Are inline RF wattmeters really averaging

Hams often speak of inline RF wattmeters as being ‘averaging’, but are they?

The ITU defines Mean Power (or average power) as the average power supplied to the antenna transmission line by a transmitter during an interval of time sufficiently long compared with the lowest frequency encountered in the modulation taken under normal operating conditions.

Principles of operation of simple inline RF wattmeters

The simplest instruments use a directional coupler feeding a half wave rectifier with capacitor input filter, a series resistance and d’Arsonval meter movement (moving coil).

The moving coil meter deflects as a result of the torque due to coil current opposing the hair springs, and the mean deflection is proportional to the mean or average current, so the deflection responds to average current.

If the rectifier / filter was ideal, the filter capacitor would charge to the peak RF voltage, and the meter current would be proportional to the peak RF voltage. Since peak RF voltage is proportional to the square root of RF power, meter deflection would be proportional to the square root of RF power, and to the extent that the meter responds to average current, then deflection would be proportional to average of the square root of RF power, not average or mean power.

Practical rectifier / filters have imperfections, the diode characteristic has a knee, the diode and capacitor have internal resistance, so the filter doesn’t charge instantly, or hold its charge indefinitely and the effect of these imperfections varies from instrument to instrument and complex waveform to complex waveform and is quite unpredictable.

The following video is a practical demonstration using a transmitter of 20W continuous sine wave keyed on and off with 50% duty cycle so that its average power is 50% of the continuous waveform.

In this case, the response to the 50% duty cycle waveform is 7W, less than the correct value of 10W as a result of the practical aspects of the rectifier circuit.

In fact it is quite challenging to produce a low-cost inline wattmeter that responds to the average power of complex waveforms such as that used, or speech.