How to Maximize Range in Radar Systems Based on GaN Power Amplifiers

Radar has become indispensable for countless applications, including military surveillance, air traffic control, space missions, and automobile safety. Among the most challenging situations for designers is long-range radar, where the return signal is extremely weak, ambient and circuit noise degrade the signal-to-noise ratio (SNR), and “pulse droop” becomes an issue.

While power amplifiers (PAs) based on gallium nitride (GaN) offer significant efficiency and other benefits compared to devices using older processes, designers need to take a system-level approach to minimize pulse droop and its effects. This will ensure superior performance in long-range radar systems.

This article briefly reviews radar operation and examines the pulse-droop problem. It then introduces a state-of-the-art S-band GaN PA from Analog Devices and an accompanying evaluation board, and suggests tactics for compensating for and minimizing pulse droop.

Radar principles and issues

The principle of radar is straightforward: a system transmits a short on/off pulse of RF energy, and a receiver picks up the signal reflected from the target. The time lag between the transmitted pulse and its echo determines the distance (range) to the target, as both propagate at the speed of light.

While this simple pulse is sufficient in principle, it is inadequate for the real world of multiple targets, especially at distances of tens, hundreds, and even thousands of miles. These longer-range radar systems face two issues:

• The return signal from a distant target is very weak, and the SNR is poor.
• Distinguishing between multiple targets at a distance requires resolving closely spaced echoes, assuming their return signals have not been distorted and overlap.

Signal strength is very low due to the unavoidable physics of the situation and the 4^th-power rule. This is shown by the classic radar equation that relates factors of the radar performance and practical effects:

 Equation 1

Where:

P_r is the expected receive power

P_t is the transmit power

G_t is the antenna gain

G_r is the receive gain

λ is the wavelength of radar operation

σ is the effective cross-sectional area of the target

R is the range from the antenna to the target.

The equation shows that round-trip attenuation primarily determines range losses, as R, raised to the fourth power, is in the denominator.

The obvious way to overcome range losses is to increase the transmitted signal’s peak power and lengthen the pulse to increase its overall energy. However, this approach blurs the return and has overlap to the extent that multiple objects appear lumped together (Figure 1).

Figure 1: These radar-image sketches show an ideal pulse response (left) and a degraded pulse response and range (right). (Image source: Analog Devices)

A more sophisticated way to improve performance is to shape, modulate, and “compress” the transmit pulse to improve the range resolution and SNR. Pulse compression allows the radar system to resolve multiple targets in a tight grouping rather than see them as blurred return pulses overlapping at the receiver.

Droop pulse-power problems and solutions

While increasing the pulse power is possible, this creates other issues. One is that higher power aggravates the PA-centric phenomenon of pulse droop (Figure 2).

Figure 2: This nominally rectangular radar pulse shows overshoot, pulse width, rise/fall times, and droop. (Image source: Analog Devices)

Pulse droop is the undesired reduction in the pulse amplitude from beginning to end, typically characterized in decibels (dB). This reduction decreases the range over the pulse length since the combination of pulse amplitude and width determines the range of the radar as an integrated power level.

Droop occurs even when using efficient solid-state GaN PAs such as the state-of-the-art ADPA1106ACGZN from Analog Devices. This 46 decibel referenced to 1 milliwatt (dBm) (40 watts) device, with 56% power added efficiency (PAE) across a bandwidth of 2.7 gigahertz (GHz) to 3.5 GHz, is well-suited to the pulse-power needs of S-band radar systems.

What causes pulse droop?

Droop is mainly due to two distinct mechanisms:

1: The PA performance is changed by the sudden pulse current. This introduces dissipation and other thermal effects that result in the shifting of critical device performance parameters. As the GaN PA transistor-channel temperature increases due to Joule self-heating, which is the product of current density and electric field, the amplifier's output power is reduced. Figure 3 illustrates the relationship between the channel temperature, drain current, and drain voltage for one operating point of a GaN transistor with a pulse width of 100 microseconds (µs).

Figure 3: Shown is the relationship between the channel temperature, drain current, and drain voltage for one operating point of a GaN transistor with a pulse width of 100 µs. (Image source: Analog Devices)

Although GaN devices are relatively efficient, some power is lost to heat, so effective thermal management is required for the best results. Depending on the pulse width, pulse repetition frequency (PRF), and duty cycle, a combination of one or more cooling approaches such as fans, heatsinks, cold plates, or liquid cooling will be needed.

