CUSTOM DESIGN // EMBEDDED COMPUTING // RADIO ASTRONOMY
Data Collection & Processing Module (DCPM)
A fully custom, self-contained computing and signal-processing engine designed to operate directly at the antenna feed — eliminating RF losses, enabling simultaneous multi-receiver operation, and bringing real-time scientific workflows to the most demanding deep-space missions.
DesignerAlex Nersesian K6VHF
StatusOperational
Form Factor7″ × 4.5″ × 3.5″ / 3 lbs
Storage640 GB internal + 4 TB USB
3
Simultaneous Receivers
640GB
Internal Storage
10A
DC-DC 5V Output
2
OS Simultaneous
5Gb/s
USB 3.0 per Port
Promotional Video// DCPM IN ACTION
0:00 / 1:33
DCPM — Data Collection & Processing Module | Animated technical overview | Designer: Alex Nersesian K6VHF | Duration: 1:39
Overview01
The Data Collection and Processing Module (DCPM) is a specialized, high-performance computing system designed to collect, process, store, and analyze data directly at the antenna feed. By moving computation to the source, the DCPM eliminates the limitations of traditional control-room PCs and enables powerful real-time scientific workflows.
Built for flexibility and field operation, the DCPM acts as a self-contained computer, data-storage platform, and signal-processing engine, supporting multiple receivers and complex mission-critical tasks simultaneously.
◆ Design Philosophy
All hardware, firmware, and software — including the custom automation scripts, multi-OS integration layer, and remote management system — were designed in-house by Alex K6VHF. No off-the-shelf data acquisition solution existed that met the demanding requirements of simultaneous multi-receiver deep-space operations at the feed point.
Problem & Solution02
✗ Traditional Setup — Limitations
RF losses across long coaxial cable runs from feed to control room
Single-device limitation — typically only one receiver at a time
High latency; processing far from the antenna source
Complex wiring harness and control-room dependency
Bandwidth bottleneck between antenna and processing PC
Difficult to expand without major infrastructure changes
✓ DCPM — How It Solves Them
Processing at the feed — zero long-coax RF loss
Up to 3 simultaneous SDR/receivers via USB 3.0 at 5 Gb/s each
Real-time processing and analysis with sub-millisecond latency
Fully self-contained — no control-room dependency for acquisition
High-speed fiber or Ethernet backhaul for processed data only
Modular — multiple DCPM units can be interconnected
Key Capabilities03
📡
Multi-Receiver Support
Simultaneously operates up to 3 SDRs or receivers via USB 3.0 at full 5 Gb/s bandwidth per port — enabling parallel science missions on a single platform.
⚡
Feed-Point Processing
Computation happens directly at the antenna feed. Eliminates all RF loss from coaxial runs and maximizes signal fidelity entering the digital domain.
💾
Massive Local Storage
640 GB internal NVMe/SATA storage with support for external USB 3.0 SSDs up to 4 TB — enabling uninterrupted high-bandwidth data capture for extended sessions.
💻
Dual-OS Architecture
Linux (Raspberry Pi 5) and Windows (NUC Box G5) operate simultaneously. Run Linux signal pipelines alongside Windows-only scientific tools — no compromises.
🔗
Remote Access
Full TCP/IP remote desktop access to both operating systems simultaneously via Ethernet or fiber optic. No monitor, keyboard, or mouse needed at the feed point.
⚙
Custom Automation
Fully scriptable in Python, C++, C#, Java, and Bash. Custom automation handles scheduling, data tagging, real-time analysis pipelines, and mission-specific workflows.
🔌
10A Power Rail
Built-in DC-DC 5V converter capable of delivering up to 10A — powers the compute modules, active cooling, and connected SDR devices from a single feed-point supply.
🔥
Harsh Environment Ready
Housed in a rugged aluminum enclosure designed for outdoor antenna feed deployment. Sealed against moisture and rated for field operation in challenging weather.
📈
Expandable Architecture
Multiple DCPM units can be networked together for additional SDR channels, distributed sensing, or custom sensor module integration — linearly scalable.
System Architecture04
Fig. 1 — DCPM V1.05 block diagram: Raspberry Pi 5 (Linux) + NUC Box G5 (Windows 11) linked via LAN hub, with 128 GB SD + 256 GB SSD + 256 GB SSD storage, DC-DC 5V/10A power rail, fiber optic SC/APC adapter, and external Ethernet/USB 3.0/HDMI interfaces
The DCPM integrates two independent compute nodes — a Raspberry Pi 5 running Linux and an Intel NUC Box G5 running Windows — into a single compact enclosure. Both nodes share a common 5V / 10A power rail and communicate via internal Gigabit Ethernet.
