DWG NJ-FD5-001REV B Blue-team eyes · Red-team teeth

NIGHTJAR FD-5

A clamshell Raspberry Pi 5 DualOps cyberdeck — field-ready hardware for authorized blue-team and red-team work.

Build a field-ready Raspberry Pi 5 cyberdeck from the board up — two 7-inch displays, a 1TB NVMe spine, split power rails, and three network lanes in one clamshell deck for authorized blue-team capture, red-team simulation, and field forensics. Seat each part in the drawing below, then wire, test, and commission it.

COMPUTE
Pi 5 · 16GB
STORAGE
NVMe 1TB
DISPLAY
2× 7-inch
POWER
3 split rails
NETWORK
3 lanes
MODES
5 postures
General arrangementSHEET 1 OF 5 · NJ-FD5-001
Click a balloon to seat a part, again to wire, again to test — then commission on Sheet 5.
FIG 1 · GENERAL ARRANGEMENTItem 01 — Compute bay
NIGHTJAR FD-5 — keyed assembly drawingTop-down exploded plan of the clamshell deck: lid with two displays and an OLED strip, a base deck carrying the Pi 5, NVMe carrier, switch panel, input, and auth bay, a side rail of power, network, radio, and sensor modules, and an optional AI expansion on the PCIe path. Numbered balloons key each part to the parts schedule.HINGECHASSIS04OLED09DISP 1 · CTRL07DISP 2 · WORK08PI 5 · 16GB01M.202NVMe 1TB03AUTH BAY15SWITCH PANEL10INPUT SET11PSU 5.1V/5A05PD BATTERY06ETH ×212WI-FI ×213GPS + RTC14AI HAT+ (OPT)16SCALE NTS · TOP-DOWN EXPLODED

Click balloon 01 to seat it into Compute bay.

[01]Raspberry Pi 5 — 16GBUNSEATED
Role
Primary compute board
Placement
Compute bay, base deck, on shock-isolated standoffs with airflow clearance.
Connects
Rail A power in. PCIe lane out to M.2 carrier. HDMI/DSI to lid. USB to hub.
Depends
Root of the build. Everything downstream assumes the brain is seated and boots clean.
Socket
Compute bay · rail A
Modes
BLUE · RED · FORENSICS · FIELD · LABBEAST
Pass
Boots from known-good media and reports stable telemetry.
PARTS SCHEDULE0/15 core seated
ITMDESCRIPTIONRLST
Optional · deliberate tradeoff
○ unseated ◔ seated ◑ wired ● tested0/15
UNIT
NIGHTJAR FD-5
DWG NO
NJ-FD5-001
POSTURE
POWER
RAIL A
SCOPE
N/A
BUILD
EMPTY 0/15

Field manual — the ordered build

Sixteen steps from a bare board to a commissioned deck: 15 core stages plus one optional AI tradeoff. Work top to bottom — each step states its goal, actions, and pass criteria.

01

Verify the Brain Boots

Prove the Pi 5 is healthy before it disappears into the chassis.

  • Inspect the Pi for physical and connector damage.
  • Attach active cooling temporarily.
  • Boot from a known-good microSD image.
  • Confirm HDMI, keyboard, mouse, Ethernet, Wi-Fi, and USB all respond.

“Seat the brain before you build the body.”

Pass — Pi boots, outputs video, accepts input, reaches the network, holds a stable temperature.

02

Prepare NVMe Storage

Make NVMe the main boot and data path.

  • Install the M.2 HAT+ or the selected integrated carrier.
  • Mount the 1TB NVMe SSD securely.
  • Flash Raspberry Pi OS 64-bit (or chosen base) to NVMe.
  • Keep the microSD as rescue media only.

“NVMe is the spine. microSD is only the parachute.”

Pass — System boots from NVMe and exposes the expected capacity.

03

Partition + Encrypt Data Layout

Create a clean operational storage model that contains dirty data.

  • Create the OS root.
  • Create an encrypted captures volume.
  • Create an encrypted case-data volume.
  • Create labs and wordlist volumes.
  • Create a scratch volume for disposable PCAPs, extractions, and caches.
/ 80–120GB OS root
/opt/tools 100GB tool workspace
/data/captures Encrypted packet captures
/data/cases Encrypted case + engagement records
/data/labs Docker-compose labs + local ranges
/data/wordlists Compressed wordlists + indexes
/data/reports Notes, screenshots, exports, reports

“Separate evidence, labs, reports, and scratch. Dirty data stays contained.”

Pass — Encrypted volumes mount only after authentication; every path is clearly labeled.

04

Install Cooling + Chassis Base

Build a stable thermal and mechanical foundation.

