TTD for wideband phased array beamforming: Eliminating beam squint in AESA

Radar, communications, and sensing systems increasingly rely on phased array antennas to achieve agile, precise beam steering without mechanical motion. As bandwidths widen to support higher data rates and better resolution, traditional phase-shifter-based architectures reach their limits, particularly due to beam squint and degraded performance at band edges.

True time delay (TTD) techniques, implemented using time delay units (TDUs), address these challenges by providing frequency-independent delay across the array, enabling accurate wideband beamforming in advanced active electronically scanned arrays (AESAs).

Phased array antenna fundamentals: AESA architecture and electronic beam steering

Phased arrays use multiple antenna elements arranged in uniform linear arrays (ULAs) or uniform rectangular arrays (URAs), with each element driven by a controlled phase or time delay. By adjusting these controls, the array forms narrow beams in desired directions and suppresses unwanted directions, eliminating the need for mechanical steering.

In AESA architectures, each element—or small group of elements—typically has its own transmit/receive (T/R) module, enabling multi-beam and multi-frequency operation for radar, satellite links, and advanced wireless systems.

Figure 1 Simplified phased array antenna features multiple elements with controllable phase or time delay, enabling electronic beam steering without mechanical motion. Source: Qorvo

Phased array performance may be summarized by two system-level figures of merit. Equivalent isotropically radiated power (EIRP) is expressed in dBm (referenced to 1 mW) or dBW (referenced to 1 W) and equals transmit power plus antenna gain (in dBi), assuming no cable or connector losses.

Gain-to-noise-temperature ratio (G/T) measures receive sensitivity by comparing antenna gain to system noise temperature; higher G/T values correspond to better detection and signal quality, especially important for satellite and deep-space links.

Wideband beam squint: Why phase shifters fail in high-bandwidth phased arrays

Wider instantaneous bandwidths have benefits: operation over multiple channels or bands, improved range resolution in radar, and higher data rates in communications. However, wideband operation exposes a key limitation of pure phase-shifter-based beamforming: the phase shift needed for a given steering angle depends on frequency, so a single-phase setting at the center frequency does not steer all frequencies to the same angle.

This misalignment creates beam squint, where the main beam moves with frequency across the band. At the center frequency, the beam points at the desired angle, but at the band edges, it becomes under-steered or over-steered, redistributing gain and degrading EIRP and G/T in the intended direction.

Figure 2 Illustration of beam squint in a phased array: a single-phase setting at the center frequency steers the beam correctly at F₀, but under‑steers at F_max and over‑steers at F_min. Source: Qorvo

Why beam squint occurs for wider-band systems is illustrated in Figure 3. Beam steering using phase shifts per element attempts to ensure that energy at each element in the array arrives at the same phase for coherent summation in the beamforming network. When the beam is steered off-boresight (θ > 0), some energy arrives at one edge of the array before it arrives at the opposite edge; the extra path length is N * d * sin θ, where N is the number of elements across the array, θ is the beam steering angle, and d is the element spacing.

Figure 3 Path‑length difference in an off‑boresight phased array, showing how the extra distance N * d * sin θ at the farthest element translates into a frequency‑dependent phase shift requirement and leads to beam squint when only fixed phase shifters are used. Source: Qorvo

Dividing this longer path length by the wavelength at F₀ and multiplying by 360 degrees gives the phase shift that must be applied at the farthest element. At F_max the wavelength is shorter, so the required phase shift is larger; at F_min the wavelength is longer, so the required phase shift is smaller, which is why a single phase shift per element at F₀ leads to under‑steering at F_max and over‑steering at F_min.

In high-gain, narrow-beam arrays, even modest squint can cause large link-budget penalties at the band edges, while lower-gain, wide-beam arrays are more tolerant of the same angular movement.

Whether an array needs true time delay depends on the relationship between beam squint and beamwidth. Smaller arrays with broader beams can often accept the squint induced by phase-only steering, whereas large arrays with narrow beams and wide bandwidths require TTD to maintain beam pointing and gain across the full band. Design factors such as array size, element spacing, scan angle, and center frequency all feed into this assessment, with wide scan angles and high frequencies typically increasing sensitivity to timing errors.

Figure 4 Example gain patterns for the same beam movement in high‑gain, narrow‑beam and low‑gain, wide‑beam antennas, showing how beam squint produces much larger gain loss at the band edges in high‑gain arrays and necessitates the use of true time delay. Source: Qorvo

True time delay vs phase shifters: Performance comparison for wideband arrays

Traditional phase shifters implement a fixed insertion phase at a given frequency, which works well for narrowband systems since phase and delay can be treated interchangeably over a small bandwidth. As bandwidth increases, the frequency dependence of phase shift becomes problematic: the constant phase setting no longer corresponds to the correct time delay across the entire signal spectrum, causing beam squint and waveform distortion.

True time delay elements, by contrast, provide a fixed time delay, so the resulting phase shift increases linearly with frequency. This linear phase slope ensures that signals of all frequencies within the band experience the same effective delay through the array, aligning their phases at the combining point and maintaining beam direction. So, for wideband, high-performance arrays, especially those requiring fine resolution or long range, TTD becomes essential rather than optional.

