Building a Custom Car Computer in 2006: Honda Accord PC Integration Project

This is the story of a pioneering car computing project that predated Android Auto and Apple CarPlay by nearly a decade. What started as curiosity about integrating PC functionality into a factory car display became a journey through reverse engineering, custom hardware solutions, and software development that would make any modern car enthusiast envious.

The Vision

Back in 2006, smartphones were still in their infancy, the original iPhone wouldn't arrive until 2007, and most people were still carrying flip phones or early BlackBerry devices. But those early glimpses of mobile computing were enough to spark my imagination. Seeing the potential of having computing power in your pocket made me think: why couldn't I have that same kind of integrated computing experience in my cars?

Looking at my Honda Accord's factory navigation system with its crisp built-in display, I had an ambitious idea: what if I could somehow get my Dell PC's output to show on that screen instead of being limited to whatever the factory navigation system could do?

The Honda Accord came with a sophisticated satellite navigation system for its time, complete with a DVD drive and a quality display that was already integrated into the dashboard. Rather than replace this entire system with an aftermarket solution, I wanted to find a way to hijack the existing display and use it for computer output.

I should apologise in advance for the photo quality throughout this project documentation but remember, this was 2006, when phone cameras were still measured in potato-pixels and taking a decent photo required actual effort, lighting and luck. The grainy, blurry images you're about to see are an authentic representation of the state of mobile photography when flip phones ruled the world and the original iPhone was still a year away from existence.

The Detective Work

The first challenge was figuring out exactly what kind of video signal the Honda's display was expecting. This wasn't information you could just Google in 2006 and Honda certainly wasn't publishing technical specifications for hackers to use.

I decided to take the direct approach of reverse engineering. I carefully accessed the wiring between the DVD satellite navigation drive and the display screen. Using an oscilloscope, I looked for what appeared to be video signals and found four promising candidates. Now came the moment of truth; I cut these wires one by one to see what would happen. The first wire I cut caused the video to start scrolling vertically, bingo, jackpot! This indicated it was the sync signal. However, I had actually expected that losing sync would also cause me to lose a colour, but I didn't - this threw me off at the time.

The other three wires were the Red, Green, and Blue signals. Cutting each one in turn would change the screen colour, clearly indicating which colour signal each wire carried. Through this methodical process, I had successfully identified all four components of the RGBS signal format. My initial expectation had been to find a standard composite video signal, which is why discovering RGBS was so surprising and confusing at first.

With the signal format now understood, I could create the custom cable needed to inject my own video signal into the Honda's display system. You can see in the photos the cable I built with BNC connectors for each of the RGBS signals, this is the one key component that enabled this whole project, it allowed me to control the video source to the Sat Nav display. I was absolutely over the moon, the four wires that I had identified via probing with the oscilloscope were the exact ones I needed, though I was still surprised to discover a European video standard (RGBS is often referred to as EuroSCART) in a Japanese car.

The close-up photo below shows the actual video interception point behind the dashboard where I tapped into the Honda's internal video signals. You can see the tin foil (yep that is actual tin foil straight from the kitchen!) wrapped around the connections - my "advanced interference screening" solution to prevent signal degradation and interference from the car's electrical systems. While it might look improvised, electromagnetic shielding was actually crucial for maintaining clean video signals in the electrically noisy environment of a car's dashboard.

Technical Discovery: As mentioned above, after probing various wires and tracing signal paths, I made a crucial discovery. The Honda navigation system wasn't using standard composite video or S-video like many car displays of the era. Instead, it required RGBS - that's Red, Green, Blue, and Sync - a format that was more common in European markets and certain professional applications.

This was both good news and bad news. Good because RGBS could potentially deliver better image quality than composite video. Bad because it meant I needed a very specific type of converter, and options were extremely limited in 2006.

The Hardware Hunt

Finding a device that could convert VGA output from a PC to Honda Sat Navs RGBS format turned out to be a significant challenge. After extensive searching, I managed to locate a scan converter that could handle the job, but it was far from ideal. This converter was enormous, bulky, and seemed to consume an surprising amount of power for what should have been a relatively simple signal conversion task.

