Memblers

There's on a few days left on this, sorry this post was a bit late. Check it out if you like. (spammy disclosure: I'm supplying the boards for this release, so I stand to make a couple bucks (literally) from the cart sales.) It's pretty neat, nice looking, the music is fun. It's made by Optomon, who's well known for some cool Castlevania hacks, but this is completely a homebrew. Clearly a very talented developer to have pulled all this off on their own, the only other credit is to the artist for the cover art and manual illustrations. The NSF is on the Kickstarter if you want to hear the music.

I backed at the $60 tier, but if there were a cheaper cart-only option, I also would have gone with that instead. I just like the carts more than the boxes, manuals, and all that. In a way it makes the release seem a little more classic-style, it's not like you could ever really just buy brand new, loose carts anywhere. It seems like KHan has released NES games in a lot of different ways, so it's always interesting to see how it turns out. There was one (NES port of Atari 2600 ET) that was only available through a mail-in order form that was included in the box with another of his releases.

3 days left on this. I wanted to point out also, if anyone was planning to try getting a cart after the Kickstarter, it would be best to order now. With this game, I expect the non-KS carts will sell out quickly, and like the way a lot of homebrew stuff goes, soon after that after that you can usually only buy it second-hand.

https://www.kickstarter.com/projects/nescape/nescape-an-escape-room-game-for-the-original-nes/ NEScape! is a new NES game being released soon by KHAN Games. It's fully developed, Kickstarter is being used to raise funds for a CIB physical release. You can also purchase the ROM for $10. Despite it using a relatively unique hardware, the ROM can be played on PowerPak and Everdrive flash cartridges, as well as recent emulators such as Mesen and FCEUX (interim builds since 2016 or so, including RetroArch). Note for full disclosure, I'm supplying the boards for this project, so I stand to make a couple bucks (literally) per cartridge order. The kickstarter page talks a little about the GTROM board the game uses. It's a particularly good board for those who are self-publishing. for it being low cost and coming preassembled and tested (512kB FlashROM, 32kB RAM). The kickstarter has less than 2 weeks remaining.

The Hi-Def and AVS are both great options. Just FYI, the jailbars are an analog effect, presumably caused by electrical interference getting into the power supply and/or video signal. It can't possibly exist in the digital HDMI output. So the countryside is safe, for now.

I talked to Rob (the developer) and it sounds like it should work on the NES and SNES classic. When you beat a boss, it will ask if you want to save, you will need to select no. It uses flash memory for saving, so if it's not emulated then mostly likely nothing will happen (but maybe it could lock up, so better to select no). I have an NES classic but haven't tried modding it yet. If it's running FCEUX, then that does support flash saving.

Candelabra: Estoscerro by Slydog Studios is 100% complete, and this Kickstarter campaign ends in just 3 days..! Check it out: https://www.kickstarter.com/projects/1545426105/candelabra-estoscerro-a-new-game-for-the-nes/ You can here the full soundtrack here (open in youtube if you want to skip around in the playlist): Here also is a short video review: If a dungeon crawler isn't your cup of tea, check out Rob's other games here (full versions for free): https://slydogstudios.org/

There was an early prototype of the 7800 that was done on single-sided PCB (i.e., traces only on one side.. and an insane amount of wire jumpers). I wonder if that was where "not enough room for a soundchip" started. On the actual 2-layer version, I think they could have made room for another chip. It seems crazy to put a sound chip into every cartridge. I wonder if they would have considered integrating the audio, mapper, and game ROM into the same chip. Possible, but I don't know at what quantity that becomes practical. I haven't programmed anything for the 7800, but have looked into it. I know the NES inside and out. On a technical level, the NES just stomps it, mostly due to the PPU being generally more autonomous and having it's own memory bus brought to the cartridge. While on the 7800, the graphics are more software-controlled, it has some flexibility but it comes at a high cost of CPU time. Pretty much the 7800 can display things that would be really hard to do on NES, but overall it can't keep up with the NES. If you could put a CPU in a 7800 cartridge that runs the game, while the 6502 handles MARIA and other I/O, then the 7800 would definitely be more competitive with the NES. On NES, mappers and the various MMC chips can add a lot of capability, but it's something like >60% of the NES library where the mapper is dead simple, they only bankswitch 16kB or 32kB program pages, and big 4kB pages (that's the full set of 256 tiles), if it pages graphics at all. That includes tons of games like Megaman 1&2, Castlevania, Contra, Zelda 1&2, etc. Would be fun to make a new game and/or cartridge for 7800, I just don't have enough time for new projects.

