Duty temperature decides the construction before anything else does. At 80 °c to 120 °c, alumina ceramic thick film heaters is a candidate; the sections below set out what it will hold, what has to be true for it to hold it, and what fails first when it does not.
What the 80 °C to 120 °C Band Requires
Battery operating windows, incubators, fluid warming and display heating. The upper limit of PET film, and the band where a low-cost printed heater stops being a candidate.
| Band | 80 °C to 120 °C |
|---|---|
| This construction, continuous | 400 °C |
| Limited by | fired glass dielectric on the substrate |
The band sits comfortably inside the 400 °C ceiling, which is where a heater should be specified: the margin between the working temperature and the insulation limit is what absorbs a poor thermal path, a warm ambient and a control overshoot without shortening the life of the part.
Why the Boundary Is Where It Is
120 °C is where PET film stops. The polyester softens from about 130 °C and loses its dielectric strength as it softens, so the margin between working temperature and failure is thin — and a failure here is an electrical one, not a cosmetic one.
This is therefore the last band in which the cheapest thin heater is a candidate. Above it, the construction changes to silicone or polyimide, and the cost per square metre changes with it.
What an Alumina Ceramic Thick Film Heaters Is
The same printed construction on 96% alumina, for higher temperature and for applications that need the element electrically isolated from everything around it.
Dielectric and resistor pastes screen printed onto stainless steel or alumina and fired at around 850 °C, then laser trimmed to resistance. There is no separate element and no interface: the heated surface is the substrate itself. On alumina the dielectric is the substrate, so the isolation is excellent and the temperature capability is the highest here — at the cost of a brittle part that must be clamped evenly and never asked to take a bending load.
The heat leaves through a clamped joint, and a clamped joint is only as good as the pressure and the flatness behind it. A thermal interface material or a thin film of thermal grease fills the gaps that machining leaves; without one, contact happens at the high spots and the element runs hotter than the plate it is bolted to.
| Element | fired resistor paste, laser trimmed |
|---|---|
| Insulation system | fired glass dielectric on the substrate |
| Total thickness | 1 mm to 3 mm total |
| Maximum size, one piece | 150 x 200 mm |
| Continuous working temperature | 400 °C |
The continuous limit of 400 °C is set by fired glass dielectric on the substrate. The element alloy would run far hotter; it is never the element that decides.
Which Temperature Is Being Held
Three temperatures matter and they are not the same number: the element temperature, the temperature of the surface being heated, and the ambient. A heater rated for a continuous working temperature is rated at the element — the hottest point in the assembly, and the one nothing on the outside is measuring.
The gap between element and surface is the thermal path. Bonded to metal it may be a few degrees; across an air gap, a void in the adhesive or a dry clamped joint it can be tens of degrees, and every one of them appears at the element rather than at the load.
Which is why sensor position is a drawing item. A sensor laminated into the heater protects the heater; a sensor on the load controls the load. Assemblies that have to do both carry two: a control sensor where the process is, and a limit device against the element.
Watt Density: the Two Figures
Two watt densities are quoted because there are two thermal situations. With nothing conducting heat away, 10 W/cm² is what the insulation survives. Bonded to a heat sink over the full outline, 80 W/cm² is. Anything in between depends on the quality of the joint, and the joint is the customer's to make.
The failure is silent and quick: the element does not know how it is mounted, so at 80 W/cm² in free air the heat simply accumulates in the insulation until it degrades. This is the single most common way these heaters are destroyed, and it happens on the first run rather than after a year.
Thick film reaches these densities because the heated surface is the substrate — there is no interface between element and load to conduct across. That is also why dry-run protection is not optional: with the fluid gone, so is the entire thermal path, and the fired layers reach their limit in seconds rather than minutes.
Control and Protection
This heater has no way of limiting its own temperature. Resistance heating draws roughly the same power whatever the temperature, so with the load removed, the fluid gone, the fan stopped or the sensor open circuit, the element keeps producing watts until something stops it.
The design must therefore contain a control device and, separately, a protective one: a controller with a sensor to hold the set point, and a bimetal thermostat or a one-shot thermal cutoff wired in series with the element to open the circuit if the loop ever fails. The two are not interchangeable — a controller that has lost its sensor cannot protect anything, which is exactly the fault the cutoff exists for.
On a fluid heater the limit device belongs on the substrate, close to the resistor, not downstream in the water. Downstream is where the temperature has already gone wrong; the substrate is where it goes wrong first, and at these watt densities the interval between the two is short.
Sensors can be built into the part: NTC thermistor, PT100 or PT1000 RTD, or a K-type thermocouple, fitted under the same insulation and tested with the heater as one assembly. A sensor laminated into the heater reads the heater; a sensor on the load reads the load. Both are legitimate, and the drawing has to say which is wanted.
