Understanding Forward/Reflected Power
Misconceptions can cause problems for engineers when measuring power in RF designs.
By: Richard A. Scholl, Chief Technology Officer, Advanced Energy Industries
Contents
•The background
•Forward/reflected power
•What reflected power represents
•What it really means
•Recap
•FET amplifiers
Measurements made in the RF domain are steeped in tradition. Like most traditions, RF concepts are passed on by word of mouth as well as by written legend. Also like most traditions, the concepts are imperfectly understood by many who need to know better. In particular, the concept of forward and reflected power is often misunderstood. By properly understanding these concepts, engineers can improve performance in their RF designs.
The background
The use of RF in the first half and more of this century was centered around radio communication, as the very name implies. In a radio station, power is generated by a transmitter, usually housed comfortably in a shack, and the RF power is led to the (usually) remote antenna by a cable, or transmission line, which in short wave stations can be several wavelengths long.
An antenna tuner is generally placed at the antenna end of the transmission line. The function of this circuit is to match the antenna. Thus, the tuner transforms the complex impedance of the antenna into a fixed real impedance.
This is desirable for two reasons. First, every transmitter has some impedance into which it can deliver the most real power (usually set by design to 50 W). If the antenna's impedance is different from this ideal value, more voltage and/or current will be required of the transmitter to get a given radiated power than if the load was matched to the transmitter. This limits the maximum power of the transmitter.
Second, if a line is terminated in its characteristic impedance (also usually 50 W), the RF voltage along that line will be uniform. If it is not terminated this way, a standing wave will be set up along the line. This can stress the line, because the voltage peaks, for a given power, are higher than the voltage on a matched line, and the line could arc over at these voltage maxima.
For both reasons, the RF engineer matches the antenna, meaning that he or she adjusts the antenna tuner until the transformed impedance is equal to the line impedance, which is generally also the transmitter's ideal load resistance. To detect when the antenna is matched, the engineer does not physically walk along the line to measure the actual standing wave. Rather more conveniently, the engineer uses a device known as a directional wattmeter, which can be places at any convenient point in the cable.
The wattmeter measures the voltage, current, and their relative phase. It reports on its dials quantities called forward power and reflected power.
Forward/reflected power
A standing wave on a transmission line can be thought of as being composed of two traveling waves, one moving toward the load (the "forward" wave) and one moving in the opposite direction (the "reflected" wave). These waves, moving through the transmission line, interfere with one another to produce the standing wave. Each of these waves have a voltage amplitude—the "forward" voltage Vf and the "reflected" voltage Vr. The conceptual problem comes from the dials of the directional wattmeter, which are not calibrated in volts but rather in watts.
To derive watts from voltage, the wattmeter (or coupler) assumes that the forward wave is terminated in a 50 W resistor, and calculates |Vf|2/50 and presents the result as forward power. Similarly, it assumes the reflected wave is also terminated in a 50 W resistor, and calculates |Vr|2/50 and presents the result as reflected power. This can cause serious confusion because usually the waves are not impressed on any such resistor.
If the antenna and tuner are replaced by a 50 W resistor, then the forward wave is terminated in this resistor, and the indicated forward power actually will be the power dissipated in the load resistor. Since the 50 W line is terminated in its characteristic impedance, there will be no reflected wave and no standing wave, because the line is matched, and Vr and the reflected power are zero. [If the tuner is adjusted to transform the antenna impedance to exactly 50 W, the same thing occurs, except that the power is radiated from the antenna instead of being dissipated in a resistor.]
If the load at the end of the line is not equal to 50 W, then there will be a standing wave. Since the forward power is defined as the power in a 50 W load, and since the load is no longer 50 W, the forward power no longer represents the load power.
What reflected power represents
One very common misconception is that the reflected power represents the power that must be absorbed by the generator. To best portray this misconception, let's look at an extreme case.
Suppose a lossless line is terminated by a pure open circuit, and suppose that the line is exactly one wavelength long at the operating frequency. In this case the current at the generator will be zero. As a result, the current in the generator's internal impedance will be zero. Hence, there is no power dissipated in the generator.
