To view this content, you must be a member of Andi's Patreon at $1 or more
Already a qualifying Patreon member? Refresh to access this content.
I’ve always loved retro tech, but sometimes working with older video equipment can be… frustrating. That’s why the GBS Control project caught my attention—it takes an affordable video scaler and transforms it into a powerful tool for retro enthusiasts like me. So, when I decided to tackle a long-forgotten V9900 scaler from my project pile, I knew GBS Control was the way to go.
The GBS 8200, a budget-friendly scaler from the arcade and retro console world, is functional but far from perfect. Latency issues and subpar video quality made it less than ideal for anyone serious about retro displays. Enter GBS Control, an open-source firmware upgrade that changes everything. With just a microcontroller like the ESP8266 or ESP32, you can unlock low latency, improved color accuracy, scan lines, and even web-based remote control.
The V9900 isn’t officially supported on the GBS Control wiki, but it seemed close enough to the GBS 8200 to give it a shot. My VC9900 had been sitting around for ages after an initial failed attempt to use it, so I figured it was time to see if I could make it work. Spoiler: it did, and the results are amazing.
Base instructions GBS Control Wiki
First, I needed an ESP8266. I found one on eBay for an absurdly cheap £3, shipping included. When it arrived, it looked fine despite some shipping damage—nothing a little pin-straightening couldn’t fix.
NodeMCU ESP8266 UK, NodeMCU ESP8266 USA
The first step, according to the GBS Control wiki, was to bypass the RGB input potentiometers. While the instructions suggested a destructive approach, I opted for a cleaner solution by removing the pots and replacing them with simple links, dressed up with heat shrink tubing for a neater finish.
Next, I wired up the debug pin using Dupont cables, which are perfect for temporary connections. The debug signal is sourced from pin 30 of the V9900’s IC6 chip, the MTV230 microcontroller. Once wired, the ESP8266 takes over, unlocking full control of the scaler’s video processor.
Connecting the ESP8266 was straightforward using the diagrams and instructions from the GBS Control wiki. I made custom cables for power and data, skipping optional add-ons like clock generators and OLED displays for now. I wanted to confirm everything worked before adding complexity.
You will also need to close the jumper to put the MTV230 in it’s place, jumper is located next to the two headers to the side of IC6.
The colored blocks, boxed in pink, on the pin out diagram below reflect the wiring and colors I used in my build. For clarity I use White for Vin and Grey for ground connecting to the 5v header.
Installing the GBS Control software was simple as long as I followed the instructions carefully. With everything connected, I ran initial tests using my Amiga 500. Cycling through ROMs and booting up a few games showed promising results—beautiful, crisp visuals with none of the latency issues I’d experienced before.
I powered the board with a Zip-Drive 5v supply, but any regulated 5v supply with a Centre positive barrel jack should work fine.
While the setup worked, it wasn’t practical—just two boards loosely connected by wires. To fix this, I built a mounting board from leftover plastic from another project (thanks, A2000 undertray!). After marking, cutting, and drilling, I mounted the V9900 and ESP8266 securely.
The result? A compact, reliable solution that handles retro video signals beautifully and outputs them via HDMI. All for less than £25 ($30)—a fraction of the cost of high-end alternatives like the OSSC or RetroTINK.
The difference is night and day. My repair footage now looks professional, and retro games come to life on modern displays. This build is perfect for anyone looking to upgrade their retro experience on a budget.
I did this to install MorphOS on a PCIe G5 PowerMac
Click “Show older versions”
4.07 (Versions not in order)
I have a pre made disk image here Dos6.22 ATI.img
at the prompt type
cd ATI
then to check the contents type
dir
You should see ‘X1900GT.ROM’ and ‘ATIFLASH.EXE’ listed.
We need to know the adaptor number, type
atiflash -i
note the adapter number of the installed X1900 GT, which is identifiable by the ‘R580+’ tag in the middle column.
Usually, the adapter number to the far left of the ‘R580+’ will read ‘0’, provided the card was installed to the PC’s first (or only) PCIe x16 slot.
Note In the following steps it assumes the adaptor number to be ‘0’, replace the 0 with the correct adaptor number if it differs.
Optional If you want to backup the X1900’s original ROM to the current directory, type
atiflash -s 0 x19back.rom
To flash the card type
atiflash -p -f 0 x1900GT.rom
Your screen may flicker for a couple of seconds. Afterward, it should tell you that the flash was successful, at which point use the power down the machine.
