It is officially Ronin. The first PCB is ordered.

The project has a name now: Ronin. Same BC250 build, same frosted PCTG covers, now with a neon samurai wordmark to match the personality it has been developing.

The first custom PCB is ordered, too. It cost more than I expected for a first assembled order. Next time, the plan is to order a stencil and handle the reflow myself. For this revision, I am waiting for the board to arrive so I can check it on the bench.

Build it, remix it, and make it your own. Ronin is a rapidly evolving work in progress, so parts, CAD and instructions will change. If you have questions, feel free to reach out.

Give the LEDs their own power

The RONIN ESP32 LED ADAPTER / REV A is designed to take LED power from the PSU's SATA supply while the ESP32-P4 display remains USB-powered. The board measures 44 × 34 mm and provides protected LED power, two data channels and 3.3 V-to-5 V data translation. A wired JST VH input joins it to the PSU's SATA connector; there is no 15-pin SATA socket on the PCB.

My two V1.0 Rainbow on a Matchstick boards by Blamm each have ten LEDs and no daisy-chain output connector. The adapter feeds them independently: channel A goes to one stick and channel B to the other. The intended firmware can then address twenty pixels without adding a data-output connector to either board. That physical LED output is still work ahead, not a tested feature.

The board has a shared manual 5 V / 12 V selector for its LED outputs. The Matchsticks require the 5 V setting. The 12 V option is for a compatible 12 V addressable strip in a different configuration. The data signals remain 5 V, and the selector does not detect the strip voltage for you.

These are visualizations of the ordered fabrication layers and current CAD, not photos of an assembled board. The IC and fuse package bodies are simplified; the 5 V shunt shows the intended setup rather than a confirmed factory assembly. The PCB is now modeled in the SolidWorks assembly, but it is awaiting arrival, electrical and thermal checks, and a physical fit test. For now, the public repository gets images and documentation. The PCB source, Gerbers, BOM and placement files stay out of the repository.

Storage gets a mount and a visible cable route

The updated assembly includes the relocated 2280 NVMe SSD, its extension board, two printed extension mounts and an explicitly modeled ribbon cable. The aim remains to put the drive in the main cooling fans' intended airflow path while making mounting and cable routing easier to review.

The public CAD download remains an earlier snapshot; this post documents the updated local assembly. A cable drawn in CAD is not a cable tested in a case. Printed fit, screw lengths, connector access, real bend limits and service slack still need checking. Then the installed extension needs a storage-link stability test and a sustained-load temperature check. Moving the drive does not change the BC250's host-link generation.

The extra PCB and final cable routing also belong in the next airflow-model update. The existing CFD report and animation remain rough, with unresolved geometry, mesh and convergence limitations. They do not establish chip temperatures or a safe overclock.

Four harness branches, no guessing by wire color

The harness now has four branches: SATA power to the adapter, display signals to the adapter, adapter to Matchstick A, and adapter to Matchstick B. The earlier plan for a cable from the first LED stick to the second has been superseded for these V1.0 boards.

The draft harness and crimping guide explains connector and terminal matching, open-barrel crimping, contact retention, strain relief and continuity checks. It records the adapter labels and pin assignments, while keeping unresolved display-side mating parts and physical dimensions clearly marked.

The short version: use the correct contacts for the housing and wire, crimp a sample first, check that it locks into the housing, and meter every completed lead before powering it. Match electrical nets and the actual mating view. Wire color and a similar-looking plug are not proof of pin order.

The display's USB supply must stay separate from SATA LED power. First bring-up will start with one stick at low brightness, then add the second. Power sequencing, data behavior and board temperature must pass before the loom is tucked behind the backbone. Keep cable slack out of the cooling corridor and leave enough room to remove the magnetic covers.

The display work carries on

The current Windows touch-button setup has passed user testing: all six existing application windows come forward without duplicates, and a closed OBS launch-and-focus test passed. The panel also retains live resource readings, its animated scenes and audio, display Sleep/Wake and scene-audio Mute/Unmute.

Linux remains the primary platform, initially Bazzite on the BC250. Both platforms share the companion code and panel protocol; Linux hardware, desktop-focus and Gaming Mode tests are still ahead. Physical RGB output, gamepad host wake and the voice-companion workflow remain unfinished. Read the display update.

The Ronin case study keeps the Voron-inspired commercial first, the calm cinematic film second, then the complete preliminary CFD animation. The old films and stills are still there. The new harness supplement will feed the next manual revision; the downloadable Rev0.3 PDF remains its original draft snapshot.

Thanks remain due to Blamm for Rainbow on a Matchstick, Mosfet.Party for the EPS Power Adapter and PS_ON Adapter, and MandicReally for the single remixed 92 mm fan shroud. Ronin is an independent project with no official Voron affiliation.