GBD-DART: Diamond Array Radio Telescope

• GBD-DART is a modular low-frequency radio interferometer that uses diamond-shaped LPDA tiles for dual-polarisation, enabling detailed studies of pulsars and solar transients.
• Its innovative antenna design with triangular weighting and low sidelobe response ensures precise beam control across the 50–80 MHz and 130–350 MHz frequency windows.
• Advanced digital receiver systems and real-time processing pipelines facilitate accurate calibration, sensitive transient detection, and scalable upgrades for future astrophysical research.

The Gauribidanur Diamond Array Radio Telescope (GBD-DART) is a modular, low-frequency, transit-mode radio interferometer optimized for time-domain astrophysics, especially pulsar studies and monitoring of solar radio bursts. Located at the Gauribidanur Radio Observatory, India (13.604°N, 77.427°E), GBD-DART employs diamond-shaped tiles of log-periodic dipole antennas (LPDAs) arranged for dual-polarisation and low sidelobe response. The system's architecture and digital signal processing provide high sensitivity, polarization capability, and scalable design for targeted scientific applications across the 50–80 MHz and 130–350 MHz radio windows (Bane et al., 2022, B et al., 31 Jan 2026, B et al., 31 Jan 2026).

1. Scientific Motivation and Key Objectives

GBD-DART is designed to address the unique challenges and opportunities inherent to low-frequency radio astronomy. The primary scientific drivers include:

• Pulsar Observations: Detection and timing of bright, steep-spectrum pulsars across 50–350 MHz, exploring emission properties, rotation measures (RM), and spin evolution.
• Solar Phenomena: Continuous monitoring of solar radio transients (notably Type III bursts), leveraging substantial fractional bandwidth and fast time resolution.
• Astrophysical Transients: Enabling studies of fast radio bursts (FRBs) and propagation effects—dispersion, scintillation, and Faraday rotation—amid high sky background.
• Training and Instrumentation: GBD-DART acts as a scalable educational platform for radio instrument development, digital signal processing, and data analysis (B et al., 31 Jan 2026).

Observation at these frequencies is significant for probing pulsar radius-to-frequency mapping, spectral turnovers, propagation through the interstellar medium, and for cross-comparison with contemporaneous facilities such as LOFAR and the MWA.

2. Antenna Array Architecture

GBD-DART comprises several generations of array designs, unified by their use of LPDAs:

• 50–80 MHz System: The initial array features 16 broadband LPDAs with impedance ≈50 Ω and VSWR <2 over 40–440 MHz. Arranged along a 75 m north–south baseline (5 m spacing), two 8-element subarrays are coherently combined. LPDAs are vertically oriented in E-plane, providing an east–west fan beam with full-width half-maximum (FWHM) ≃110°, and a 3° FWHM in declination at 65 MHz (Bane et al., 2022).
• 130–350 MHz Tiles: Subsequent systems (“DART-I” and “DART-II”) utilize diamond-shaped tiles each containing 64 LPDAs (organized as 32 dual-polarised elements), within a 5.9 m × 5.9 m square, rotated by 45° to achieve diagonals of ≈8.35 m. Each dual-pol unit is realized using off-axis LPDA pairs, tilted 23° from zenith for beam shaping. The array factor employs a triangular (Pascal’s triangle) weighting ([1,2,3,4,3,2,1]) along both axes to suppress sidelobes (<−20 dB) in the E–W and N–S directions (B et al., 31 Jan 2026, B et al., 31 Jan 2026).

The following table summarizes the two core architectures:

Array Generation	No. of LPDAs	Frequency Range (MHz)	Tile Footprint
50–80 MHz Prototype	16	50–80	75 m N–S, 5 m spacing
130–350 MHz Tile	64	130–350	5.9 m × 5.9 m diamond

The diamond configuration, paired with the triangular column weighting, yields measured primary beams of ≃15° FWHM at 200 MHz (element HPBW ≈60° in E-plane, ≈90° H-plane), with high sidelobe suppression critical for time-domain studies in the presence of strong interfering sources (B et al., 31 Jan 2026).