As the duty cycle increases at a constant pulse width, the time the PA spends off between pulses decreases. This means the PA has less time to cool and is at a higher temperature at the rising edge of the subsequent pulse. In the limiting case of a 100% duty cycle (continuous wave (CW)), there is no time for the PA to cool, and its temperature is constant at its maximum.

This leads to a tradeoff. As the duty cycle increases, the part's average temperature increases, reducing peak and average output power. However, the magnitude of the temperature rise during the pulse decreases, which means there is less droop and more consistency over the pulse width. Thus, the tradeoff becomes a balance between less droop and more power.

2: The second consideration is the power supply. Due to the fast transient of the pulsed power, the PA power supply is challenged to cope with the sudden demands for high power while maintaining the voltage rail at the required value. As with the thermal problem, the solutions are known, but the implementation is critical.

It begins with adding large charge-storage (bulk) capacitors along the PA bias line and placing ceramic or tantalum bypass capacitors close by. This is seen in the ADPA1106-EVALZ evaluation board (Figure 4, left) which has decoupling capacitors placed close to the amplifier, and its associated “pulser board” with large charge-storage capacitors that maintain power levels during wide pulse widths (Figure 4, right).

Figure 4: The top of the ADPA1106-EVALZ evaluation board (left) shows the unique layout and tight positioning of the decoupling capacitors; the bottom side shows the aluminum heat spreader (middle); the associated pulser board holds the high-value bulk capacitors used to supply needed current during pulse transients (right). (Image source: Analog Devices)

The evaluation board is designed to address the unique challenges of optimizing the application of the ADPA1106. It comprises a two-layer printed circuit board (pc board) fabricated from a 10 mil Rogers 4350B copper-clad board mounted to an aluminum heat spreader. The spreader assists in providing thermal relief to the device and mechanical support to the pc board. Mounting holes on the spreader allow it to be attached to a heatsink. Alternatively, the spreader can be clamped to a hot and cold plate.

Although using large-value storage capacitors is not ideal as they increase the radar array’s size, weight, and cost, they are often the only viable approach. Further, the relative position, orientation, and type of decoupling capacitors used near the amplifier will influence their effectiveness and pulse fidelity. At the RF frequencies of PAs, such as the ADPA1106, the impact of parasitic capacitance and inductance must be considered carefully and factored into the design.

Droop results versus pulse width, repetition frequency

The ADPA1106 PA was tested for droop performance in two ways: by varying the pulse width under constant pulse repetition frequency, and by varying the duty cycle while maintaining constant pulse width. In both tests, pulse droop was measured from 2% into the pulse period through the end of the pulse to remove the effect of the initial overshoot.

The first test uses a varying pulse width at a fixed pulse repetition frequency of 1 millisecond (ms) (Figure 5). There is a high correlation between increasing pulse width and increasing pulse droop. At the maximum tested pulse width, the droop approaches 0.5 dB, which is the maximum level of droop that is usually acceptable at the system level.

Figure 5: Testing with a fixed pulse-repetition frequency of 1 ms shows the correlation between increasing pulse width and increasing pulse droop. (Image source: Analog Devices)

In addition, due to thermal effects, the peak and average output power decreased slightly with increasing pulse width, while the downward slope at the tail end of the longest pulse width slightly increased. This may indicate that self-heating effects are beginning to affect the thermal management of the package and the heatsink below it.

To assess the effects of the duty cycle, the ADPA1106 was tested again using a constant pulse width of 100 microseconds (µs) while changing the duty cycle (Figure 6). As the duty cycle increases towards 100%, the PA has less time to cool between pulses and is at a higher temperature at the rising edge of the subsequent pulse. As a result, the part's average temperature increases, pulse amplitude decreases, and the magnitude of the temperature rise during the pulse decreases.

Figure 6: Using a constant pulse width while varying the duty cycle shows that the change in magnitude variation decreases as the duty cycle increases. (Image source: Analog Devices)

This demonstrates the tradeoff. It shows the negative impact of reduced peak and average output power due to the higher absolute temperature of the part. However, there is the benefit of less droop and greater output-power consistency over the entire pulse width because the PA’s temperature change is less over the duration of the pulse.

Conclusion

Achieving maximum range in radar systems requires a system-level approach to minimizing pulse droop. This includes effective thermal management and the addition of bulk capacitors to the power supply. To demonstrate how to balance the required tradeoffs, this article used actual test data utilizing the ADPA1106 high-efficiency PA to assess droop by varying two critical pulse parameters and using suitable cooling. The results showed that the device delivered very low droop under 0.3 dB over a typical range of pulse conditions.