The device worked by taking the VGA signal from the Dell PC, which comprises of separate horizontal and vertical sync signals (RGBHV), and converting it to the RGBS format. While the converter was bulky and power-hungry, it was literally the only option I could find available at the time that could perform this specific conversion.

Looking back: How tech has moved on huh, today you can buy compact devices like the VGA2SCART converter that do the same job in a package smaller than a deck of cards and use a fraction of the power. But in 2006, if you wanted to convert VGA to RGBS, you had to make do with industrial-grade equipment that was designed for professional video applications, not car modifications.

The Power Problem

With the video signal mystery solved and the bulky scan converter doing its job, I thought the hardest part was behind me. I was wrong. The next challenge was something I hadn't fully anticipated was keeping a full blown PC running reliably in an automotive environment.

My solution for powering the Dell PC seemed straightforward enough - a 12V DC to 240V AC inverter that would convert the car's electrical system to standard UK household power that the PC expected. Initial bench tests were promising. Connect the inverter, plug in the PC, and everything worked perfectly. The computer booted up normally and ran without any issues while parked with the engine off.

The Problem: The moment I turned the key to start the engine, disaster would strike! The PC would instantly lose power and shut down abruptly, leaving Windows in that unhappy state that anyone who's experienced sudden power losses knows all too well.

The problem was fundamental to how car electrical systems work. When you crank the engine, the starter motor draws enormous amounts of current sometimes 200-300 amps or more. This massive power draw causes the car's electrical system voltage to drop significantly, often falling from the nominal 12V down to 8V or even lower during those crucial few seconds of engine start. My inverter simply couldn't maintain stable output when the input voltage collapsed like that.

Engineering the Tank Circuit Solution

The solution came from understanding how critical systems in vehicles handle this exact problem. I needed what's called a "tank circuit" essentially a backup power system that could maintain steady power to the PC during those brief moments when the main electrical system was compromised by engine starting. A mini UPS (Uninterruptible Power Supply) for the CarPC, if you like.

The tank circuit consisted of a small 12V lead acid battery that would automatically kick in whenever the main battery's voltage dropped below a certain threshold. During normal operation, this backup battery would stay fully charged from the car's alternator. But when the starter engaged and the main system voltage collapsed, the tank circuit would seamlessly take over, keeping the inverter and PC running smoothly through the engine start sequence.

Implementing this wasn't trivial, it required careful selection of the backup battery size, proper charging circuitry, and a voltage-sensing relay system to ensure the handoff between main power and backup power was transparent to the PC. The relay circuit would monitor the main system voltage and automatically switch to the tank circuit whenever it detected the voltage drop that occurred during engine cranking. Once the engine started and the alternator brought the main system voltage back to normal levels, the relay would switch back and the backup battery would begin recharging for the next start cycle.

The satisfaction of solving this power management challenge was immense, and I felt genuinely proud of applying my electronics knowledge so effectively. It actually made me reflect on my career path - had I chosen the right direction with IT, or should I have pursued electronic engineering instead? This project certainly revealed there's definitely an engineer inside of me, one who thrives on understanding systems at a fundamental level and creating elegant solutions to complex problems.

Once properly installed, the tank circuit completely eliminated the shutdown problem. The relief was immense, the PC and the screen would keep running happily regardless of what was happening with the engine, it was trully amazing to see the screen remain on during cranking (I don't even think modern cars do that). With this solution in place, the CarPC was finally, truly practical for everyday use

The Software Challenge

With the hardware finally working reliably, video converting properly and power staying on through engine starts I faced the next major hurdle: what software to actually run on this custom car computer system. Once again, This turned out to be more challenging than I'd anticipated. But hey, as you've gathered I thrive on challenges.

My first attempt was with open source software. While it technically worked, the user interface was rather clunky. When you're trying to operate something while driving (safely, of course), you need large buttons, clear graphics, and intuitive navigation things that desktop-oriented open source solutions weren't optimised for. It did however give me an idea of the potential of the system that I'd put in place.

