genesis(ish) project I have a hardware / software project for this community which I will start as soon as I have enough free time Thats all im going to say for now :nod:
Something for the genesis developers Ive been doing a little research on "Play Tv" which is basically a genesis console on a chip containing a rom programmed with 6 games and can be picked up extremely cheaply What I plan to do make a small development pcb that attaches to the play tv pcb which will contain 128-256MByte DRAM, SD Card Socket for games / applications, Flash rom containing Bios and FAT filing system, COM Port for pc link, Improved audio output If I can find an RGB output on the Play TV then i'll also throw in a Jamma connector and Audio amplifier At the moment its just a few scribbles on paper but I hope to have some spare time within the next few months to get started on this I hope there will be some developers here interested in this gizmo
Why not just make it for normal consoles so anybody can use it? Also what good is 128MiB of RAM when the largest game is 5MiB after bankswitching and even 8MiB would be obscenely large, especially for homebrew? COM port as in RS232/USB? What for since you can already do printf debugging over controller/expansion port? Improved audio how? You realize that clones always have innately worse audio than real consoles right? In what will you implement the DRAM MUX and state machine, and the various save methods? I hope you don't think you'll use a MCU like the mapper thread MCU would be smart for SD reading speed, but of little any other use, therefore like PowerPak I'd recommend directly controlling the card to cut down on components. If you use the system's SMS mode, your chip count could be as little as 8-bit Flash BIOS, 8-bit SRAM, ~3000 gate ASIC, 16-bit DRAM. Also are you sure you can pull this off? I've heard countless people say they are going to develop copiers but few if any have gotten anywhere, they're always overoptimistic about their digital skills and usually can't get past the DRAM interfacing.
"Why not just make it for normal consoles so anybody can use it?" I wanted to make a tiny stand alone multigame system but I could always make a cart version too, thanks "Also what good is 128MiB of RAM when the largest game is 5MiB after bankswitching and even 8MiB would be obscenely large, especially for homebrew?" Yes its a little large but pc dram cards are cheap I wasnt thinking of using this just to run game copies, this is going to be a board for developing new software on so theres no reason why you cant go crazy with the graphics/sampled sound and fill the ram "COM port as in RS232/USB? What for since you can already do printf debugging over controller/expansion port?" As far as I can tell the Play tv only has a single game port so I was thinking of adding a second for the pc link but point taken I could use the single port it already has "Improved audio how? You realize that clones always have innately worse audio than real consoles right?" Yes the YM2612 emulation on this thing is a little "off" but not so bad, when I said improved audio what I meant was that the output mixing is a little clipped so I was going to rework the audio section, I cannot do anything about the emulation "In what will you implement the DRAM MUX and state machine" Generating the ras/cas is not a problem and can be generated with simple logic gates and dare I say it "old school r/c timing !", addess mux with simple data selectors or tri-state buffers controlled by the same logic thats generating the ras/cas timing "and the various save methods?" Well this system is for development really so your software can use the bios's fat filing system to save files to the sd card, im not interested in copied game saving "I'd recommend directly controlling the card to cut down on components" Yes this is exactly what I am going to do with the sd card, no need to over complicate things. I am trying to keep this thing extremely cheap and easy to construct "Also are you sure you can pull this off? I've heard countless people say they are going to develop copiers but few if any have gotten anywhere, they're always overoptimistic about their digital skills and usually can't get past the DRAM interfacing." Yep my skills are "good enough", only spare time is against me Thank you for your comments :thumbsup:
They're also really power hungry, you could probably get a single suitable SDRAM chip to do the job. Loading that much RAM serially off SD cards directly with the MD will take ages, under the spirit as (no) traditional save memory, how about having people load game data on demand and using a smaller (but still huge for MD) RAM? Yeesh, that's a lot of discrete logic, especially for addressing and bankswitching that much RAM. Using RC constants to generate the setup times really is crazy, too much fine tuning with a scope IMO... Good luck! Also are you sure anybody out there will develop for your system? Perhaps you should run it by the MD/Gen homebrew communities to see if people want what you've outlined. I know many homebrewers will only develop their games for common cartridge hardware to add a touch of authenticity and make their games theoretically less difficult to publish.