In practice, many systems adopt hybrid architectures that combine phase shifters and TDUs. Phase shifters can handle fine steering around a nominal direction or serve narrowband modes, while TDUs provide coarse or wideband delay control to prevent squint across the full band. The design challenge is to balance cost, die area, power, and complexity against bandwidth and performance goals, choosing where in the array hierarchy (element, subarray, or tile level) TDUs should be inserted.

Integrating TDUs into AESA tile and subarray architectures

In AESAs, antenna elements are often grouped into modular tiles, each containing beamforming ICs (BFICs), RF front-ends, and other control circuitry. Within each tile, TDUs can provide precise synchronization across elements or subarrays, ensuring that signals combine coherently in the desired direction over wide bandwidths. Typical TDUs offer delay steps in the picosecond range, allowing fine-grained control of beam pointing and compensation for channel-to-channel variations.

Wide instantaneous bandwidth is particularly important for high-resolution radar, where shorter pulse widths improve range resolution but demand larger bandwidth. For example, a ULA operating at a 10 GHz center frequency with a 1.5 GHz signal bandwidth and 16 elements with half-wavelength spacing may be limited to a 60-degree scan range if phase-only steering is used and signal degradation must be held within acceptable bounds. Beyond such limits, true time delay, rather than approximate phase-based delay, is required to preserve beam integrity and resolution across the full spectrum.

Calculating TTD requirements for ULA design

In uniform linear arrays, the delay required at each element can be derived from the desired scan angle, element spacing, and operating frequency. A phase shifter with a given resolution, say a 6-bit device with a least significant bit (LSB) of 5.625 degrees, corresponds to a particular minimum time delay increment at the operating frequency; at 10 GHz, this phase step translates to approximately 1.5 picoseconds of delay. To replace such a phase shifter with a TDU, the delay unit must support at least the same or finer time resolution to maintain equivalent steering granularity.

For arrays with up to 16 elements per side and scan angles up to about 60 degrees, the total required delay at the outer elements can reach around 650 picoseconds. Designers may implement this total delay using cascaded TDUs or a combination of coarse and fine delay stages, distributing the delay across the RF chain to meet both performance and implementation constraints. This approach enables larger or more agile arrays while maintaining precise control of beam pointing over wide frequency ranges.

TDU topologies: Switched lines, ATLs, and LC networks

TDUs can be realized with several circuit topologies, each offering trade-offs in terms of noise, insertion loss, die area, and linearity. Switched delay lines use multiple physical line lengths selected by RF switches to create discrete delay values; they tend to provide low noise and low insertion loss but require more chip area, especially at longer maximum delays.

Artificial transmission lines (ATLs) use synthetic line structures to achieve compact delay implementations, trading size for higher loss and potentially increased noise. LC-based delay networks implement analog delay using lumped inductors and capacitors, offering fine control and tunability but adding design complexity and sensitivity to component tolerances.

Digital delay lines, familiar with digital signal processing, use switched digital paths to provide quantized delays, making them attractive for architectures that need both coarse and fine control and may integrate closely with digital beamformers. Each topology represents a different balance between delay precision, footprint, insertion loss, and linearity, and system-level simulations are typically required to identify the optimal choice for a given application.

Integrating TTD in RFICs: Monolithic microwave circuits for wideband phased arrays

Historically, true time delay could be implemented with coaxial cables, optical fibers, or microstrip and stripline networks, but these approaches often struggle with size, weight, and cost in large, high-frequency arrays. Monolithic microwave integrated crcuits (MMICs) now provide a more practical and scalable solution, integrating delay elements, switches, and equalization within compact RFICs.

Advances in CMOS, GaAs, and MEMS technologies have reduced the size and power consumption of TTD circuits while improving bandwidth and delay resolution.

A typical MMIC-based TDU may combine switched or artificial transmission lines with wideband distributed gain amplifiers and gain-slope equalizers to compensate for frequency-dependent loss and maintain flat group delay across the band. Integration with beamforming ICs and RF front-end modules allows designers to place TTD functionality at the element, subarray, or tile level, trading semiconductor area against array-level performance and flexibility.

Why TTD matters: Wideband array performance and future-proof AESA design

From a system perspective, true time delay provides consistent beam steering across wide frequency ranges, improving signal quality, resolution, and link margin at the band edges. By effectively eliminating beam squint across the operating bandwidth and preserving coherent combining across the array, TTD enhances both EIRP and G/T, directly impacting radar detection performance and communications reliability.

In demanding environments and applications, such as defense radar, satellite payloads, and next-generation wireless backhaul, this can be decisive in meeting performance requirements. Beyond immediate performance gains, TTD also supports future proofing. As bandwidth requirements continue to grow and spectral environments become more complex, architectures that already incorporate wideband-capable TDUs and hybrid TTD/phase-shifter beamforming are better positioned to adapt without wholesale redesign.

For engineers and decision-makers, the key takeaway is that while phase shifters remain suitable for narrowband or cost-sensitive systems, TTD is becoming a critical enabler of competitive, high-performance wideband phased arrays.

David Schnaufer is technical marketing communications manager at Qorvo, where he leverages his extensive technical and strategic experience to develop insightful, thought leadership content. Throughout his career at Qorvo, he has served several roles, including senior manager of strategic marketing and product marketing manager.

TTD for wideband phased array beamforming: Eliminating beam squint in AESA

Become a member

Become a member

Become a subscriber

Become a sponsor