Cold resistance, hot resistance, and the inrush nobody plans for
A heater that measures 147 Ω cold and 529 Ω hot is not faulty. It is tungsten, and it will trip a fuse sized on the hot number.
Tungsten has a large temperature coefficient
Resistance follows R(T) = R₂₀ [1 + α(T − 20)]. For tungsten, α is about 4.5 × 10⁻³/K — one of the largest of any common conductor. At 600 °C the bracket works out to 3.6, so a heater whose hot resistance is 529 Ω measures around 147 Ω cold. Nothing is wrong with it. That ratio is the physics of the element, and it is the reason a 100 W heater can open a 1 A fuse.
The printed resistor is not pure tungsten
An MCH circuit is a printed paste — tungsten with glass frit, often with Mo-Mn — so the measured ratio comes in below the pure-metal calculation. Ratios of 2 to 3 at 600 °C are typical for production parts. The number you must not guess is yours: measure the cold resistance, measure the current at operating temperature, and put both on the datasheet.
What the inrush actually breaks
The surge lasts a second or two, which is long enough to do damage and short enough to be ignored during qualification.
- Fuses — a 1 A fast fuse opens around 2 A in under a second; a 3.6× surge welds it
- Thermostat contacts — arc energy at make scales with inrush, and contacts pit
- Relays and triacs — rated for steady resistive load, not for four times cold current
- Traces and connectors — sized on steady state, hot at every switch-on
The fix is rarely a bigger fuse
Size the protection against the inrush waveform, not just the steady-state number: read the time-current curve, confirm the fuse rides through the cold start, and confirm it still opens on a real fault. Where the surge is unavoidable, add zero-cross switching, a soft-start element, or a relay with a higher make rating. A fuse that survives the first second and still protects the steady state is the design target — those two requirements pull in opposite directions, which is why this gets skipped.
Why the datasheet says ±10% cold resistance
Cold resistance is the only electrical parameter you can verify at goods-in without a thermal rig, and it is the proxy for power: at fixed voltage, 10% low resistance is 10% high power. The tolerance looks wide until you compare it with the 3.6× change between cold and hot. Ten percent is tight in that context.
Measuring it so the number means something
Most disagreements about resistance come from the measurement, not the part.
- Four-wire (Kelvin) measurement — two-wire clips add 0.1–0.3 Ω, which is 20% on a low-voltage part
- 25 ± 1 °C, after two hours of stabilisation — not straight off the line
- Zero the meter against a shorting bar, not against the probes in air
- Record meter, fixture and date on the certificate, so a later drift is provable
A worked number
230 V, 100 W. Hot resistance 529 Ω, hot current 0.43 A. At a 3× ratio the cold resistance is 176 Ω and the first-cycle current is 1.3 A. A 1 A fast-acting fuse typically melts near 2 A and will hold 1.3 A — but only just, and only at 20 °C. Move the same product to a 0 °C warehouse and the resistance drops another 10%, pushing the surge to about 1.4 A. Put two heaters on one circuit and the fuse opens on a cold morning and never on the bench. That is how intermittent field failures are born.
Fuse selection is an I²t problem
Fuses are sold on current but they clear on energy. The melting integral I²t describes how much energy the element absorbs before the wire opens. A 3× surge lasting 1.5 s carries about 9× the I²t of steady state, so a fuse chosen on the current rating alone is consumed a little at every switch-on. Datasheets publish a pulse derating curve; the working rule is to keep the surge I²t below about half the melting integral. At twenty cycles a day, that derating is not optional.
Soft start, three ways
In rough order of cost:
- Zero-cross switching — removes first-cycle asymmetry and the worst di/dt, cheap with a triac
- Series NTC — simple, but it has to cool between cycles or the second start gets no help
- Current-limited supply or a PWM ramp — the cleanest answer, and the one that needs a controller
Why the first cycle is the worst
The element is coldest at the first switch-on after storage, the contacts are clean, and the fuse is at ambient. Every later cycle starts from a warmer element with higher resistance. A design that survives its first cold start usually survives the rest — which is also why a single bench test at room temperature proves very little.
When the ratio drifts
A cold resistance that creeps upward over the life of a part is a circuit degrading — a trace crack, an oxidised lead joint, a hot spot. If your field returns show rising cold resistance before failure, the heater was running hotter than the design assumed, and the root cause is usually the mounting rather than the element.
Specifications
Questions buyers ask
No. It is normal behaviour for a tungsten-based element. A part that measured the same hot and cold would be a different material, not a better heater.
Yes, in low-voltage and small-thermal-mass designs. Size it for the steady-state current after it heats, and check it does not steal voltage from the element at start-up.
The ratio is the same, but a 12 V design carries about 19× the current of a 230 V design at the same power, so the absolute surge is far larger and every conductor has to be sized for it.
Because ambient sets the cold resistance. Ten degrees colder is roughly another 4.5% lower resistance and 4.5% more surge, which is often exactly the margin the fuse did not have.
Only with one that captures the first 100 ms. An averaging clamp meter shows the steady current and tells you nothing about the surge.
Tell us the supply and the protection
Voltage, power, fuse or relay part number and the ramp you need — we will check the cold-start current against it.