Remove the newly-flashed X1900 GT from the PC, and install it into the G5.
If all went well, the Apple logo should come up, and you should now be in Mac OS X. Verify the GPU information via Graphics / Displays in System Profiler, and rejoice!
You have successfully flashed a graphics card completely on your own, and now possess a Radeon X1900 GT PPC Mac Edition.
If you’ve been following my journey, you might recall a video I did a while back where I converted an old Amiga external floppy drive into something cool and retro-inspired. If you missed that one, no worries—I’ve linked it below. We’re taking things a step further. I’m about to transform this incredibly grubby, stained Cumana drive into something truly special, and I’m super excited to share the process with you.
Ever since I started this project, I’ve had one thing in mind: creating a purple OLED screen. I know it sounds a bit ambitious, but that’s exactly what I’m aiming for. Imagine pairing that vibrant purple screen with the Charity Amiga’s purple case—how cool would that be? Today, we’re going to see if we can bring that vision to life.
Before we dive in, though, I have to issue a little disclaimer: please remember that I’m not exactly a professional. In fact, I’m just an enthusiast with a lot of curiosity and a willingness to experiment (sometimes wildly). So, if you decide to follow along, do so at your own risk!
Like with any good project, I had to start by cleaning everything. And I mean everything. The drive, the cables—nothing escaped my cleaning spree. I won’t bore you with the details of rubbing down cables, but trust me, it was necessary.
Now, let’s talk about the case. It started off in a horrible state—stained, yellowed, and just plain ugly. But with a little bit of elbow grease and a lot of black paint, it’s undergone a serious transformation. It’s now a sleek, glittery black, which I think gives it a much more retro and sophisticated look.
You might remember from the previous video (linked bove) that I built a black Gotek drive with a blue OLED display. It looked pretty cool, but I had my heart set on purple this time around. Finding a purple OLED display, however, turned out to be more of a challenge than I anticipated. After scouring the internet, I came up empty-handed—purple just isn’t a color that’s readily available in OLED displays. I found plenty of blue, orange, green, and white options, but purple? Not a chance.
But I wasn’t about to give up. Drawing on my days in theater and lighting design, I decided to try a little hack: using a color filter. By cutting some purple lighting gel to size and securing it with double-sided sticky tape, I managed to give the white OLED display a purple hue.
And guess what? It worked better than I expected! The purple actually looks really good. One trick I used to enhance the effect was turning the OLED contrast up to its maximum setting. You can adjust the contrast in the FF.CFG file, which you load onto a USB drive—just like any other configuration for FlashFloppy. If you’re interested in trying this yourself, I’ll link the GitHub page for FlashFloppy along with the specific configuration details you need.
Flash Floppy Config Wiki : https://github.com/keirf/flashfloppy/…
oled-contrast=255
To add a little extra flair, I decided to make the inside of the drive pink, giving it a neon accent that ties in nicely with the purple Amiga aesthetic. I’m really pleased with how it turned out—I think the combination of colors gives it a unique, retro-futuristic vibe that’s hard to resist.
I hope you’ll agree with me that the final result looks pretty amazing. This project was a lot of fun, and it’s always satisfying to see a rough, grimy piece of tech turned into something stylish and functional. And if you liked this project, why not give it a try yourself? Who knows, you might surprise yourself with what you can create!
In the world of retro computing, where nostalgic gamers reignite their passion, the Amiga stands as a titan of vintage technology. My latest project has taken me on an epic journey to create a ROM switcher for these beloved machines. This isn’t just an upgrade; it’s a labor of love to preserve and enhance a classic.
I set out with a simple goal: to build a straightforward, passive ROM switcher for the Amiga. The challenge arose from the fact that not all Amigas were made equal. The early revisions of these machines had a hardware bug where the ROM socket wasn’t correctly wired. Specifically, address line 17 was connected to Byte select, which wasn’t an issue with the small ROM sizes of Kickstart 1.3 but became a problem with later, larger ROMs.
My mission was clear: create a ROM switcher that is compatible with all 68k Amigas with OCS or ECS chipsets—A500s, 600s, 2000s, and CDTVs. I wanted to switch between at least two ROM images, and that’s where my prototype journey began.
I started with a 27C800 chip, allowing for two ROM images. My initial prototype worked functionally but was a bit of a hack job. I even considered adding LEDs to indicate the selected ROM, but that idea didn’t pan out. The key missing feature was compatibility with early revision Amigas, where the ROM socket was miswired. By correcting this in the switcher, I aimed to create a truly universal solution.