Next up, I tried a commercial offering, CentraFuse 2.0, this cost me £60, a significant investment at the time. This had a much better user interface that was clearly designed for automotive applications, but it came with its own set of limitations. The functionality was restricted, and more importantly, there was very little ability to customise it to my specific needs and vision.

What I really wanted was something that could be fully customisable to create the ultimate in-car computing experience. I envisioned a system with an up-to-date satellite navigation system, a sophisticated media player, internet access capability, and even screen mirroring functionality. In fact, I was particularly excited about getting screen mirroring working with both Windows Mobile and Android devices essentially creating my own version of what would later become Android Auto and CarPlay, but years before those technologies existed.

When discussing my project at work, it piqued the interest of two of my colleagues. They could see the potential in what I was trying to achieve, and more importantly, together we had the programming skills to help make it happen.

We decided to write our own custom software from scratch using Delphi. We called it MediateICE, and it was designed specifically for in-car computing applications. Unlike the off-the-shelf solutions I'd tried, MediateICE was built around our exact requirements and could be continuously refined and expanded as we discovered new needs and possibilities.

Credit where credit is due, Andrew and Shaun undertook the bulk of the development work, leveraging their programming expertise to bring the vision to life. The software was their baby, I supplied a real world setup for testing and feedback.

The software represented the culmination of everything we'd learned about what worked and what didn't in car computing interfaces. You can see in the photo how MediateICE integrated seamlessly with the Honda's factory display, presenting a clean, automotive-focused interface that looked like it could have been factory Honda software. The main menu shows the various functions available navigation, multimedia, and system controls all designed with the large buttons and clear graphics necessary for safe in-car operation.

The background image in the interface shows a Honda Accord coupe a subtle nod to Honda's sportier offerings and perhaps a hint at automotive aspirations beyond the practical saloon. The 3.0-liter V6 Accord coupe represented the performance variant of Honda's midsize lineup, combining the reliability and comfort of the Accord platform with more power and style.

The User Input Challenge

With video output solved and the system running reliably, the next logical challenge was user input. How do you control a PC-based system while driving safely? The Honda's touchscreen was right there, but reverse engineering the touch interface proved to be a bridge too far.

I spent considerable time researching how I could intercept the Honda touchscreens signals, but unfortunately concluded that it would require serious amounts of dismantling to access the touch interface electronics. The touchscreen controller was deeply integrated into the head unit, and tapping into those signals would have meant essentially disassembling the entire Sat Nav unit. I had to swallow my pride and accept that this particular challenge was beyond the practical scope of the project.

The Touchscreen Dilemma: While I could intercept and replace the video signals relatively easily, the touchscreen input system was far more complex and inaccessible. Sometimes knowing when not to pursue a particular technical challenge is as important as solving the ones you can tackle.

Instead, I opted for a much more elegant solution: a Sony RM-X2S Car Audio Joystick Remote Control. This compact device was designed specifically for automotive use and could be easily mounted to the side of the steering wheel, within easy reach without having to let go of the wheel. The joystick provided around 20 different inputs including directional controls, multiple buttons, and even analog controls more than enough for comprehensive system control.

The Sony remote was connected to a Velleman K8055 USB Experiment Interface Board a versatile digital and analog I/O board that could convert the joystick's various button presses and analog movements into USB signals that the PC could understand. The K8055 was perfect for this application because it provided multiple digital inputs, analog inputs, and was designed for exactly this kind of custom interfacing work.

Smart Control Integration: MediateICE included sophisticated input mapping capabilities, allowing complete customization of what each button and joystick movement would do. You could map inputs to media controls (next song, previous song, volume), navigation functions (screen select, OK, back), or any other system function.

Physical Installation and Power Management

The physical installation was carefully planned to distribute components throughout the vehicle for optimal space utilization and thermal management. In the boot, I mounted the Dell small form factor PC alongside the power inverter, tank circuit, and Velleman interface board. Under the drivers seat, I positioned the HDMI to VGA converter and the large scan converter keeping these heat-generating components away from the main equipment bay and closer to the dashboard where the video signals needed to terminate.

