What is the difference between corporate and series feeds?

Corporate feed uses a binary tree of power dividers (Wilkinson, T-junction, or hybrid) to split the input power into N branches, each feeding one element. Every element has an independent path from the input, making the amplitude and phase of each element independently controllable. Path lengths to all elements are equal (for broadside radiation), providing frequency-independent beam pointing. The disadvantage is complexity and size: an N-element array needs N minus 1 dividers and log2(N) divider stages, with total feed loss of 0.5 to 2 dB depending on N and frequency. Series feed connects all elements along a single transmission line, with each element tapped off at a point along the line. It is much more compact and simpler to fabricate, but the path length to each element is different, creating a frequency-dependent progressive phase shift that causes beam squint: the beam direction changes with frequency. This limits bandwidth to applications where beam squint is acceptable or desired (frequency-scanned arrays).

How does amplitude tapering reduce sidelobes?

A uniform amplitude distribution (all elements with equal power) produces the narrowest beam but the highest sidelobes (minus 13.2 dB for a linear array). Tapering the amplitude so that center elements receive more power and edge elements receive less reduces sidelobes at the cost of wider beamwidth and reduced directivity. Taylor distribution provides a specified sidelobe level with minimum beamwidth degradation: minus 25 dB SLL has approximately 1.1 times the beamwidth of uniform. Minus 30 dB SLL: approximately 1.15 times. Minus 40 dB SLL: approximately 1.25 times. Chebyshev distribution provides equal sidelobes across the entire pattern (all sidelobes at the same level). Taylor is preferred in practice because it transitions smoothly to the far sidelobe region, while Chebyshev has a step that causes implementation sensitivity. The taper also reduces the aperture efficiency: minus 25 dB Taylor has approximately 0.9 taper efficiency (0.46 dB loss), minus 40 dB has approximately 0.75 (1.25 dB loss).

How are feed networks designed for phased arrays?

In active phased arrays, the feed network architecture depends on whether beamforming is done in the analog or digital domain. Analog beamforming: a corporate feed distributes the signal to phase shifters at each element. The phase shifters (PIN diode, MEMS, or ferrite) steer the beam. Each element has a T/R module with PA, LNA, and phase shifter. The feed network must handle the full TX power and maintain amplitude and phase accuracy. Digital beamforming: each element has its own ADC/DAC, and beamforming is done in the DSP. The feed network is simply the RF distribution to the T/R modules, with no beamforming function. Digital beamforming enables simultaneous multiple beams, null steering, and MIMO. Hybrid beamforming (used in 5G mmWave): combines analog and digital. A small number of digital chains (4 to 8) each feed a sub-array of elements through analog phase shifters. The feed network for each sub-array is a compact corporate or series feed. The trade-off is flexibility (fewer beams than full digital) versus cost and power (fewer ADCs and DACs).

Antenna Array

Feed Network

/feed net-werk/

Distributes power to array elements with controlled amplitude/phase. Corporate: binary tree, freq-independent pointing, 0.5-2 dB loss. Series: single TL, compact, beam squint with freq. Amplitude taper: Taylor −25 dB SLL (1.1× BW), Chebyshev (equal SLL). Phase: progressive shift steers beam. Phased array: per-element phase shifter. Digital BF: per-element ADC/DAC. 5G hybrid: sub-array analog + digital.

Corporate: freq-stable

Series: compact

Loss: 0.5-2 dB

Understanding Feed Networks

The feed network is the circulatory system of an antenna array. It determines how power is distributed among the elements (amplitude taper for sidelobe control) and the relative phase at each element (beam steering). A poorly designed feed network degrades the entire array's performance through loss, amplitude/phase errors, and impedance mismatch.

For modern 5G mmWave arrays with 64 to 256 elements, feed network design is one of the most challenging aspects of the system. The network must fit in a compact form factor, maintain tight amplitude (<1 dB) and phase (<10°) accuracy across the band, handle the required power, and operate at millimeter-wave frequencies where losses are high.

Feed Architecture Comparison

Type	Beam Squint	Size	Loss	Application
Corporate	None	Large	0.5-2 dB	Phased array, BTS
Series	Yes	Compact	0.3-1 dB	Freq-scan, radar
Hybrid (corp+ser)	Partial	Medium	0.5-1.5 dB	2D arrays
Space feed	None	3D volume	Low	Lens, reflectarray
Active (T/R)	None	Per-element	N/A (active)	AESA, 5G

Common Questions

Frequently Asked Questions

Corporate vs series?

Corporate: binary divider tree, equal path lengths, freq-stable pointing. N-1 dividers, 0.5-2 dB loss. Independent element control. Series: single TL, compact, but beam squints with frequency. Path length difference = progressive phase. Good for freq-scanned arrays, limited BW for fixed beam.

Sidelobe control?

Uniform: −13.2 dB (narrowest beam). Taylor: specified SLL with min BW penalty. −25 dB: 1.1×BW, 0.46 dB gain loss. −40 dB: 1.25×BW, 1.25 dB loss. Chebyshev: equal SLL everywhere. Taylor preferred (smooth transition). Trade: lower SLL = wider beam = less directivity.

Phased array feeds?

Analog: corporate + per-element phase shifter (PIN, MEMS, ferrite). Full TX power through feed. Digital: per-element ADC/DAC, BF in DSP. Simultaneous multi-beam, MIMO. Hybrid (5G mmWave): 4-8 digital chains each feed sub-array with analog phase shifters. Trade: flexibility vs cost/power.

Related Terms

← Feedthrough Ferrite →

Array Design

Request a Quote

Need feed network design, array simulation, or beamforming optimization? Contact our team.

Get in Touch