Rear I/O — Ethernet · Fiber SC/APC · RPi USB3.0 · DC 12V · CTRL · PC I/O
I/O Panel Features
ETHERNET — 1 GBit sealed connector FIBER — SC/APC optical uplink RPI USB 3.0 — Direct SDR connection DC 12V — Field power input CTRL — Multi-pin control bus PC I/O — NUC Box G5 interface
SDRs and receivers connect directly via USB 3.0 to whichever compute node runs their driver stack. Processed data flows out via a single Ethernet or fiber uplink to the control room, dramatically reducing cable complexity while eliminating all analog RF loss.
Hardware05
DCPM V1.05 — Top view, specification label
DCPM — Internal layout, dual-node wiring
DCPM — Rear I/O: Ethernet · Fiber · USB 3.0 · DC 12V
Rear I/O panel features field-ready mil-spec connectors: Ethernet (1 GBit), Fiber Optic SC/APC, RPi USB 3.0, DC 12V in, CTRL multi-pin, and PC I/O — all sealed for outdoor antenna feed deployment.
DCPM internal view — Raspberry Pi 5 (left), NUC Box G5 (right), DC-DC power rail, LAN hub, fiber adapter, and wiring harness. Designed and assembled by Alex K6VHF.
Technical Specifications06
Raspberry Pi 5 — Linux Node
CPU
Broadcom BCM2712 — Quad-core ARM Cortex-A76 @ 2.4 GHz, 64-bit
5V output, up to 10A — powers both nodes and connected SDRs
Enclosure
Aluminum — 7″ × 4.5″ × 3.5″, 3 lbs
Application Examples07
1. Interferometer / Hydrogen Line / DSN / Pulsar Reception
Multi-channel data acquisition for interferometry baselines, hydrogen line spectroscopy, DSN probe tracking, and pulsar detection and timing — all running simultaneously on one DCPM unit.
2. B210 / Pulsar Reception & Processing
USRP B210 integration for wideband coherent receiver applications, pulsar timing, and high-bandwidth signal processing leveraging the Linux node's GNU Radio / UHD driver stack.
3. Multi-Unit Expansion Configuration
Two or more DCPM units networked together, doubling the receiver count and storage capacity. Ideal for multi-dish interferometry or simultaneous multi-band observations.
4. 4-SDR Simultaneous Reception
Four independent SDR receivers operating simultaneously on a single DCPM unit. Each receiver runs its own software stack, enabling parallel multi-frequency or multi-polarization captures.
Mission Applications08
🛰
Artemis-II
High-bandwidth reception and real-time processing support for the DSES Artemis-II lunar mission tracking campaign.
🌍
Deep Space Network
Detection and tracking of DSN spacecraft signals. Multi-receiver simultaneous coverage of X-band, S-band, and Ka-band probe downlinks.
🎊
Hydrogen Line
Long-duration 1420 MHz drift-scan and pointed observations. Local storage enables multi-hour uninterrupted data capture without network dependence.
☆
Pulsar Detection
High-time-resolution pulsar observations requiring dedicated compute at the feed. Sub-millisecond timing precision with the Pi 5 RTC and GPS PPS synchronization.
📉
Interferometry
Multi-baseline VLBI-style experiments with synchronized data capture at each antenna. Local storage + timestamped files for offline correlation.
🌌
EVE / EAE Programs
Earth-Venus-Earth and Earth-Apophis-Earth radar experiments requiring high-bandwidth coherent reception and real-time Doppler tracking.
Dual-OS Architecture09
🐧 Linux Node — Raspberry Pi 5
GNU Radio + UHD for SDR signal processing
Python automation scripts and pipelines
PRESTO pulsar timing suite
GQRX, SDR++, rtl_sdr tools
Custom C++ real-time signal analyzers
SSH + VNC remote access
💐 Windows Node — NUC Box G5
WSJT-X, SDR Console, SDRuno
Spectrum Lab, SpectrumSpy
Windows-only vendor drivers and tools
C# / .NET scientific automation
Remote Desktop (RDP) access
Java-based mission control software
Both operating systems run simultaneously and are accessed remotely via TCP/IP — no monitor, keyboard, or mouse is required at the feed point. The DCPM presents itself on the network as two independent machines, both accessible from the control room over a single Ethernet or fiber uplink.
Designer10
AK
Alex Nersesian K6VHF
RF Engineer — CETUS-RF | Senior Space Scientist — DSES
Alex designed and built the DCPM entirely from scratch — hardware integration, enclosure design, firmware configuration, multi-OS architecture, custom software stack, and automation scripts. The DCPM is one of several custom engineering solutions Alex has developed for amateur radio astronomy and deep-space mission support, combining professional RF engineering with practical embedded computing.
DCPM V1.05 — Top view, specification label
DCPM — Internal layout, dual-node wiring