  • Install the active cooler or case-integrated cooling.
  • Mount the Pi + NVMe carrier on standoffs.
  • Verify the airflow path is clear of cables and panels.
  • Secure the fan cable away from hinge movement.

“Heat kills field gear. Airflow is a component.”

Pass — Pi idles and loads within thermal limits with no cable interference.

05

Build Split Power Rails

Stop displays, radios, hubs, and storage from fighting over power.

  • Create Rail A for the Pi 5 + NVMe.
  • Create Rail B for the two displays.
  • Create Rail C for the USB hub, radios, GPS, OLED, and Ethernet accessories.
  • Add fusing or protection where practical.
  • Label every rail physically and in the UI.

“One weak rail makes the whole deck unreliable. Split the load.”

Pass — Deck stays stable with both displays, NVMe, keyboard, Ethernet, and radios attached.

06

Install Hardware Switch Panel

Give the operator physical control over major subsystems.

  • Install four low-profile toggles.
  • Switch 1 → display rail.
  • Switch 2 → USB hub / accessory rail.
  • Switch 3 → fan override.
  • Switch 4 → auxiliary module rail.

“Software profiles are nice. Physical switches are decisive.”

Pass — Each switch visibly and electrically controls its assigned subsystem.

07

Mount Dual Displays

Create a two-screen clamshell workstation.

  • Mount the touchscreen as the left / lower CONTROL display.
  • Mount the HDMI panel as the right / upper WORK display.
  • Use short right-angle HDMI adapters where needed.
  • Route the USB touch cable separately from power.
  • Test both displays at boot and after wake/resume.

“Left screen controls the deck. Right screen does the work.”

Pass — Both screens initialize reliably and hold correct orientation.

08

Install Input System

Support both mobile and serious typing workflows.

  • Pair the mini wireless keyboard + trackpad.
  • Add a dock / stow slot for the 60% mechanical board.
  • Add a mouse, trackball, or pointer if desired.
  • Map hotkeys: mode switch, screenshot, terminal, capture start/stop.

“The mini keyboard is for movement. The mechanical board is for work.”

Pass — All input devices work without interfering with Wi-Fi or USB storage.

09

Install Network Interfaces

Create separate management, capture, and wireless lanes.

  • Onboard Ethernet → management / stable uplink.
  • USB gigabit Ethernet → capture / isolated lab lane.
  • Wi-Fi adapter 1 → authorized monitor-mode auditing.
  • Wi-Fi adapter 2 → connectivity / AP / comparison.
  • Physically label every interface on the chassis.

“Three network lanes: manage, capture, radio.”

Pass — Each interface has a stable name, a visible role, and a clear profile assignment.

10

Install OLED Status Display

Expose essential state without opening a dashboard.

  • Wire the OLED over I2C / the selected display bus.
  • Show host, mode, primary IP, VPN, CPU temp, battery, free storage, capture state.
  • Use low-refresh logic to save power.

“Status belongs on hardware, not buried in a menu.”

Pass — OLED updates reliably and stays readable in field lighting.

11

Install GPS + RTC

Improve field metadata and offline time accuracy.

  • Install the RTC with a coin cell or backup power.
  • Install the GPS module where the antenna has sky exposure.
  • Verify time persists after power loss.
  • Keep GPS metadata optional and operator-controlled.

“Bad timestamps ruin evidence. Fix time before field work.”

Pass — Offline boot has correct time; GPS status is visible only when enabled.

12

Install Hardware Auth Bay

Make strong authentication part of the physical deck.

  • Cut a recessed slot for the YubiKey / Nitrokey.
  • Configure SSH with hardware-backed keys.
  • Configure Git signing if needed.
  • Wire up the password-vault unlock workflow if needed.

“The key is part of the machine. No key, no operator state.”

Pass — Admin and sensitive workflows require the token where appropriate.

13

Build the OS Profile Layer

Run multiple operational postures from one deck.

  • Create a base OS: stable desktop, terminal, networking, update path.
  • Create mode profiles for Blue, Red, Forensics, Field, and Lab Beast.
  • Each profile powers a different hardware subset and exposes different status fields.

“Five postures. One deck. Different power, tools, and stance.”

Pass — Operator can boot or switch into the right profile without reconfiguration chaos.

14

Install Tooling by Category

Build a structured tool environment, not a junk drawer.

