PWM Heater Controllers: Check Timer Resolution at the Required Low Power

System boundary

Temperature controller, timer, switching stage, source and resistive heater. Low-power reachability depends on both digital timing and actual pulse delivery.

Integration interfaces

Integration risks

Follow the command from temperature error to heater terminals

Identify the controller output variable, timer conversion, gate driver, switching device and heater connection. A floating-point duty value can contain many decimal places while the timer accepts only integer counts. The driver can then impose another limit before that digital pulse becomes useful voltage across the heater. Record each conversion rather than assigning the software display directly to physical power.

This discussion concerns a resistive heater under a voltage-switched PWM drive. It assumes the on-state voltage and resistance are sufficiently constant within one short period for an initial power estimate. Current regulation, source shutdown, appreciable reactive behavior or a changing on-state voltage need their own model. The timer resolution remains relevant, but it no longer determines power through the simple equation alone.

Calculate the available duty increments

Let the effective timer count frequency be ft and the PWM period contain N equal count intervals. The period is N/ft and frequency is ft/N. An on-time of n intervals gives duty D = n/N under the stated edge-aligned counting convention. Here ft is in hertz, N and n are counts and D is dimensionless.

Actual peripherals have specific prescalers, compare behavior, update rules and endpoint handling. Obtain those from the selected hardware documentation. A timer's register width does not mean that every bit is used at the chosen frequency: only the count intervals in the configured period determine this duty spacing. Center-aligned modes or special high-resolution features require their actual timing convention, not an assumed reuse of the equation.

Convert one timer count into a nominal heater power step

For a constant resistance R and on-state heater voltage V, the ideal on-state power is V squared divided by R. Average power in this simplified model is D times that value, and one timer count changes average power by V squared divided by R N. V is in volts, R in ohms and power in watts; off-state power is assumed zero.

Assume an effective count frequency of 48 MHz, a 20 kHz PWM rate, a 24 V on-state heater voltage and 12 ohm resistance. The period contains 2,400 counts and lasts 50 microseconds. On-state power is 48 W, so the nominal average-power increment is 0.020 W per count. These assumed values demonstrate the calculation, not a selected timer, heater rating or validated operating point.

Check the smallest pulse that reaches the heater

Now assume the complete driver requires an on-command of at least 2 microseconds to deliver a repeatable usable pulse. At 48 MHz that corresponds to 96 counts, or four-percent duty at the chosen 20 kHz rate. If the on-state rectangular model applies, the minimum such repeated pulse corresponds to 1.92 W average power, far above the 0.020 W one-count increment.

A requested 0.5 W would correspond to 25 ideal counts, or about 0.521 microseconds. It lies inside the timer's numeric range but below the assumed effective-pulse limit. Do not silently round every nonzero request up to 96 counts: that would deliver much more heat than commanded. Define how unattainable requests are represented and handled by the controller.

Evaluate frequency changes without confusing two different limits

Reducing the illustrative PWM rate to 5 kHz while retaining the 48 MHz count clock gives 9,600 intervals per period. The nominal power increment becomes 0.005 W. The same 2 microsecond minimum effective pulse still occupies 96 counts, but now represents one-percent duty and 0.48 W average power in the ideal model.

This change improves the low-power command map, but it can alter thermal ripple, electrical interference, audible behavior and controller timing. It does not establish that 5 kHz is suitable for the actual installation. Compare candidates against the complete application requirements and retain the difference between numerical count resolution and a driver-imposed minimum effective on-time.

Treat pulse skipping as a different time-domain delivery method

A controller may distribute qualified pulses across multiple periods to obtain a lower long-term average than one pulse every period. That is a different modulation pattern, not a fractional pulse produced by the existing timer. For example, one assumed 2 microsecond pulse in every four 20 kHz periods gives 0.48 W averaged over that four-period interval in the same ideal electrical model.

The heater and temperature sensor respond to the resulting sequence, not only to its long-term mean. Evaluate pulse energy, thermal ripple, measurement synchronization and the maximum gap between delivered pulses. Keep protective interruption independent of this averaging strategy. A controller must still be able to stop heating as required instead of completing an unnecessary burst merely to preserve its arithmetic average.

Verify the low-power command map at the heater boundary

Sweep commands through the low-count region while recording actual heater-terminal voltage and current with suitable instrumentation. Identify missing pulses, pulse-width distortion and the first repeatable delivery region. Compare measured energy per pulse and average power with commanded values. The software timer register is not evidence of the electrical energy that reached the heater.

At short pulses, voltage-current channel timing mismatch can strongly bias calculated power. Establish the measurement timing separately rather than shifting traces until the desired average appears. Repeat relevant checks as heater resistance and source conditions change. A map established with a cold resistance and ideal bench source may not describe the installed hot-state operating point.

Give the temperature loop a realistic actuator specification

The temperature-control design needs the reachable power levels, minimum effective pulse, update timing and off-state definition. It can then select an appropriate strategy for low demand and avoid integrating indefinitely toward an unreachable command. Changing control gains alone does not create an actuator level between two physical delivery states.

For a ChipSimple heater enquiry, provide the finished resistance range, driver and timer configuration alongside the required minimum maintenance power. Include mounting and temperature-ripple requirements when evaluating a frequency or burst change. Keep the heater's electrical and thermal limits under their established owners. The completed review should explain how each low-power request becomes real terminal energy, not merely how many decimals the controller calculates.