Anything is possible. Depending on the game, the difficulty would range from extraordinarily difficult, to very easy (if you have the skills/experience and the target program is hospitable). 2 major factors to consider: CPU time - does the game have enough idle time that you can take over? This varies during gameplay, and you have to fit inside the very worst case scenario (and your music program has it's own worst case timings to consider). Still fits in the "possible" category because, in theory, one could rewrite and optimize the game. This would be ridiculous though (unless one greatly enjoys RE'ing software). If there's not enough CPU time, better to just walk away. Or add your own CPU to the cartridge.. Memory - where in memory will the music data and variables sit? This actually may be kind of easy to overcome, as long there is a board with larger ROM/RAM you could migrate it to. If the game's source code was available that would be helpful to refer to, but overall it's of limited use. Assembling 30+ year old source code, or more likely converting it to a modern assembler, may be a bit of a project in itself. But I've never tried to assemble any of the old Atari source code, so I could be wrong. It's amazing that any of the source codes are even available at all. I wouldn't assume that they are the final versions, until you can assemble and do a binary compare.

Colecovision was my first game system, then we got the Atari expansion for it. Then later I was given a 7800, and that was my main Atari machine. Had maybe 15 games for it, but most of the other kids in the area had one with different games, too. I remember being really excited when we got a new TV, because I was able to move the old one into my bedroom, and soon I grabbed a butter-knife (as a screwdriver) and figured out how to hook it up. Maybe it was xmas 1990 when my grandma gave me a huge box that turned out to be filled with 150+ 2600 and 7800 games, bought the collection from one my relatives. I got a lot of use out of those. It's kinda funny that the very common CX40 joystick mostly reminds me of playing Atari at other people's houses when I was a kid, because I didn't have a CX40 or an original VCS until the mid-90s. By that time, I was more often using a Sega controller.

There is another way, excuse me if sounds a little crazy (because it is). It wouldn't require a modification to the system, however you couldn't use the RF output for audio to your TV (you'd need a separate audio amp/speaker or use an A/V mod to the system). One could make a pass-through cartridge, and plug your normal Atari game into the other end of it. This cartridge would have it's own audio output, and the audio synthesizer of your choice. The cartridge would spy on the CPU bus (let's say with a small-ish FPGA), and react to TIA writes. You would have a table of every possible combination of writes the game could make to the TIA, and use this to handle every combination of output you want to replace this with. With this, I imagine maybe just making a new pitch table so you have the same LFSR synthesis, but with the tones being closer to musically in-tune frequencies. This is a lot of work, for minimal improvement. If you change it too much, what would sound good on one game, probably will sound terrible on another. This meets the goal of requiring no modification to the games, but it is impossible to just look at the TIA writes alone and know the intent of the sound. To go further, the cartridge needs more intelligence. Inevitably, this becomes a thing where you'd have to make game-specific configurations. At this point you might as well also patch the ROM as a Game Genie does, then you can mute the TIA audio and not need the separate audio output anymore. I'm not saying this would be easy to make, just possible. I didn't exactly come up with this just for this thread, I'm actually designing hardware that will have these kind of capabilities.. Not for the 7800 though (and I didn't have this exact type of audio patching in mind), but it's definitely an interesting system that could use a soundchip. I feel the same way about the TIA like you said in your first post, I love it in the 2600, but to me it's just like 7800 doesn't even have a sound chip. I've seen the thread about the HOKEY project, I don't know what chip it is exactly but it's probably similar in capability to what I'm working with. Other than all that crazy hardware, I am working on my own game-music-inspired synth chip (which includes TIA and Lynx-type sounds), but now that I'm seeing that 7800 homebrew actually exists, I wonder if there would be any interest in bringing it to the Atari? But I suppose if the HOKEY turns out well, there won't be much room for another audio expansion.

