FPGA Based Videogame System

[Reaperman]

Perhaps Krikzz may add Everdrive support for RetroUSB unrom512 and Memblers GT_ROM (I've not upgraded my firmware in a while). Not sure if flash saving will work properly without a new hardware revision on the everdrive n8. For games that just keep high score tables it's not an issue, but those that need flash writes for game progression, might run into issues.

 

 

Yeah, there's apparently already 3rd party everdrive support for mapper 30 (here). Weirdly, I don't happen to have one of those (since powerpak and nt mini), and I do get retrousb wanting to protect their homebrew authors--but there's already a fairly large wave of freely distributed mapper 30 WIP's out there, and I expect demand for the mapper will only continue to grow. Supposedly, it's fairly easy for me to flash my own games from the nesmaker ide using the infinite lives device, but it's not easy to trade nesmaker projects with others so they can use the official flash solution to try out others' stuff. It's actually kind of odd how poorly the projects package up with that software. It needs a bit of cleaning up in terms of how it handles paths internally, and where it wants to look for assets.

 

The infinite lives device appears to be much more difficult to use outside of nesmaker. Proper documentation only seems to exist for the older flasher, it's command line only, and there's some kind of unsigned driver flap which I'm sure is causing a ton of problems for me.

Edited September 22, 2018 by Reaperman

[cybercylon]

Infinite lives has been discontinued. They are beta testing a new version... INLretro Dumper-Programmer. You can actually order one in it's beta format. I am holding off until it is more finalized and a gui is in place for using it. It has slots for Genesis, N64 and GB/GBA too, so it may be able to do much more once all is said and done.

 

Note: I have no idea if this newer version will help

Edited September 22, 2018 by cybercylon

[Reaperman]

Note: I have no idea if this newer version will help

 

The new one's the one I have. It comes with their little nesmaker kit, but with the non-nes slots unpopulated. Yeah, I probably shouldn't have bought it so early, but the nesmaker kit page doesn't have all the red texted beta warnings like the page for the dumper/flasher does. The old kazzo dumper they sold is a considerably better documented product with a nicer interface that of course isn't super compatible with my device.

Edited September 22, 2018 by Reaperman

[CZroe]

Don’t forget about the Universal Mega Dumper in development from Renee at dbelectronics.

[Jurai]

there's a 15% off ebay coupon PICKSOON going live tomorrow afternoon fyi, could be useful if thinking about getting an SD2SNES, I personally ordered from retro_saikou and had no issues (cart was a rev.H sd2snes), there are a couple cheaper listings but I can't speak to whether they are also official krikzz cards or clones out of china

[seastalker]

I still say Analogue's "potential existence" would be taken to the next level if they created an "arcade console": An FPGA machine that would accept custom cartridges that would contain both the actual (licensed) arcade ROM set of an arcade game, and the custom FPGA core required to run that ROM set. Then you'd have different levels of cartridges, from the "budget" ones like Pac-Man, Pepper II, Time Pilot, Centipede, etc., to the more expensive arcade games like Altered Beast, Final Fight, Street Fighter II, Out Run, R-Type, and plenty of others I'm not thinking of right now.

 

The base console would need a strong FPGA setup and a huge chunk of RAM, but I believe it's not outside the realm of technical feasibility. And all those arcade cartridges would become instant collectibles. Heck, they could even offer rotatable screens as a deluxe set, for when you want to play TATE arcade games with the proper screen orientation, or alternate controllers (like a trackball controller, or a paddle controller for playing Arkanoid, or a dual-joystick controller for playing Robotron 2084 and Space Dungeon).

 

Any way you look at it, you gotta admit there's a huge potential market right there.

 

I really like this idea and have thought about this too. Those thinking about an arcade cabinet may be intimidated by the perceived 'spaghetti' wires inside a case or learning about JAMMA and superguns. Aquiring, storing and swapping PCBs can really keep people from the enjoyment. A small, arcade-perfect FPGA console would attract a huge market. A Sega-16 cart, Neo Geo AES/MVS one, and Capcom CP systems would be crown jewel launch titles. An open-source dev cart could encourage new FPGA developers to bring a one-off favorite to the table, as well as develop and sell NEW arcade games or downloads. Instead of a PC running MAME in an arcade cabinet, the ability to mount the system with easy cart swap access, no exposed boards, modern day low power, neat wiring to an lcd or CGA converter- this would be a dream arcade for many. Maybe some company would manufacture a blank cabinet for us to customize the artwork or with convenient swappable control panels. If it REALLY took off, maybe a new arcade monitor company would form where the screen could rotate from horizontal to vertical.