After proving the concept with my crude prototype, I moved on to a more polished design. The first iteration was functional but had some shortcomings, like the floating bite mode select line in early revision mode. This line needs to be high to enable the 16-bit word mode that Amiga runs by default. By adding a bodge wire to tie that pin directly to VCC, the switcher worked perfectly.
The second revision of the board addressed these issues and extended the functionality. It supported a 27C160 chip, allowing for up to four ROM images on a single switcher. These ROMs are relatively easy to come by, provided they are no slower than 100 ns. This version came very close to what I envisioned, but I still had a few tweaks to make.
For instance, the new revision couldn’t program ROM images through the switcher—a feature that existed in the first revision. This was because the byte select pin, now tied directly to VCC, couldn’t handle the programming voltage. However, I saw potential in this functionality. It would allow for easier ROM image management, eliminating the need to manually combine images or navigate the complexities of ROM writing software.
With the final version of the ROM switcher, I achieved my goals: it fixes the early revision Amiga issues, allows for easy writing of ROM images, and supports multiple ROMs. The switcher can hold four ROMs and even allows for writing new images individually, making it flexible and user-friendly.
In the end, the project was more than just about creating a piece of hardware. It was about preserving the legacy of the Amiga and making it more accessible for enthusiasts and collectors. Along the way, I also explored building an adapter for the 27C4096 chip, a more straightforward and logical design than the 27C400. This new adapter worked perfectly on the first try, a rare occurrence for me!
I’d like to give a special thanks to PCBWay for sponsoring this project. They offer PCB prototype fabrication, CNC machining, and 3D printing services, all of which were instrumental in bringing this project to life. Their support made it possible to turn my designs into reality.
If you’re interested in building your own ROM switcher or need other PCB services, check out PCBWay.com. And if you’re curious about the ROM switcher, it will be available on PCBWay’s shared projects.
I would also like to than Chris “Cathers” for providing the ROMs and the inspiration for the 27C4096 adaptor and donating an Amiga 1200 case to the channel 🙂
Stay tuned for more adventures in retro computing, and if you’re curious about my process or other projects, check out my other videos!
This was my journey creating a universal Amiga ROM switcher. It’s been a fascinating and challenging project, but seeing it work perfectly in the end made it all worthwhile. If you’re a fan of vintage technology or just love a good DIY project, I hope this inspires you to dive into the world of retro computing.
Part 1 is here -> Building a Frankenstein Amiga: A Journey of Restoration and Customisation
Welcome to the final part of my Amiga 500 Frankenstein build, which I’m doing for the More Fun Making It charity auction. This journey began with a couple of neglected and yellowed A500 cases. Well, to be precise, it involved parts from at least three different cases that had all seen better days. One of the bottom halves was particularly damaged, so I swapped it out for a better one. When this project is finished, it’s going to be more of a Cinderella transformation rather than an ugly sister. However, I must warn you, the final color might be a bit of a Marmite situation – you’ll either love it or hate it.
First things first, I had to remove all the badges from the case and give it a thorough clean. This thing needed to be absolutely pristine before I could change the color.
With the case cleaned, the first phase of the color transformation involved spraying it with a black undercoat. This black layer acts as a plastic primer and base coat for the wild color that’s coming next. I must admit, it did look pretty good in black, but black Amigas are quite common these days. We’re going for something a bit more unique here. I planned on adding another two, maybe three, coats of the crazy color, followed by some lacquer to ensure the finish is robust enough to withstand regular use.
After about two weeks of painting and letting it dry, I decided to take the case with me to RMC the Cave to get Lee’s opinion. Here’s a picture of me at the RMC Retro shop.
With the paint now well and truly dry, I could finally put the finishing touches on the case and place the Amiga 500 board in its new home. The next stop was kickstart, where Reese showed his skills by playing any Doom game on any platform, even if it’s not an Atari. I spent much of the day chatting with people and hanging out with other YouTubers in the community room.
At the end of the day, we had a small signing ceremony, which was a great way to wrap things up. Here’s a quick look at that moment.
You might think that’s the end of the journey, but when I got back a couple of weeks later, I needed some footage for the end of this video. When I tried to fire up the Amiga, it was dead. So, I immediately began troubleshooting. After pushing some chips and trying again with no luck, I knew it was time for the big guns. I started by eliminating the CPU and Gary, the usual culprits for a black screen on an Amiga 500. I then beeped out the entire CPU socket and found one pin that wasn’t connected correctly. Turning the board over, I checked the trace and found continuity. Confused, I turned the board back over and lost continuity.