"They're also really power hungry, you could probably get a single suitable SDRAM chip to do the job." Sorry I used the term dram to decribe the type of ram (dynamic ram), A single 128MByte sdram (which is dynamic ram) is what I had in mind "Loading that much RAM serially off SD cards directly with the MD will take ages, under the spirit as (no) traditional save memory, how about having people load game data on demand and using a smaller (but still huge for MD) RAM?" Loading data on demand is the way I imagined the programs should be written, with this much ram the data can stay there once it has been loaded making the program run faster on subsequent runs or you can go crazy and fill it up with huge sample files and graphics in one load if you really like "Yeesh, that's a lot of discrete logic, especially for addressing and bankswitching that much RAM. Using RC constants to generate the setup times really is crazy, too much fine tuning with a scope IMO... Good luck!" Yes I do have a slight concern with the rc timing but it has worked well for me in the past on z80, 6502, 6800 homebrew systems, if it is a problem i'll use a purely digital solution "Also are you sure anybody out there will develop for your system? Perhaps you should run it by the MD/Gen homebrew communities to see if people want what you've outlined. I know many homebrewers will only develop their games for common cartridge hardware to add a touch of authenticity and make their games theoretically less difficult to publish." Thats a very good point, maybe I should make this a more compatible cartridge based board so theres a least a slim chance someone would want to program it, I will run it buy another forum as see how the feedback (flaming ?) goes Thanks again for your comments :thumbsup:
What is useless since there are better clones out there now, like the Gen Dual?, the GAME JOY, or the new MD which is in Play-Asia. As Calpis said, why do you want so much RAM? I think that 128 MBit is more than enough (wasting RAM actually). Why use SD? It's serial (aka, slow), and it has some strange propietary protocol, when CF is parallel and ATA. Again, as Calpis said, useless. And even, if you don't want to sacrifice one controller port, you even have a third port on the back for debugging, and you can use the two front ones for controllers. And again, as Calpis said, clone's sound is awful, metallic, some channels are distorted and some channels are louder than others, which are nearly muted, some things aren't emulated or aren't emulated properly, etc... It has no RGB. And, for JAMMA, I'd seriously rather use a real Genesis, having a clone in an arcade when you can get bare genesises for a teener it's a tad bit... Any ideas about how you're gonna code it / design everything? Are you gonna use a FPGA for controlling everything? Of course! :110:
"What is useless since there are better clones out there now, like the Gen Dual?, the GAME JOY, or the new MD which is in Play-Asia." Ok its not the best around but its small and very cheap in the UK (£6.95) "As Calpis said, why do you want so much RAM? I think that 128 MBit is more than enough (wasting RAM actually)." Already answered this question "Why use SD? It's serial (aka, slow), and it has some strange propietary protocol, when CF is parallel and ATA." They are not too slow for this application, reading 1Mbyte from the card in spi mode only takes around 2 seconds The protocol is no problem, I have already written initializer/read/write routines for other projects in C (pc) and ASM (pic microcontroller) which work fine "Again, as Calpis said, useless. And even, if you don't want to sacrifice one controller port, you even have a third port on the back for debugging, and you can use the two front ones for controllers." I think you are confusing yourself, my post is about interfacing the play tv which has only 1 port and this is why I am adding another for the pc link "And again, as Calpis said, clone's sound is awful, metallic, some channels are distorted and some channels are louder than others, which are nearly muted, some things aren't emulated or aren't emulated properly, etc..." I have also said that the sound is not perfect but good enough "It has no RGB. And, for JAMMA, I'd seriously rather use a real Genesis, having a clone in an arcade when you can get bare genesises for a teener it's a tad bit..." So no jamma or amplifier then, no biggy "Any ideas about how you're gonna code it / design everything? Are you gonna use a FPGA for controlling everything?" I will most likely code the bios in C and use ASM for the more time critical functions, the hardware will be using a few simple logic gates, buffers, decoders, latches An FPGA would use a smaller footprint and be more flexible but im trying to keep it simple for the constructor, small, cheap, no exotic parts A cartridge version would allow the same functionality for the genesis
I will do thanks, I just hope I can share it with you guys here p.s. spec is subject to change without warning
I just started a mod that turns my super famicom into a hot asian girl who wants me 24/7. I swear I'll complete it when I get the time
Tell me when your finished with that So I Buy it off of you.Ive been needing a replacement, My Girlfriend is driving me insane.
I don't want to discourage. I think projects are great. It's just I don't usually announce something until I'm completely certain I will finish it...and in a reasonable amount of time from the point of announcement. A lot of the time I announce a project after it's completed. This's because there will be things I'm doing where I'll not be sure if it will work out or not. So I figure there's no point in announcing that I'm working on something if I'm just going to hit a dead end.