 

Of course I have no EXPECTATIONS of all this happening, but I like putting good thoughts out into the world all the same.

[Pixelboy]

I really like this idea and have thought about this too. Those thinking about an arcade cabinet may be intimidated by the perceived 'spaghetti' wires inside a case or learning about JAMMA and superguns. Aquiring, storing and swapping PCBs can really keep people from the enjoyment. A small, arcade-perfect FPGA console would attract a huge market. A Sega-16 cart, Neo Geo AES/MVS one, and Capcom CP systems would be crown jewel launch titles. An open-source dev cart could encourage new FPGA developers to bring a one-off favorite to the table, as well as develop and sell NEW arcade games or downloads. Instead of a PC running MAME in an arcade cabinet, the ability to mount the system with easy cart swap access, no exposed boards, modern day low power, neat wiring to an lcd or CGA converter- this would be a dream arcade for many. Maybe some company would manufacture a blank cabinet for us to customize the artwork or with convenient swappable control panels. If it REALLY took off, maybe a new arcade monitor company would form where the screen could rotate from horizontal to vertical.

 

Of course I have no EXPECTATIONS of all this happening, but I like putting good thoughts out into the world all the same.

Glad to see someone thinking like I do.

 

The big hurdle here is of course a technical one. Not only is the idea of an FPGA reproducing something as advanced as Street Fighter II a pretty crazy idea (it's probably not doable with current affordable FPGAs), even the primitive arcade games like Galaga have crazy custom hardware running under the hood, and for those reasons programming an FPGA core is a daunting task for most arcade games. The open-source MAME source code can help to make sense of each individual hardware setup, but it doesn't help a whole lot when it comes to programming an FPGA core.

 

It's evidently easier and cheaper to create an emulation-based machine (like MAME with arcade ROMs installed) and the resulting game experience is mostly the same for the end user, although it's never quite as plug-and-play as a dedicated ROM cartridge. Doing the same thing via a cartridge-based FPGA solution is really a purist pipe dream, but on the other hand, any company that puts such a cartridge-based arcade console on the market is going to make waves. Owning actual arcade games on cartridges is an attractive idea, technical hurdles non-withstanding.

 

Perhaps the best approach would be to remove the FPGA component from the equation: Instead of having a custom FPGA core in every cartridge (together with the licensed arcade ROMs) an "executable program" would be included. This executable program would follow a format similar to a binary Java file, and would be designed to run on a virtual machine. So the game console, no matter what hardware and operating system it actually contains, would then copy the contents of the cartridge to RAM at boot, and then run the pre-installed virtual machine program, which would itself run the "executable program" portion of the cartridge.

 

The advantage here is that the cartridge software would not be coupled to the game console hardware: Any fast computer that is equipped with the proper cartridge port and can run the virtual machine application could play these proposed arcade cartridges. And the virtual machine could set itself up the same way that emulators do: An arcade ROM set requires one master CPU, two secondary CPUs and two different sound chips? No problem, the virtual console configures the required virtual components and the arcade ROM has no idea it's running in a virtual environment.

 

Another advantage is that the virtual machine could be upgraded for bug fixes and new features added to the binary language, such new features being used by the "executable program" of more "advanced" cartridges. The contents of the cartridges themselves would remain unchanged, and new versions of the virtual machine would always be backward-compatible with all released cartridges.

 

For gamers who want to play licensed arcade games but are not interested in collecting cartridges, the ("executable program" + arcade ROM set) combos could be released as DLC. The end goal would be to create a sort of software standard by which any arcade game could be encoded as an ("executable program" + arcade ROM set) combo, and it is that combo that would be legally licensed.

[Andromeda Stardust]

Glad to see someone thinking like I do.