3. Analog Signal Chain and Array Performance

Each LPDA feed incorporates a balun and low-noise amplifier (LNA; ≈20 dB gain, noise figure NF ≃0.7–1.35 dB), followed by short low-loss cables and phased combiners to preserve signal integrity and polarization purity. The analog chain includes:

• Filters: Chebyshev/TEM high-pass (HPF) and low-pass (LPF) designs set passbands (e.g., 130–350 MHz, or 50–80 MHz).
• Gain Stages: Successive amplifiers yield ≈50 dB net gain per chain.
• RF-over-Fiber: For 130–350 MHz, analog signals are converted to optical for transmission over ≈250–350 m single-mode fiber to the receiver room, ensuring phase stability and minimizing cable loss (B et al., 31 Jan 2026, B et al., 31 Jan 2026).

System Temperature: At these frequencies, the sky temperature—dominated by Galactic synchrotron emission—exceeds 8,000 K at 65 MHz and remains ≳200 K at 175 MHz; receiver noise is subdominant ($T_{\text{sys}} \simeq T_{\text{sky}}$). The system equivalent flux density (SEFD) is approximately 25,000 Jy for a single 16-pair tile at 175 MHz (B et al., 31 Jan 2026).

Sensitivity Metrics: For a 16 MHz bandwidth and 1-hour integration, the minimum detectable continuum flux density is ≈50 mJy (SNR=1). For typical sub-array integrations (e.g., 0.5 s with 2 MHz bandwidth), the sensitivity degrades to ≈250 Jy, validated via drift scans of strong radio sources (Cyg A, Tau A, Vir A) (B et al., 31 Jan 2026).

4. Digital Receiver Systems and Signal Processing Pipelines

Multiple digital backend architectures are deployed:

• 50–80 MHz System: Uses a CASPER ROACH board with Xilinx Virtex-5 FPGA and AD9480 8-bit ADCs sampling at 90 Msps (2nd Nyquist zone, 45–90 MHz). Channelization employs a 4-tap FIR polyphase filterbank (2048-point FFT, Δν ≃ 44 kHz), with adjacent channel isolation ≃–49 dB, and native time resolution of 1 ms (Bane et al., 2022).
• 130–350 MHz Tiles: Digital backend is centered on a portable dual receiver (PDR) capturing 16 MHz IF bands with 8-bit ADCs at 33 MSps; data is timestamped, packetized (UDP), and streamed to disk. RAM-disk buffering permits low-latency triggered dumps for transient capture. Real-time coherent and incoherent dedispersion is executed via DSPSR, PRESTO, and PSRCHIVE in a pipeline architecture, employing multiprocessing and multi-threaded implementations to achieve near real-time reduction (1:1 observation:processing time) on AMD R9/Intel i9 platforms (B et al., 31 Jan 2026).

Channelization, Stokes parameter calculation, folding, and single-pulse detection are implemented, with support for both high-resolution (128 µs × 64 kHz) and low-resolution spectra. The pipeline accepts digital baseband (DADA) formats and supports both real and simulated calibration data for validation.

5. Observing Modes, Calibration, and Instrumental Response

GBD-DART operates exclusively in transit mode—sources drift through the fixed primary beam (nominally 15° FWHM at 175 MHz) over ∼1 hour. The beamforming design, combined with cable-length and digital delay tuning (granularity 11.11 ns at 50–80 MHz, and calibrated phase alignment at 130–350 MHz), supports precise control of beam shape and mainlobe phase. Artificial pulsed noise-diodes are injected at the first analog stage for amplitude and polarization calibration, while in-field VNA measurements ensure coherence and minimal X–Y leakage (<–20 dB) (B et al., 31 Jan 2026). The instrumental response is cross-validated via drift scans of satellites (e.g., ORBCOMM), solar continuum transits, and simulated polarised pulses.