Yet there is a standing wave on the line, so there are forward and reflected waves. The directional wattmeter will show a forward and reflected power, and in fact they will be equal to one another, and equal to the power which would be dissipated in a 50 W resistor if the voltage at the generator were impressed upon it. But there is no such resistor and so there is no power lost anywhere (again assuming no line loss).
What it really means
The above example raises a few questions in an engineer's mind. Is there any measure of sense that we can get from the forward and reflected power readings? What do they really mean, anyway, if forward power doesn't represent the power delivered to the load and reflected power doesn't represent power absorbed by the transmitter?
First of all, the difference between the forward and reflected power has a concrete meaning: it is the power delivered to the load (see sidebar). In this sense, one can imagine that the forward power is being impressed upon the load, which is "rejecting" the reflected part and absorbing the balance.
An engineer has to be careful not to take this imagery too far. It must not be inferred, for example, that the generator actually has to generate the forward power. The generator has to produce the power lost in the load, the line, and its internal impedance, and no more, because any power produced must be delivered somewhere, and these are the only places for it to be lost.
In the case of the open lossless line, there is no power lost in the load, line, or the generator, and so there is no real power lost anywhere, but there are forward and reflected power figures. These powers are not real energy flows, but rather are calculated figures which represent the power that would be generated in a 50 W resistor if the forward and reflected waves were impressed upon it. Since there is no such resistor, there is no such power.
Recap
Let's recap. What can be inferred from the forward and reflected power readings on our directional wattmeter and what cannot?
- If the reflected power is zero there will be no standing wave on the line and no excess stresses on it.
- If the reflected power is zero the generator can deliver its maximum rated power into the load.
- The difference between the forward and reflected powers is the real power delivered to the load, but:
- The generator does not actually generate a power equal to the forward power, and:
- The reflected power figure is not absorbed by the generator.
If the last above is true, then why are generators rated to handle some maximum reflected power? The answer relates to the maximum voltage and current the generator can deliver. If a generator is rated to deliver, say, 1000 W, this does not mean that it can do so into any load. As the load resistance is increased, to get 1000 watts one needs more and more voltage (P=V2/R, so V
R) and sooner or later the generator has produced all the voltage it can, and the power is limited. Similarly, as the load resistance is decreased, the current goes up (P=I2R, so I/R) and sooner or later the generator cannot deliver any more current.
Finally, it is remembered by those who used to work with tube-type generators that a mismatched load can cause the plate of the tube to glow red. Does this not mean that excess power is being dissipated there, and doesn't this indicate that reflected power is being absorbed by the tube?
Certainly the higher temperature of the anode indicated that the plate power was increased, but this does not mean that the reason was absorption of some kind of power coming in from the outside. In a tube-type transmitter, the tube has a substantial voltage across it during conduction, resulting in large conduction losses.
Actual loss depends upon the design mode (A, AB, B, C, etc.) but can easily be larger than, and are rarely less than 1/3 of the output power. To achieve high efficiency the tube must be biased well beyond cutoff and must be operated on a tight load line. If there is reflected power as a result of a reactive load, the load line becomes an open curve. This means that for a portion of the cycle, substantial voltage exists across the tube at the same time high current is being drawn through it. This results in the large plate dissipation.
FET amplifiers
In the case of modern field-effect transistor (FET) amplifiers, the output FETs are operated as switches. The FET, operated as a switch, is nearly an ideal one, with typical "on" resistance of considerably less than 1 50 W. Thus, properly designed FET amplifiers feature losses well under 10% of the output power. These losses can increase or decrease, depending upon the load impedance, and in some cases the presence of reflected power can even result in decreased dissipation. Modern protection circuits for such amplifiers look only at the power lost in the output device and the voltage across it, and reflected power is not measured.
About the Author:
Richard Scholl, Chief Technology Officer, Advanced Energy Industries, Inc., 1625 Sharp Point Dr., Fort Collins, CO 80525. Tel: 970-407-6426; Fax: 970-407-5426.