 

The big hurdle here is of course a technical one. Not only is the idea of an FPGA reproducing something as advanced as Street Fighter II a pretty crazy idea (it's probably not doable with current affordable FPGAs), even the primitive arcade games like Galaga have crazy custom hardware running under the hood, and for those reasons programming an FPGA core is a daunting task for most arcade games. The open-source MAME source code can help to make sense of each individual hardware setup, but it doesn't help a whole lot when it comes to programming an FPGA core.

 

It's evidently easier and cheaper to create an emulation-based machine (like MAME with arcade ROMs installed) and the resulting game experience is mostly the same for the end user, although it's never quite as plug-and-play as a dedicated ROM cartridge. Doing the same thing via a cartridge-based FPGA solution is really a purist pipe dream, but on the other hand, any company that puts such a cartridge-based arcade console on the market is going to make waves. Owning actual arcade games on cartridges is an attractive idea, technical hurdles non-withstanding.

 

Perhaps the best approach would be to remove the FPGA component from the equation: Instead of having a custom FPGA core in every cartridge (together with the licensed arcade ROMs) an "executable program" would be included. This executable program would follow a format similar to a binary Java file, and would be designed to run on a virtual machine. So the game console, no matter what hardware and operating system it actually contains, would then copy the contents of the cartridge to RAM at boot, and then run the pre-installed virtual machine program, which would itself run the "executable program" portion of the cartridge.

 

The advantage here is that the cartridge software would not be coupled to the game console hardware: Any fast computer that is equipped with the proper cartridge port and can run the virtual machine application could play these proposed arcade cartridges. And the virtual machine could set itself up the same way that emulators do: An arcade ROM set requires one master CPU, two secondary CPUs and two different sound chips? No problem, the virtual console configures the required virtual components and the arcade ROM has no idea it's running in a virtual environment.

 

Another advantage is that the virtual machine could be upgraded for bug fixes and new features added to the binary language, such new features being used by the "executable program" of more "advanced" cartridges. The contents of the cartridges themselves would remain unchanged, and new versions of the virtual machine would always be backward-compatible with all released cartridges.

 

For gamers who want to play licensed arcade games but are not interested in collecting cartridges, the ("executable program" + arcade ROM set) combos could be released as DLC. The end goal would be to create a sort of software standard by which any arcade game could be encoded as an ("executable program" + arcade ROM set) combo, and it is that combo that would be legally licensed.

Main issue here is cost (in time investment) versus reward. A custom core has to be programmed for EVERY arcade game, as opposed to one core to play hundreds or thousands of games with home consoles.

[Jurai]

My problem with the idea of some kinda pay as you go arcade cartridge model is at that point most people will just justify building a mame cab unless costs were ridiculously low. I really wish that SNK had done the Neo Geo Mini as a full blown bartop mini cab instead of the dinky thing they ended up with. Thought about maybe building a lil bartop picade or something but haven't spent time researching yuet

[vanfanel]

We're not far away from what you're saying. For only 50 Euros you can have a ZX-UNO, which is a small FPGA board supporting A LOT of arcade games right now:

 

http://www.zxuno.com/forum/viewtopic.php?f=60&t=1870

 

It's not CPS-1 level hardware, that woudn't fit the FPGA in that board, but you talked about Galaga... well, I have Galaga in FPGA using the ZX-UNO and it works great.

 

The same crew has just released the ZX-DOS, but we're short on cores right now.

Edited September 27, 2018 by vanfanel

[Andromeda Stardust]

We're not far away from what you're saying. For only 50 Euros you can have a ZX-UNO, which is a small FPGA board supporting A LOT of arcade games right now:

 

http://www.zxuno.com/forum/viewtopic.php?f=60&t=1870

 

It's not CPS-1 level hardware, that woudn't fit the FPGA in that board, but you talked about Galaga... well, I have Galaga in FPGA using the ZX-UNO and it works great.

 

The same crew has just released the ZX-DOS, but we're short on cores right now.

Dumb question. I noticed you have a scart plug attached. Will this device work on ntsc tvs?

 

My tv sets (tubes and flat panels) only support 60hz ntsc composite, 60hz ntsc rf, and 24/50/60hz over hdmi. I don't have any device capable of receiving pal formatted analog video, or any 240p rgb signal, scart or otherwise.

