The Pantasy model
The starting point was the Pantasy 85005 Retro PC, which I received for Christmas.
MAKING OF / DEVELOPMENT
Developing BrickDOS came with quite a few challenges, but above all it was a lot of fun.
Even better, the result is not just a model or a technical prototype, but a system that actually works and can be used. What I like most is being able to take the original idea behind the Pantasy model one step further: the nostalgic look is now joined by an actual nostalgic gaming experience.
BrickDOS is part of my hobby and brings together several of my biggest interests: LEGO, electrical engineering, computer science and gaming. I am not pursuing any other purpose with it. The project exists to be fun – not to make money.
Most of the graphics and designs on this site are AI-generated works based on my own ideas and direction. Electrical engineering, computer science, quality management and, above all, creative tinkering are much more my area. I am happy to leave the artistic part to professionals – or machines. ;-)
BrickDOS is purely a hobby project and is not intended to make money. If you still enjoy the project, I am of course happy about every Chocomel or schnitzel someone buys me. ;-)
If you have questions about BrickDOS, the hardware, the software or the conversion, feel free to send me a message. Just do not expect miracles from my answers. ;-)
You can find more projects, information and other things from me on my personal website.
ORIGINS
The starting point was the Pantasy 85005 Retro PC, which I received for Christmas.
The model quickly inspired the idea of putting a real computer inside it and actually playing DOS games on it.
I needed a computer that was small, affordable and powerful enough. In the end, I chose a Raspberry Pi.
In parallel with the hardware build, the software foundation was created using Raspberry Pi OS, DOSBox-X, ScummVM, OpenBox and the later BrickDOS interface.
The games were first prepared on a desktop PC, configured individually and tested extensively – sometimes more extensively than planned.
At the same time, I started looking for a 4:3 display that would physically fit inside the monitor and work well with the Raspberry Pi.
The right AUO panel was eventually found through Alibaba. After that, there was only one thing to do: wait.
The display, controller, Raspberry Pi, audio, power supply and software were first tested together outside the case.
Resolution, boot sequence, audio, game configurations and mechanical integration were adjusted step by step.
Only once the system worked reliably was all of the hardware permanently installed inside the monitor case.
Then came the most important test of the entire project: power it on, choose a game and simply play.
LESSONS LEARNED
Speakers were not part of the original plan at all. During development, it became clear that a real retro PC somehow feels incomplete without its own sound.
Suitable speakers were found quickly – but the housings required another round in BrickLink Studio.
The display raised several questions at once: which interface works best with the Raspberry Pi, what size physically fits inside the monitor, and where can a suitable 4:3 panel still be found at all?
MIPI DSI would have been technically more elegant, but suitable panels were difficult to find. In the end, I went with an AUO LVDS panel and a separate HDMI-to-LVDS controller sourced through Alibaba.
The computer had to be small enough for the monitor case, affordable and at the same time fast enough for the planned emulation. The Raspberry Pi offered the best overall package.
In hindsight, a less powerful model would probably have been enough for DOS-only use.
Some difficulties were initially interpreted as Raspberry Pi performance problems. Switching from classic DOSBox to DOSBox-X showed, however, that some of the issues were not caused by the hardware at all.
DOSBox-X therefore became the better foundation not only because of the originally planned Windows 98 support, but also for normal DOS use.
The original plan was to use Windows 98 SE as the central visible interface and launch both DOS and Windows games from it. However, graphics emulation on the ARM system was not reliable enough.
The concept was therefore simplified to Raspberry Pi OS as an invisible foundation with a direct launch into the BrickDOS ASCII interface.
The display fits surprisingly well inside the Pantasy monitor and its 4:3 aspect ratio also matches the intended era visually. For longer gaming sessions, however, the usable screen area is simply small.
In addition, the rounded monitor housing covers parts of the display corners.
The Raspberry Pi system on its own initially seems relatively inexpensive. But the display, controller, power supply, adapters, audio, cooling, cables and additional LEGO parts quickly add up.
BrickDOS is therefore not a cheap way to play old games. There would have been much simpler solutions for that – just none that would have been as much fun to build.
None of the individual components is especially exotic on its own. The real work was combining the hardware, power supply, display, software, emulators, games and LEGO housing in such a way that they became one coherent system.
TODO / IMPROVEMENTS
BrickDOS works – and that is exactly why the things that could still be improved are now becoming obvious. This list is less a formal specification and more a collection of items that will probably make me pick up the screwdriver, soldering iron or text editor again at some point.
The externally accessible connections should be integrated into the monitor housing in a cleaner and less conspicuous way.
If possible, the image rotation currently handled in software should be moved to the display controller.
An accessible Ethernet port would make classic network games such as Duke Nukem 3D much more interesting.
The problems with Windows 98 game emulation should be investigated further. If a reproducible DOSBox-X issue turns up, contributing to the project would also be an option.
Menu heights, line breaks and the display of very long game titles can still be made more robust and consistent.
Parts of the simulated boot sequence are cut off by the rounded corners of the monitor housing and should therefore be repositioned.
In the long term, a system-wide solution would be useful that keeps important content away from the obscured edge areas without unnecessarily shrinking the entire image.
The wireless devices are convenient and avoid visible cables. A wired setup, however, would fit the retro feeling better.
Add new titles, refine problematic configurations and continue optimizing existing games during longer-term use.