Mobile Solid State Linear Amplifier Basics

By Telstar Electronics (Manufacturer of the Skywave DX350)

Mobile RF amplifiers have been in existence for many years. Alas, in the beginning only vacuum tube types were available. Tubes functioned well but have some inherent problems:

  • Large physical size
  • Require high voltage
  • Require a significant warm-up period
  • Dissipate considerable power (heat) while idling
  • Fragile to mechanical shock

With the advent of solid state technologies, the transistor solved many of the vacuum tube shortcomings. Transistors have improved steadily over the years in power handling capability, frequency range, and reliability. There are basically two types of transistors used today in RF linear amplifiers. The BJT (bi-polar junction transistor) is basically a current controlled device, while the FET (field effect transistor) is a voltage controlled device. Each have their own advantages/disadvantages. Generally, in the mobile RF linear amplifier arena the BJT is the preferred device due to its superior gain at lower source voltages. Some typical examples of high power RF BJTs are the Toshiba 2SC2290 and 2SC2879. These devices were designed to produce adequate power and gain at 28 MHz, while operating from lower voltage sources (12-14VDC) that are present in a mobile environment.

What is a linear amplifier? Simply put, a linear amplifier is a device that increases an input signal level without distorting the original wave shape. In an RF amplifier, maintaining the original modulated signal integrity at the output is essential in reduction of harmonic and spurious emissions. These distortion related byproducts are directly responsible for audible anomalies at the receiving station as well as interference to other bands/frequencies. Getting into this “linear region” of transistors involves maintaining a level of DC current in the device, which will ensure that transistor is turned on even when no signal is applied. The level of this idling current is dependent on “class of operation” that the designer chooses.

The amplifier class of operation chosen depends on a number of factors including power output, degree of linearity required, efficiency, topology, and signal type. Linearity is always a trade-off with efficiency. Although there are many classes of amplifiers defined, the two that are in common use for most high power RF amplifiers are class-AB and class-C. Class-C amplifiers have very high efficiencies, but don’t operate in the linear mode. That means that for input signals that are varying greatly in amplitude, the transistors are turning on and off. This distorts the output signal and creates the harmonic and spurious emissions mentioned earlier. This unwanted behavior would certainly apply to AM and SSB signals. However due to its high efficiency, class-C is an excellent choice for signals that don’t vary in amplitude such as FM and CW. The method of choice for a high power RF amplifiers (tube or transistor) that will operate in a linear fashion on AM, SSB, FM, and CW is class-AB. Class-AB has been derived as a compromise between class-A (very linear but low efficiency) and class-B (borderline linearity but good efficiency). Class-AB maintains only a small bias idling current during amplification to ensure the active devices stay on, in their linear region. This-AB biasing technique gives the designer the best of both worlds, good linearity and efficiency. Since we are concentrating this discussion on BJT type devices it’s only fair to mention that although class-AB is an excellent method of bias for this type of amplifier, it comes with some design challenges.

The major problem bi-polar transistors have operating in class-AB is that once the idling bias is set, it drifts significantly with changes in transistor temperature. As the transistor heats, the bias current steadily creeps upward. This leads to increased heating of the transistor, further increasing the current. This condition is known as thermal runaway. If allowed to continue, this runaway bias current will continue to rise until the transistor overheats and is destroyed. Because of this destructive phenomenon, some type of bias temperature compensation must be incorporated in the amplifier design. This compensation often takes the form of a sensing diode. This diode is “thermally attached” to the transistor, most often using the amplifier heat sink. This diode now has the ability to thermally track the transistor and provide compensation to the bias current via additional circuitry. Another design challenge with class-AB stems from the inherent low impedance of the typical high power BJT device. A designer must realize that certain steps must be taken to provide an extremely low impedance feed point for the bias. Failure to do this will result in the bias being cut off when amplifier input signals are applied. This can effectively negate the class-AB biasing under certain signal conditions, reverting the amplifier to a non-linear mode.

Let’s discuss the topology of an amplifier. Topology here refers to the basic configuration. A single transistor amplifier is often referred to as “single-ended”. Amplifiers that contain several BJT transistors almost always use a “push-pull” configuration. Sometimes these configurations are used together where a single device is used to drive a push-pull arrangement. The “push-pull” configuration is very attractive to designers for a number of reasons:

  • Adds the power of two transistors
  • Increases overall input & output impedances for easier matching
  • Symmetric configuration cancels even-order harmonics
  • Configuration supports class-AB biasing for linear operation
  • Easily accommodates the use of transformer coupling

For these reasons, the push-pull arrangement is usually preferred. These push-pull modules can also be conveniently combined for even greater output power.
In keeping with the “maximum power transfer theorem”, matching is one of the most important aspects of any RF amplifier. Without proper input matching, some of the driving signal to the amplifier will be lost. This condition is evident by a high SWR (standing wave ratio) looking into the amplifiers input. Without proper output matching, some of the power developed within the amplifier cannot be delivered to the load. This wasted power remains within the amplifier as excess heat. This is certainly undesirable, and if left unchecked can destroy the amplifier. Matching comes in two forms, narrowband and broadband. Narrowband matching is accomplished using tuned LC (inductor/capacitor) circuits. Bandwidth and other parameters are modified by various configurations and component values. Broadband matching is accomplished with transmission-line or conventional transformers. A great majority of solid-state amplifiers use transformers because they require no tuning over a wide band, and this scheme fits well with the push-pull topology.

In any well designed amplifier along with the basics, consideration should be given to protect the unit under certain adverse operating conditions. Protections can include:

  • Reverse polarity (DC power)
  • Excessive current (DC power)
  • Excessive heat sink temperature
  • Excessive output SWR
  • Excessive RF input power

Some or all of the above mentioned protections could be included in a given amplifier. Keep in mind that all these functions add a significant cost factor. At minimum, the reverse polarity, excessive current, and excessive heat sink temperature protections probably should be incorporated in the design.
We have only really scratched the surface of this subject, but hopefully this has given you some insight on a few of the factors involved in the design of solid state BJT amplifiers for mobile operation…