6. Representative Scientific Results

GBD-DART has demonstrated capabilities for both solar and pulsar time-domain science:

• Sky Background and Solar Drift: At 65 MHz, drift scans reveal the expected sky brightness distribution, matching theoretical convolutions with the ∼110° wide fan-beam (Bane et al., 2022). Solar drift at 200 MHz shows excellent agreement with all-sky models.
• Solar Bursts: Type III solar radio bursts have been recorded (e.g., 27 July 2024 at 200 MHz), with subsecond time and ∼64 kHz spectral resolution. Event morphology aligns with parallel international observatories (B et al., 31 Jan 2026).
• Pulsar Detections: First light included B1919+21 (SNR ≃23 at 65 MHz, ≃7 at 175 MHz), subsequently confirming Crab, J0953+0755, J0837+0610, and J1136+1551 with full-Stokes and single-pulse archives (Bane et al., 2022, B et al., 31 Jan 2026, B et al., 31 Jan 2026).
• Polarimetric Results: Rotation measure and polarisation fraction measurements are routinely achieved. Crab pulsar spin-down monitoring over ∼200 days yields period derivatives consistent with Jodrell Bank ephemerides to parts in 10³; full-Stokes timing achieves typical ToA precision ≾100 µs per epoch (B et al., 31 Jan 2026).

Key observational parameters are summarized below:

Pulsar	SNR	RM (rad/m²)	Linear Pol. (%)	Single-Pulse Rate
J0953+0755	300	+3.08 ±1.14	≈70	3978/40 min (≈1.7 s⁻¹)
J0534+2200 (Crab)	10–16	–39.4 ±1.8	≈50	706/11.3 h (≈1 min⁻¹)
J1136+1551	120	+1.42 ±1.9	≈65	≃690 total

The measured minimum detectable ΔS_min, S/N, and instrument noise closely match theoretical predictions based on system temperature and effective collecting area (Bane et al., 2022, B et al., 31 Jan 2026).

7. Scalability, Limitations, and Future Upgrades

GBD-DART is designed for modular expansion:

• Bandwidth and Tiling: Planned digital backend upgrades using next-generation FPGA (RFSoC) receivers will support full 200 MHz instantaneous bandwidth. Commissioning of additional 64-LPDA tiles is planned to double collecting area and sensitivity ($\text{SEFD} \lesssim 12$ kJy), enabling sub-25 mJy detection in 1-hour integrations (B et al., 31 Jan 2026).
• Beamforming: Time-delay and FFT-based digital beamforming will facilitate multi-beam operation and tracking across larger zenith angles, enhancing coverage for transient searches.
• Calibration and RFI Mitigation: Enhanced on-site calibration (hot/cold loads, drone-based beam mapping, and improved noise diode procedures) are targeted to minimize systematics and improve polarimetric fidelity.
• Pipeline Enhancements: Bandwidth expansion and real-time GPU acceleration for coherent dedispersion; pipeline cluster scalability (e.g., Kubernetes deployments) to manage high-throughput data for FRB capture and archiving (B et al., 31 Jan 2026).

These upgrades aim to support a wider set of astrophysical targets—including deep searches for new pulsars, intensive timing of binary/millisecond systems, and rapid response to transient phenomena—and to provide a scalable training resource for next-generation radio instrumentation specialists.

• (Bane et al., 2022) "A prototype for pulsar observations at low radio frequencies using log periodic dipole antennas"
• (B et al., 31 Jan 2026) "GBD-DART-I : Pulsars and transient source observation between 130 MHz and 350 MHz at Gauribidanur"
• (B et al., 31 Jan 2026) "GBD-DART-II: 175 MHz Polarimetric Observation of Pulsars from Gauribidanur and a New Pulsar Signal Processing Pipeline"