The PC was configured to "go to sleep on losing 12V power signal" through a manual switch, providing a clean shutdown process that protected the Windows installation from corruption. This was crucial for system reliability, as abrupt power loss could damage the file system and require expensive repairs or complete reinstallation.

Battery Life Reality Check: With the ignition off, running the PC would drain the main car battery in about 45 minutes before requiring a jump start. This highlighted the substantial power requirements of the system and the importance of the manual shutdown switch.

To prevent accidentally leaving the CarPC running and finding myself with a dead battery, I implemented an ingenious solution borrowed from automotive safety systems. I installed a "headlights on" reminder circuit, but wired it to monitor the PC switch instead of the headlights. This circuit would beep if you turned off the car's ignition but had forgotten to switch off the PC - a simple but effective safeguard against costly mistakes.

Safety Considerations

Running 240V AC in a vehicle naturally raises safety concerns, so I thoroughly researched the potential risks and implemented appropriate safeguards. The key safety factor was Honda's crash detection system, which automatically disables the ignition circuit upon impact. This meant that in the event of an accident, power to the CarPC system would be immediately cut, eliminating any risk from the 240V AC supply.

Additionally, all the equipment in the boot - the Dell PC, power inverter, tank circuit, and associated components - were enclosed within a metal box for both safety and professional presentation. Unfortunately, I don't have photos of this final installation, as the images you see throughout this documentation are from the initial prototype stages. The completed system was far tidier and more professional looking than these early development photos suggest.

This attention to safety and proper installation was crucial for a system that would be used daily in a family vehicle. The combination of crash-activated power cutoff and proper enclosure of all high-voltage components ensured the CarPC installation met appropriate safety standards for automotive use.

The Perfect Platform

The 7th generation Honda Accord 2.4 proved to be the ideal foundation for this CarPC project. These cars struck a perfect balance as family vehicles and motorway cruisers, comfortable, quick (especially with certain modifications... but that's a story for another blog), equipped with all the modern conveniences, and made even better with the custom CarPC installation.

Between 2006 and 2013, I owned three different 7th gen Accords and installed this CarPC system into every single one of them.

A Unique Achievement: Looking back, this may well have been a completely unique project. While others were installing aftermarket head units or using basic PC-to-TV converters, the complete reverse engineering of Honda's RGBS video system, custom power management with tank circuits, and purpose-built MediateICE software created something truly one-of-a-kind.

The combination of technical challenges overcome from signal format discovery to automotive power management to custom software development created a system that was years ahead of its time. While modern cars now offer smartphone integration as standard, they still don't match the complete customization and control that this CarPC setup provided.

There's something deeply satisfying about a project that starts with a simple question ("Can I get my PC to display on this screen?") and evolves into a comprehensive automotive computing platform. The 7th gen Accord provided the perfect canvas for this kind of technical innovation, and the resulting system transformed these already excellent cars into something truly special.

Screen Mirroring Success

Beyond MediateICE, one of the most impressive hardware capabilities I achieved was screen mirroring from mobile devices. This wasn't actually part of the MediateICE software itself, it was more a hardware feature of my CarPC implementation that could display any video source through the same signal conversion chain.

The photo shows the system successfully mirroring an Android phone's display onto the Honda's navigation screen you can see both the car's display and the phone showing identical content at 17:47 on February 28th. This was years before Android Auto or Apple CarPlay existed, making it a truly pioneering implementation of smartphone-to-car connectivity.

The screen mirroring worked through the same ingenious daisy-chain of converters that I'd developed for the PC's VGA output, but adapted for mobile devices. The process went: MHL cable from phone to HDMI, then HDMI to VGA (using the black converter box visible in the boot photo), and finally VGA to RGBS through the large scan converter. While this multi-step conversion process might seem complex, it was the only way to bridge the gap between modern smartphone output and the Honda's proprietary display system.

MHL → HDMI → VGA → RGBS → Honda Display

If I did have a photo of the boot you would see how all the components were mounted and connected the Dell PC on the right, the various converters and cables, and you'd see the systematic approach to organizing what could have easily become a chaotic mess of wires and boxes. The black HDMI to VGA converter was a crucial link in the chain that made smartphone mirroring possible.