How This Construction Fails
- Run at the bonded watt density in free air. The insulation reaches its limit within minutes and the element burns through where the local density is highest. Specify at 80 W/cm² only where the heater is genuinely bonded to metal over its whole area; unbonded, the figure is 10 W/cm².
- Loss of the control loop. An open-circuit sensor, a welded relay contact or a controller in manual leaves the element at full power with nothing watching it — the case the series thermal cutoff exists for, and the reason it must not be omitted to save a part.
- Dry running. With the fluid gone the entire thermal path is gone, and at these densities the fired layers and the trim are destroyed in seconds. Flow or level interlocks, not a downstream thermostat, are what prevent it.
- Scale on the wetted face. A layer of scale is a thermal resistance in exactly the wrong place: the substrate temperature rises to push the same power through it, and the failure looks like an overpowered heater rather than a fouled one.
Where This Construction Is the Wrong Answer
- Do not plan on running above 400 °C. That ceiling belongs to the insulation system, not to the element alloy, and it is not a figure that derating extends — past it the insulation degrades and the dielectric strength goes with it.
- Do not flex the substrate, clamp it unevenly, or accept a dented one. The fired glass dielectric is a brittle layer bonded to metal, and it cracks before anything visible happens to the plate.
How to Specify One
This is a made-to-drawing part, so there is no catalogue number to quote. A quotation is built from the following, and the more of it that arrives with the enquiry the fewer rounds it takes:
- An outline drawing, or a sketch with dimensions: overall size, holes, cut-outs, radii and the lead exit position.
- The supply voltage, and whether it is AC or DC. The element is wound for one voltage; it is not a switchable range.
- The power required, or — better — the temperature to be reached and the mass or flow it has to be reached in. Given the second, the wattage can be calculated instead of guessed.
- What the heater is attached to, and how. This is what decides which watt density figure applies, and it is the question a quotation cannot proceed without.
- The ambient, and the lowest temperature the assembly must start from. A heater sized for a 20 °C start is undersized for a -30 °C one.
- Duty cycle: continuous, intermittent, or a warm-up to a deadline. Time to temperature is a specification, not a by-product.
- Sensor and protection: which sensor, where it goes, and which limit device is fitted in series with the element.
- The certification the end product is built to, named on the enquiry rather than after approval.
- Quantity and target price, because the construction that is right for fifty pieces is often not the one that is right for fifty thousand.
- The fluid, its flow rate and its inlet temperature, and how a dry-run condition is detected in your system.
Lead time: Samples 7-10 days from approved drawing; production 15-20 days.
Alumina Ceramic Thick Film Heaters — Capability Table
These are process limits for the construction, not the specification of one stocked item. Every line is what the factory can hold; what your part holds is fixed on your drawing.
| Parameter | Specification | Limit / note |
|---|---|---|
| Construction | Glass dielectric and resistor pastes screen printed and fired onto the substrate, then laser trimmed to resistance | Fired, bonded or encapsulated as a single assembly |
| Substrate | 96% alumina ceramic, 0.6 mm to 2 mm | The substrate carries the heat and the mechanical load |
| Thickness | 1 mm to 3 mm total | Substrate plus element system |
| Maximum size, one piece | 150 x 200 mm | Limited by the firing or curing fixture |
| Dimensional tolerance | u00b10.2 mm on the substrate, u00b10.1 mm on printed features | On the machined substrate |
| Supply voltage | 12 V DC to 400 V AC | Circuit is designed for the ordered voltage |
| Resistance tolerance | u00b110% standard, u00b15% selected | Thick film resistors are laser trimmed where a tighter band is required |
| Watt density | 10 W/cmu00b2 typical, 80 W/cmu00b2 maximum in a wetted application | Achievable because the heat is generated in the surface being cooled |
| Maximum working temperature | +400 u00b0C at the element system | Of the element system; the substrate itself tolerates more |
| Dielectric strength | 2500 V AC, 1 minute | One minute, circuit to substrate |
| Insulation resistance | greater than 100 Mu03a9 at 500 V DC | At 500 V DC |
| Terminations | Welded tabs, crimp terminals or soldered leads depending on temperature | Soldered, welded or crimped as the temperature allows |
| Mounting | Bolted, brazed or welded into the assembly; flatness of the mating face is critical | Flatness of the mating face matters more than clamping force |
| Sensor and control options | Printed PTC or NTC track, or a bonded thermocouple | Integrated into the same firing or bonded to the substrate |
| Certifications available | UL 499, IEC/EN 60335, RoHS, REACH, CE | Construction level |
| Isolation | Substrate is the insulator u2014 no separate dielectric needed | Which is what allows the highest watt densities here |
| Handling | Brittle; support fully and never clamp on a point | Nearly every ceramic heater failure in the field is mechanical, not electrical |
Capability figures are this factory's process limits for fired thick film construction. Watt densities far above any flexible heater are achievable because the resistor is printed directly onto the surface being cooled — there is no interface and no adhesive in the heat path.
