint, and use q″ = k · ΔT / t to see whether the plate gradient is acceptable. Fourth, decide who closes the safety case — an existing PID loop and protection device favour MCH, a design with no controller favours PTC. Fifth, confirm RoHS / REACH and documentation. Recording these answers on the specification sheet makes the technology choice reviewable rather than a matter of preference.

• Step 1 — temperature: > 300 °C → MCH; below it weigh self-regulation value
• Step 2 — R = V² / P; a ±10% voltage band moves power ~±20%
• Step 3 — watt density ≤ 45 W/cm²; check the plate gradient with q″ = k · ΔT / t
• Step 4 — control and compliance: PID / protection present → MCH; none → PTC; confirm RoHS / REACH

Converting an existing PTC design

Replacing a PTC with an MCH in an existing product is usually feasible, but it is a control change rather than a drop-in substitution. Keep the mechanical envelope and the supply rail, then re-derive the cold resistance from R = V² / P at the nominal voltage, re-check watt density on the printed resistor footprint, and add the thermostat, thermal fuse or closed-loop control that the PTC had provided intrinsically. Re-check the dielectric at system level with the 4,500 V / 1 s withstand figure and re-size the fuse for the tungsten cold-start current. Send the voltage, available space and temperature target for a feasibility assessment; a like-for-like outline rarely preserves the R–T behaviour, so the control loop is the item to re-validate, not the ceramic body.

• Keep the outline and supply rail; re-derive R = V² / P for the new circuit
• Re-check watt density on the resistor footprint against the 45 W/cm² ceiling
• Add the thermostat / thermal fuse / PID protection the PTC supplied intrinsically
• Re-size the fuse for cold-start inrush; confirm 4,500 V / 1 s dielectric at system level

Specifications

Questions buyers ask