I had similar implementations working with Windows Mobile devices as well. In fact, I might have even had a virtual Windows Mobile setup running on the PC itself it's been nearly two decades now, so some of the details are getting fuzzy! But the key point was that the hardware setup I'd created was flexible enough to display any video source that could be converted through the signal chain, whether it was the PC running MediateICE, a connected smartphone.

This hardware-based approach to screen mirroring meant that drivers could access their phone's navigation apps, media players, and other applications directly through the car's integrated display, regardless of what the phone's operating system was. This created a user experience that wouldn't become mainstream until nearly a decade later when manufacturers finally caught up with similar built-in solutions.

Looking Back

This project was a fascinating personal journey that happened during an era when enthusiasts had to be part electrical engineer, part detective, and part mechanic to achieve what we now take for granted with Android Auto and CarPlay. The amount of reverse engineering, custom hardware sourcing, and creative problem-solving required was substantial, but that was also part of the appeal. More importantly, it allowed me to learn extensively and let out the engineer hidden within me.

The wire probing technique that revealed the RGBS signal requirement was particularly crucial without that detective work, the entire project would have been impossible. It's a reminder that sometimes the most important tool in a tech project isn't something you can buy, but rather the willingness to dig deep and understand exactly how existing systems work before you try to modify them.

The Legacy: Looking at today's connected cars with their seamless smartphone integration, it's easy to forget that someone had to pioneer these concepts first. This project, with its custom MediateICE software and innovative screen mirroring capabilities, was essentially a working prototype of what would eventually become standard automotive technology built nearly a decade before the industry caught up.

The technical challenges I overcame from video signal conversion to automotive power management to custom software development showcase the kind of systems thinking and creative problem-solving that drives innovation. While modern solutions are certainly more elegant and user-friendly, there's something deeply satisfying about having built and understood every aspect of the system from the ground up.

For anyone inspired to tackle similar projects today, the good news is that the hardware landscape has evolved dramatically. What once required industrial-grade scan converters and complex power management can now be achieved with compact, efficient components. But the core principle remains the same: understanding existing systems, identifying the gaps, and engineering creative solutions to bridge them.

Personal Achievement: This Honda Accord PC integration project created a working car computer system that was years ahead of its time on a personal level. While it didn't impact the broader automotive industry, the screen mirroring, custom interface design, and seamless integration approaches I developed independently paralleled concepts that wouldn't become mainstream until nearly a decade later. The system even included automatic file synchronisation - pull up outside the house, the CarPC would connect to the home WiFi and run folder sync on media folders, automatically updating the car's music and video libraries without any manual intervention. Most importantly, I thoroughly enjoyed building it and I thoroughly enjoyed using it daily for many, many years.

Take On The Challenge:

If there's one thing I'd want readers to take away from this project, it's this: never shy away from tasks that look impossible at the start. When I first looked at that Honda navigation screen and wondered if I could get my PC to display on it, the technical challenges seemed insurmountable. How do you figure out proprietary video signals? How do you keep a PC running through engine starts? How do you create automotive-grade software?

But every impossible task is just a series of smaller, solvable problems. The key is breaking down the challenge, tackling each piece methodically, and never being afraid to learn something completely new. I wasn't an automotive electrical engineer when I started this project, but I became one through necessity and curiosity.

The most rewarding projects are often the ones where you have no idea how you'll solve them when you begin. They force you to grow, to research, to experiment, and to think creatively. Whether it's reverse engineering video signals, designing power management circuits, or writing custom software, each challenge teaches you something valuable that you can apply to future projects.

So whatever technical challenge you're considering, whether it's modifying your car, building custom electronics, or creating software solutions to problems that bother you, my advice is simple: start. Begin with research, gather your tools, and tackle the first small piece. You'll be amazed what you can accomplish when you combine curiosity with persistence and a willingness to learn as you go.

The engineer inside of you is waiting to be unleashed. Sometimes all it takes is finding the right project to let it out. What project do you fancy taking on?