What I understand of the built-in libraries of CC65 (which is not a lot, I use CA65 all the time but CC65 very rarely), they are more for cross-platform I/O. If you're going platform-specific, you can just ditch it all entirely and be just fine. I think you just have to make your own linker config, beyond that it's ready for whatever code you want. Though I haven't looked at newer versions, in the past it was the same with NES.. it supported it as a target platform, but no one actually used the built-in NES libraries because they were broken (were OK on inaccurate emulators, but not on the hardware). I could point you to some NES-related resources if that would help as an example, you can just just throw away the I/O routines or your replace them with your own.

One of my older plans, that I didn't carry out (because I've radically changed my design plan so many times..), was to use Bud Industries HP-3650 or similar-style enclosure. They have a replaceable panel on the end, one could design a new panel for it and produce it in acrylic with a laser cutter. Or drill out the plastic, easy enough if you just need a round hole. I started looking at some Chinese enclosure suppliers, so far I've only ordered from SZOMK and the prices for what I'm using (USB enclosures and another small type) have been much cheaper than stuff on Digikey and Mouser. They offer cutouts, drilling, silkscreen, and everything but I haven't had to ask about the pricing on those yet.

In my own experience, I find soldering SMT parts to be easier and faster than through-hole. Especially in the case of resistors and caps, you don't need to keep flipping the board over and trimming off leads. Of course I'm talking about SOIC, 0805, and 0603, going smaller than that it starts getting more difficult to hand-solder, and the solder paste stencil and toaster oven method becomes more attractive. SMT boards can be easier to design, when you've got the bottom of the board completely available for routing traces. For the boards, quality and price from China is hard to beat locally. I use MyroPCB and PCBWay, with both of those you can enter your data on their site and get bare board and assembly quotes automatically. You definitely will want SMT if you go that route. The parts themselves will be cheaper, and the assembly will be cheaper. For assembly you can roughly expect something like $0.015 per pin SMT, maybe $0.15 per pin for through-hole. These companies can buy the parts for you, IME this adds like 10% to the parts cost (for their fee and buying some spares). There are some one-time costs, like paying them to make the stencil, and the value really depends on how many of them you want to make. 20 or 50, hand soldering isn't too bad. 100+, outsourcing it probably starts to look like a better deal.

Any luck with this yet? Hi-Def NES is awesome. I don't know if this info is online yet (check game-tech's site), but I was talking to kevtris and I think he mentioned this heat issue showing up in a few cases. The fix was something like replacing a ferrite with a 0.1 Ohm resistor, and this discharges some kind of unexpected build-up on the regulator's input. This is just going by memory of something mentioned in passing in conversation, so please confirm this before modding anything.

It was kinda surreal to see that picture here, because that's my workplace. So I know what every little thing is in that picture. The thing you're seeing must be the stereo microscope, it's an awesome tool for inspecting fine pitch solder joints. It can make 0402 SMT part look like a skyscraper. Yes, I work with kevtris and see him every day, and no, he doesn't have his own massive lab with a 20 foot long workbench. But when off the clock, we do get to use all that good stuff for our own projects, some of which kev has built himself, like the reflow oven in that pic for assembling prototypes. That is really annoying how Mike K is still going around saying kevtris "threw them under the bus", what a load of crap. I'm sure he wouldn't mind me saying that we've talked about that project quite often, because we both find it interesting. He's done anything he could have done to help, short of designing the thing for them, and he's just had a realistic outlook on the chances of it succeeding based on the cost going up, amorphous specs, and bizarre marketing (no updates allowed?). And these have been almost everyone else's complaint as well. But I guess that guy is pissed that kevtris dared to speak realistically about it and show his own projects. Which is totally on-topic in this discussion, seeing as how the system was promoted everywhere as being an FPGA console running NES/etc. games. And then going on to say he's launching a competing product, while the rest of the team seems to be emphasizing that no, RVGS is actually for new games. It's completely different..! A few people are saying they want the Z3K instead, probably because it does what they were incorrectly led to believe the RVGS would do. I'll have to listen to the whole interview later, but it seems like John and Steve really know their stuff and are making something great. But it seems like almost no one thinks this fundraising campaign is going to work, there's just not any traction, hopefully they can continue to work towards something that will be funded. If I had a system where I could write in 6502, or C, and make my own hardware add-ons in Verilog, I would really like that.

