Bench Talk

**By Joe Hijuelos — WS2L** Founder, RF Boneyard / Hijuelos Sound Works LLC **Project status: Active Research — October 2026** ## Why We're Doing This The Yaesu FTDX-9000 is one of the remarkable high-end amateur radio transceivers of its era. It's an impressive piece of engineering, and these radios deserve to remain operational for generations. But there's a problem. The FTDX-9000's optional Data Management Unit (DMU) depends on aging computer hardware, a CompactFlash operating-system card, and proprietary software. As these components age, preserving the DMU becomes increasingly important. What happens when the original CompactFlash card fails? What happens when the industrial computer stops working? Can the original operating system be preserved, recovered, and eventually operated on replacement hardware? Those are the questions we're investigating at RF Boneyard. This is not a commercial replacement product announcement. It's an independent preservation, restoration, and reverse-engineering research project using my own FTDX-9000D. ## Phase 1: Preserving the Original CompactFlash Our first priority was protecting the original software. Rather than experimenting directly on the radio's CompactFlash card, we created a sector-level backup using R-Drive Image and established separate working copies for analysis. This gives us an opportunity to investigate the original operating system without unnecessarily exposing the irreplaceable original media to risk. The goal is to preserve the software, partition structure, boot configuration, and operating environment—not merely copy a few application files. ## Phase 2: Discovering the Original Linux System Our investigation confirmed that the DMU software uses a Linux 2.4.27 kernel dating from 2005. We extracted and examined operating-system files, startup scripts, filesystem configuration, and application executables. One particularly interesting discovery was a startup script launching three programs: - `OpeningRogoExe` - `TimerHandlerExe` - `CtrlExe` These appear to participate in the DMU startup sequence, with `CtrlExe` a key candidate for the main control and graphical application. Understanding how these programs communicate with the radio, initialize the display, and depend on the original hardware is a central part of the project. ## Phase 3: Getting the Original DMU Software to Boot in VirtualBox One of our biggest milestones came when we successfully brought the original Yaesu Linux software into a VirtualBox environment. We observed the Yaesu startup display and progress sequence—an encouraging sign that the original software could execute outside its original computer. We have not yet restored the complete operational DMU graphical interface. The system still requires additional investigation into its display initialization, application startup, and hardware dependencies. Nevertheless, achieving the boot sequence was an important breakthrough. It demonstrated that at least part of the original DMU operating environment could run on virtualized hardware. ## Phase 4: Investigating Replacement Industrial Hardware We're also studying a NEXCOM EBC365 Rev B industrial motherboard equipped with a VIA C3 800 MHz processor and CompactFlash support. The objective is to determine whether hardware of this type can serve as a practical preservation or replacement platform for the original DMU computer. We're examining BIOS behavior, storage compatibility, display output, and the hardware interfaces needed to support the original software. This remains experimental. We have not established drop-in compatibility. ## Phase 5: Understanding the R-Drive Backup Format Another part of this project has involved analyzing our preserved CompactFlash backup. We determined that its `.rdr` archive contains Zstandard-compressed records and identified repeating 40-byte record headers, 256 KiB data blocks, and several record sequences. We then tested an independent Linux utility called RDRKit. Its diagnostic mode successfully identified the archive's initial record headers, matching our earlier findings. However, its current disk-object listing functionality does not support one of the archive's index encodings. That means we still have work to do before reconstructing a complete raw disk image from this particular backup. We are documenting these discoveries carefully and keeping all restoration experiments separate from the original media. ## The Bigger Goal: Preserving the DMU for the Future Ultimately, we'd like to explore several possibilities: 1. Preserve and document the original Yaesu DMU software and hardware. 2. Establish a repeatable recovery procedure for compatible original equipment. 3. Determine whether replacement industrial computers can operate the original software. 4. Investigate a modern Raspberry Pi-based graphical control system capable of reproducing useful DMU functions. 5. Share verified technical findings that may help other FTDX-9000 owners maintain their equipment. A Raspberry Pi solution would require considerably more than displaying a replacement screen. We would need to investigate the radio's communication protocols, control functions, and application behavior. We're approaching this as an engineering research effort, not assuming these challenges have already been solved. ## Why This Matters to RF Boneyard RF Boneyard was founded to help keep valuable electronics operating long after manufacturers discontinue the parts and assemblies needed to service them. Our interests extend beyond selling replacement components. We believe in preserving technical knowledge, supporting restoration work, and encouraging the repair of equipment that still has years of useful life. The FTDX-9000 DMU project is one example of that philosophy in practice. ## Calling All Yaesu FTDX-9000 Owners Do you own an FTDX-9000D, FTDX-9000MP, or another configuration using the DMU? Have you experienced DMU startup problems, CompactFlash failures, display issues, or difficulty finding replacement computer hardware? We're interested in hearing about your experiences. If you have service documentation, firsthand restoration experience, or knowledge of the DMU's internal architecture, you're welcome to join the discussion. We'll continue publishing findings as the research progresses, distinguishing confirmed results from experiments and open questions. **The mission is simple: keep these remarkable radios alive.** 73, **Joe Hijuelos — WS2L** Founder, RF Boneyard **rfboneyard.com** *Independent restoration and research project. Not affiliated with or endorsed by Yaesu Musen.*