Thinking I was just too tired, I called it a night. Day two brought no change. Looking at the underside of the board with magnification, I saw that the trace was slightly damaged. The flexing of the board when turning it upside down was enough to make contact. So, I put a piece of wire across the damaged trace, and it was solid again. The Amiga fired up like nothing ever happened.
With the board now fully functional, I tidied it up and added any finishing touches I thought necessary. Then, I buttoned it up, put it back in the box, and finished the video. This time, I made sure to test it both before and after putting the lid on.
I’ll be doing something else in a similar style for another charity auction soon, so why not subscribe to my YouTube channel and make sure you don’t miss that video when it comes out? It’s free and helps the channel.
Thanks for following along on this journey!
Cheers!
Today, I’m diving into another exciting project: creating a Frankenstein Amiga from various parts I have lying around. Most of these parts come from a job lot of leftover spares I found on eBay. You might be wondering why I’m doing this. Well, I’m challenging myself and also building an Amiga for a charity auction organized by a friendly man named Lee from the YouTube channel More Fun Making It.
There’s something a little strange going on with this Amiga 500 Revision 5. It seems that not all the RAM chips are managing to output TTL logic levels. Based on what I’m seeing on the scope diagram, it’s giving me random green and blue flashes. So, I’m going to pull all the RAM out and socket it. Having the RAM in sockets will be a much better solution for the new owner, especially if any of the RAM I’m installing goes bad in the future.
The first thing I desolder, ignoring my marks, is the resistor pack. Well, let’s just style that out and pretend no one noticed (lol). Now, all the RAM has been removed, before we can solder in any of the new sockets for the RAM, we need to tack back in those resistor packs—the ones I unsoldered for no reason. Not that I’m planning to put any of the original RAM back in, but let’s go ahead and test it to see if it shows up as bad. Amazingly, all of the RAM tests good. I suppose the heat from desoldering the chips might have revived it, or maybe the logic levels I saw are enough for the tester but not for the Amiga. Or perhaps the RAM isn’t the issue, and we’ve got something else causing the Amiga not to function.
Time to start putting the sockets in: 16 * 16 pin DIP sockets, that’s 256 solder points plus the resistor packs that I unsoldered for no reason. I tack in one socket and solder it up to make sure I’m happy with the temperatures before I go ahead and install all the rest. I tacked in the rest using the two voltage pins, 5 volts and ground. The followed up by soldering the rest of the pins and clean up the flux. That’s all the sockets installed. Now they just need chips.
I remembered someone saying they hated turn pin sockets because they’re hard to line up with the chip pins. So I thought I’d showcase how I do it. This is just how I do it; it’s not advice. I’m not a chip insertion instructor. If you don’t like it, don’t do it. I’m just saying it works for me. Basically I insert one row of pins and then gently apply pressure to the other side while running my tweezers across the chip pins this locates the pins in the socket, it usually takes a couple of strokes for all the pins to align then the chip just goes in without issue. This is similar mechanics to picking a lock…
All the RAM is installed and looking lovely. I took the liberty of fixing the ROM socket so it will take a 27C400 EPROM or a later mask ROM. This is another future-proofing step for the next owner should they want to upgrade to a later Kickstart version.
We have life! I’ve cleaned up the flux and covered the bodge wires from the ROM socket modification. We have a working board, but we need more than that to build a full Amiga. In the same job lot of parts, I have several bits of keyboards. This is where it gets even more Frankenstein. The keys aren’t laid out in the same way as my other 500s. I think this is from an earlier keyboard. The controller is from a Rev 6 Green Power Light. If you can properly identify the original revision any of these parts came from, then please comment on the video.
Hooking it all back up with the keyboard and making sure it’s happy before we swap out DiagROM for Kickstart and try booting the Amiga into Test Kit. After confirming all is well with the Amiga, the next step is to get it to boot into Workbench and see if we can load StarTracker from the external GoTek drive. The plan is to load a module and play it, which will give I/O, graphics, and RAM a fairly good workout.
Since we’ve opened Pandora’s box, I thourght why not see if we can get that 512K of RAM in the trapdoor to be chip RAM? Having a total of 1MB chip RAM opens more games and options for the next owner. We just need an 8372A Agnus some little modifications. JP2 needs to be swapped, we need to cut a trace from Gary to the trapdoor, and we need to cover pin 41 on Agnus to keep it in PAL and prevent it from switching to NTSC.