  • Install by role. Every tool needs a mode and a purpose.
  • Keep categories isolated and documented.
  • Keep tool directories recoverable.
Base Terminal
tmux, shell config, fzf, ripgrep, jq, yq, curl, wget, git, Python, Go, Rust
Blue Team
tcpdump, tshark, Wireshark, Zeek, Suricata, osquery, Sigma/YARA, log parsers, IR templates
Red Team (authorized)
Nmap, range scanners, web proxy/test tools, dir enum, nuclei (owned assets), lab-only validation
Wireless (authorized)
Kismet, Wireshark, aircrack-ng suite, adapter diagnostics, GPS survey logging where permitted
Forensics
hash tools, Sleuth Kit, metadata tools, binwalk, bulk extractors, timeline utils, read-only mount helpers
Reporting
Markdown editor, screenshot util, diagramming, Pandoc, PDF export, report templates

“Install by role, not by hype. Every tool needs a mode and a purpose.”

Pass — Tool directories are organized, documented, and recoverable.

15

Create Recovery + Maintenance Plan

Make the deck field-repairable and hard to brick.

  • Create a rescue microSD card.
  • Create a labeled backup target.
  • Use stable disk IDs, not volatile /dev/sdX names.
  • Document the restore procedure in the UI.
  • Back up configs, scripts, profile files, and report templates.

“Recovery is not optional. A field deck needs a parachute.”

Pass — Operator can restore from documented media without guessing device names.

16 · OPT

Optional Edge AI Decision

Decide whether AI hardware belongs in this exact deck.

  • Default: keep the PCIe path dedicated to NVMe.
  • Use AI HAT+ 2 only if local inference beats clean NVMe architecture.
  • Alternative: USB accelerator, networked workstation, or remote inference box.

“AI is optional. NVMe is core. Do not trade the spine for a gimmick.”

Pass — Operator understands and accepts the tradeoff before seating the optional module.

W01 Split display power from Pi power to reduce brownout risk.
W02 Do not duplicate NVMe carriers. Use either the M.2 HAT+ or an integrated NVMe case design.
W03 AI HAT+ 2 is a deliberate tradeoff. Keep PCIe dedicated to NVMe unless you redesign the expansion plan.
W04 Use stable disk IDs in recovery workflows. Avoid blind /dev/sdX commands.
W05 Red Mode stays unarmed until scope is loaded and verified.

Frequently asked questions

Common questions about building and operating a Raspberry Pi 5 cyberdeck.

What is the NIGHTJAR FD-5 cyberdeck?

The NIGHTJAR FD-5 is a clamshell Raspberry Pi 5 DualOps cyberdeck — a field-ready hardware deck built for authorized blue-team and red-team work. It pairs two 7-inch displays, NVMe-first storage, split power rails, and multiple network interfaces in a single buildable chassis, and this guide walks the assembly socket by socket.

Why use a Raspberry Pi 5 16GB for a cyberdeck?

The 16GB Raspberry Pi 5 gives the deck enough memory for GUI work, browsers, packet-capture tools, local databases, containers, and multi-pane workflows at once. Less RAM forces compromises the moment you run a proxy, a browser, and a capture pipeline together, so 16GB is the practical floor for a dual-display operations deck.

NVMe or microSD — which should boot the deck?

NVMe is the spine. The build flashes the operating system and encrypted data partitions to a 1TB M.2 NVMe SSD and keeps the microSD strictly as rescue media. NVMe is faster and far more reliable than an SD card for sustained capture and database work, and a labeled rescue card is the parachute if the main drive ever fails.

Do I need the AI HAT+ on a Raspberry Pi 5 cyberdeck?

No — the AI HAT+ 2 is a deliberate optional tradeoff, not a core part. The cleanest architecture keeps the single PCIe lane dedicated to the NVMe drive, so the guide flags a conflict when you seat both. If local inference matters more than clean storage, accept the redesign and move AI to USB, a networked workstation, or remote inference instead.

Why split the power into separate rails?

Displays, radios, USB hubs, and storage all draw current in bursts. Putting them on one rail causes brownouts that look like random crashes. The FD-5 splits power into Rail A (Pi + NVMe), Rail B (displays), and Rail C (accessories), with physical switches so a weak subsystem never destabilizes the whole deck.

Is a cyberdeck legal, and what is the FD-5 actually for?

The FD-5 is a dual-use operations deck for authorized work only: cybersecurity education, lab validation, internal assessment, blue-team capture, red-team simulation, field forensics, and wireless auditing inside owned or explicitly permitted environments, CTFs, and cyber ranges. Red Mode stays unarmed until engagement scope is loaded and verified, and every target in this guide is a fictional owned-lab asset.

What are the five operating modes?

The deck boots five postures from one build: Blue (defensive capture and analysis), Red (scoped, authorized offense), Forensics (read-only, hash-first, radios cold), Field (low-power, dimmed, long battery), and Lab Beast (full-power bench operation). Each mode powers a different hardware subset and exposes different status fields.