[Newsdee]

There are a bunch of arcade cores on the MiST and MiSTer FPGAs, but don't expect CPS1 just yet. They do have some cool stuff like Defender though.

 

You can find the core list on their wiki page:

https://github.com/MiSTer-devel/Main_MiSTer/wiki

 

It supports HDMI out of the box and optionally RGB/VGA out via an expansion board. You only really need the SDRAM board on top of the basic DE10 Nano hardware (easy to find, many people build and sell them).

Edited September 28, 2018 by Newsdee

[seastalker]

@Kosmic Stardust, I can't speak for the ZX-DOS, but I have and adore my MiST FPGA which also has a nice handful of arcade games. I have CRT computer monitors, but as most of the cores are for VERTICAL alignment, I love my 2007 Dell monitor that ROTATES. It will all depend on what you DO have for a screen or are willing to get in the near future. I do get both PAL and NTSC signals working fine on my setups, are there are scart->hdmi converter boxes out there. I'll have to check dual signal compatibility and 240p on each of my screens to see which is working best.

[RetrogamerX]

Is Analogue going to announce something soon? This seems to be the right time to announce a new product.

 

Don't let us down Kevtris

[NE146]

They did finally announce something about Super NT shipping at least

 

https://support.analogue.co/hc/en-us/articles/360000634171-Super-Nt-Now-Shipping-9-28-18-Update

[RetrogamerX]

They did finally announce something about Super NT shipping at least

 

https://support.analogue.co/hc/en-us/articles/360000634171-Super-Nt-Now-Shipping-9-28-18-Update

 

Well, we all knew that already. Hope we hear something in oct after all the Super NT preorders ship out.

[Andromeda Stardust]

There are a bunch of arcade cores on the MiST and MiSTer FPGAs, but don't expect CPS1 just yet. They do have some cool stuff like Defender though.

 

You can find the core list on their wiki page:

https://github.com/MiSTer-devel/Main_MiSTer/wiki

 

It supports HDMI out of the box and optionally RGB/VGA out via an expansion board. You only really need the SDRAM board on top of the basic DE10 Nano hardware (easy to find, many people build and sell them).

Yeah, I could buy a cheap used 1280x1024 monitor and mount it vertically no problem there. Even drill holes in a plywood panel for the Vesa mounts. But I'm really not interested if it don't support 480p minimum out of the box. 480p over vga would be acceptable for monitor inputs. Cheap upscalers tend to be laggy, and quality upscalers are very expensive.

[SegaSnatcher]

Glad to see someone thinking like I do.

 

The big hurdle here is of course a technical one. Not only is the idea of an FPGA reproducing something as advanced as Street Fighter II a pretty crazy idea (it's probably not doable with current affordable FPGAs), even the primitive arcade games like Galaga have crazy custom hardware running under the hood, and for those reasons programming an FPGA core is a daunting task for most arcade games. The open-source MAME source code can help to make sense of each individual hardware setup, but it doesn't help a whole lot when it comes to programming an FPGA core.

 

How is Street Fighter 2 too advanced for FPGA lol? Nothing about the specs below screams 100K+ LE FPGA needed. Looks pretty similar to Neo-Geo specs.

 

• CPU:
  • Primary: Motorola 68000 @ 10 MHz (some later boards 12 MHz)
  • Secondary: Zilog Z80 @ 3.579 MHz
• Sound chips:
  • Yamaha YM2151 @ 3.579 MHz
  • Oki OKI6295 @ 1 MHz (7.576 kHz samples)
• Display
  • Resolution: Raster, 384x224 @ 59.6294 Hz
  • Color depth: 16-bit (12-bit RGB with 4-bit brightness value)
  • Colors available: 65,536^[1]
  • Onscreen colors: 4096^[1] (192 global palettes with 16 colors each)
• Sprites:
  • Simultaneously displayable: 256 (per scanlines)
  • Sizes: 16x16, max. 16 colors (15 unique + 1 transparent)
  • Vertical and horizontal flipping capability
• Tiles: Sizes 8x8, 16x16, 32x32 with 16 colors (15 unique + 1 transparent)
• Tile maps: 3 maps, 512x512, 1024x1024, 2048x2048 pixel
• 68K RAM: 64 KB WORK RAM + 192 KB VRAM (Shadow)
• PPU: 192 KB VRAM + 16 KB CACHE RAM
• Z80 RAM: 2 KB WORK RAM

Edited September 28, 2018 by SegaSnatcher

[SegaSnatcher]

Is Analogue going to announce something soon? This seems to be the right time to announce a new product.