While you weren’t looking, I’ve sprinkled a little cosmetic customisation over this Amiga 500. All we need now is a case.
Part 2 coming soon!
Building this Frankenstein Amiga has been a labor of love, filled with moments of frustration and triumph. I hope it finds a good home at the charity auction and brings joy to its new owner!
It all began with the power supply. After my Amiga’s original power unit gave up, I decided not to settle for an expensive or hard-to-find replacement. Instead, I went for a modern solution—transplanting an ATX power supply. This not only gave my Amiga a new lease on life but also made it more robust and reliable.
The next step involved the heart of the Amiga’s timing mechanism—a tick generator. Using a simple 555 timer IC, I initially hacked together a basic circuit to fulfill this role. However, the DIY spirit in me wasn’t satisfied with just a makeshift part. I wanted something better, something that looked the part.
This led to the creation of “JazzTick” my reimagined version of the original 555 timer hack. This new circuit wasn’t about creating a perfect solution but rather a good enough one that was both simple to assemble and effective. The design was straightforward, incorporating multiple resistors to fine-tune the resistance, thus adjusting the frequency more precisely.
I experimented with various resistors, aiming for the optimal frequency. In the UK, where PAL systems like the Amiga operate at 50 Hz, getting this right was crucial. After some trial and error with different configurations, I settled on a combination that gave me just under 400-ohms of resistance, and a 49.6Hz signal.
A project like this could have been daunting, but thanks to PCBWay, it was a breeze. They offer PCB fabrication for as little as $5 and have a vast library of shared projects which is incredibly helpful. For those who are a bit shaky with a soldering iron, PCBWay also provides assembly services. Their CNC Machining and 3D printing services further enhance the possibilities for custom DIY projects.
With the new JazzTick board ready, it was time to swap out the old hack. Replacing it with the new, neatly designed board was satisfying. Not only did it fit perfectly, but it also stuck to the power supply almost like a tick—a neat and tidy installation that I was very proud of.
The best part? This setup is flexible enough to be adjusted for a 60 Hz NTSC signal, should the need arise. For now, my Amiga is running smoothly, and the tick generator is ticking away without a hitch.
I’ve taken on the challenging yet rewarding task of bringing an Amiga 2000 back to life. If you’ve been following along, you know the journey hasn’t been smooth. The last update left the Amiga half-assembled with a problematic CPU slot and an overheating hard drive. As if that wasn’t enough, the power supply decided to give up on me.
Upon deeper inspection, it was clear that more work was needed. The power supply was non-functional, leading me to replace a suspect capacitor, though these fixes didn’t solve the problem.
Delving deeper, I decided to replace the voltage comparator and the strobe controller—fortunately, these parts were inexpensive. I installed new sockets and integrated circuits, hoping this would fix the issue.
Despite all these efforts, the power supply still failed. This led me to completely recap it, which seemed promising until it catastrophically failed again after just 15 seconds. I realized that some parts were either unknown or impossible to find.
I obtained a COMPAQ HB 146 SNQ power supply that mirrored the Amiga 2000’s requirements, I decided to adapt an ATX power supply, ensuring it matched the original’s settings. This required making it a permanent 230-volt input and using the original Amiga 2000 switch to activate the ATX’s PS-ON signal, effectively integrating it with the original system.
Testing the new setup, I used an old SCSI hard drive as a load to ensure stable voltage outputs. The results were satisfactory, with the 5V line perfectly on target, although the 12V line was slightly low—a point some might contest.
With the power supply sorted, I turned my attention back to the Amiga itself. I cleaned up acid damage on the board, replaced the battery with a more reliable one, and swapped out the old hard drive bracket for a 3D printed back plate, enhancing the setup.
However, issues persisted. Testing revealed unexplained memory discrepancies, and further investigation showed a short across two address lines on the board—a likely artifact from replacing the CPU slot. After removing the offending sliver of metal, I restored proper functionality to the memory and Zorro boards.
The journey didn’t end there. The Amiga’s ZZ9000 card, used for flicker fixing, lacked a crucial tick signal. I cobbled together a 50 Hz generator using a 555 timer and various components, which not only worked but improved the ZZ9000’s output.
This project has been a testament to the challenges and triumphs of hardware restoration. It’s a continuous learning process, filled with setbacks and victories.