 

Don't let us down Kevtris

Next month will be to celebrate a certain console's 30th anniversary.

[ApolloBoy]

They did finally announce something about Super NT shipping at least

 

https://support.analogue.co/hc/en-us/articles/360000634171-Super-Nt-Now-Shipping-9-28-18-Update

FINALLY, goddamn. My family preordered one for me as a birthday gift and I've been waiting to see if they were gonna post anything (other than pointless game pictures on social media).

[Keatah]

The big hurdle here is of course a technical one. Not only is the idea of an FPGA reproducing something as advanced as Street Fighter II a pretty crazy idea (it's probably not doable with current affordable FPGAs), even the primitive arcade games like Galaga have crazy custom hardware running under the hood, and for those reasons programming an FPGA core is a daunting task for most arcade games. The open-source MAME source code can help to make sense of each individual hardware setup, but it doesn't help a whole lot when it comes to programming an FPGA core.

 

Galaga was done in FPGA years ago.

http://www.fpgaarcade.com/kb/galaga/

 

It isn't the crazy custom configurations that make things difficult in translation to FPGA, but instead schematic availability and custom chip knowledge and amount of circuitry.

[spoonman]

I'd be happy with an announcement of an FPGA based Neo•Geo since mine is starting to have issues with its cartridge connector.

 

I'd also really like to see the "Super NT scanlines overhaul" that Kev was mentioning a few months ago.

I always leave scanlines on for 240p content, but they look too squared off here. Even the scanlines on the AVS are a bit more believable than on the Super NT.

 

I wish we could get the core to the Super Game Boy released on the SNT. I'm not sure why it is, but I notice some input latency when using the SGB on the NT. And maybe the Colecovision (because it's my favorite console)

[Andromeda Stardust]

Next month will be to celebrate a certain console's 30th anniversary.

hmmm? i wonder what console they are talking about? Could it be???

 

Release date

JP: October 29, 1988

NA: August 14, 1989

KOR: August 1990

PAL: September 1990

BRA: September 1, 1990

 

 

[Andromeda Stardust]

I'd be happy with an announcement of an FPGA based Neo•Geo since mine is starting to have issues with its cartridge connector.

 

I'd also really like to see the "Super NT scanlines overhaul" that Kev was mentioning a few months ago.

I always leave scanlines on for 240p content, but they look too squared off here. Even the scanlines on the AVS are a bit more believable than on the Super NT.

 

I wish we could get the core to the Super Game Boy released on the SNT. I'm not sure why it is, but I notice some input latency when using the SGB on the NT. And maybe the Colecovision (because it's my favorite console)

It's going to be Nintendo's old 16-bit rival, not Neo Geo, not Turografx. I wonder what it could be? Stay tuned for an October 29th announcement? I think Snatcher may be onto something!

 

Are you running the Super NT in 720p? AVS is a 720p console. To me, 5x scanlines 1200p cropped to 1080p is too razor thin, but 3x 720p is perfect.

 

SGB is perfect on my TCL TV. Could be your TV. They aren't going to get the SBG core running as you'll need an entire Game Boy. SGB is cheap and plentiful.

[SegaSnatcher]

I'd be happy with an announcement of an FPGA based Neo•Geo since mine is starting to have issues with its cartridge connector.

 

I'd also really like to see the "Super NT scanlines overhaul" that Kev was mentioning a few months ago.

I always leave scanlines on for 240p content, but they look too squared off here. Even the scanlines on the AVS are a bit more believable than on the Super NT.

 

I wish we could get the core to the Super Game Boy released on the SNT. I'm not sure why it is, but I notice some input latency when using the SGB on the NT. And maybe the Colecovision (because it's my favorite console)

Cartridge pins might have cold solder joints if you already tried cleaning.