Sorry I couldn't resist spamming this thread, but if anyone wants details on the NES stuff mentioned, here is a link (w/ pics): http://forums.nesdev.com/viewtopic.php?f=4&t=12716 In short: It will be used to release new games, and after I source decent NES cables for the USB adapter, a complete devkit for it will be available for hopefully around $30. I've seen many of kevtris' projects in person, and observing the development of this FPGA console has been fascinating, it's a lot of work and research over many years. Even back when the FPGA chip alone was probably $100 and it only ran NES, I've been trying to convince him to sell the thing somehow. Ideally, with his vision of how it should work. This project seems vague, but seemingly not incompatible with that. That's what's really disappointing with this project, it'd be a great use for those cores, but it seems bizarre to be at the stage of funding production, and not only is there no prototype, but the system specifications are part of the stretch goal. What the hell..? Nothing that's not be said already, but that's my reaction.

I found what I must have been remembering, it's referring to a proposed non-videogame project: A thornier problem is the fact that our current Yamaha contract, and presumably any future one we sign to purchase the cheaper chip, explicitly prohibits us from competing with Yamaha in the consumer music market. Ideally what would happen would be that we would design the product, do a non-disclosure agreement with them, demonstrate the product, and market the product as a joint venture using Yamaha's name and distribution channels. This is tricky, but I'm naive enough to think it could happen this way. [...] I've only just now looked into the AMY chip, it seems ahead of it's time and quite capable. I'm actually designing a chiptune-inspired MIDI-controlled synth, later on I might try experimenting with this kind of synthesis. It doesn't make sense today for sample playback since memory is dirt cheap, but for generating unique sounds, it would be quite interesting. Anyways, sorry about the off-topic posts.

In internal emails from Atari's arcade sound department, when introducing the YM2151 it was started that Yamaha would not allow Atari to use the chip in their home computers. Sucks for us, but I guess they didn't want to make Atari into their competitor, for the profit of an IC or two.

That sounds great but there is only one problem. No body wants to teach me. It would involve a good amount of self-help, but there are certainly people around who could help you learn. The more initiative you take (and demonstrate) in learning it, the more easily you will be able to receive help (i.e., try to figure something out or experiment before asking). If you want to learn how to program the NES, I could try to point you in the right direction. There are far more tutorials around for newbies then there ever were before. About porting Atari stuff to NES, the graphic systems are completely different. You would seriously have a much easier time converting a Coleco game to the NES (even though the CPU is different), they are very similar display-wise. In theory, you could program in C, and it will compile to almost any type of CPU assembly language (just expect it to be really slow and limited on these old systems) if you wanted to port the game program between systems. The graphics and user i/o would need to be completely rewritten for each platform though, and one can't do that without having a very good understanding of the assembly language and hardware of the system. You'll have a 10000% easier time by targeting just one platform, you know? Ports from NES to SNES however, would be relatively easy. NES programs would port easily to the Turbografx-16 as well.

I seriously doubt that changing any BIOS software will fix anything. If it works on one clone, it should work on all of them - there is a fundamental flaw somewhere. Also keep in mind the Tristar revisions could just as well have had hardware changes over time, independently of any BIOS updates.

Some NES homebrew developers would like to test their software in the hardware without having to pay $100.00 or $500.00 to have a prototype cart made. Hense the push for a hardworking NOAC. It has been professed that the Retro Duo 2.0 is the most powerful NOAC on the market right now but it doesn't play the PowerPak but on the Super8/TriStar, you atleast get a screen. I can agree that it would be cool if there was a clone system that was easy to develop for. Other than cost, there's no reason a clone couldn't have a serial or USB port to load programs, or use any computer-like peripherals. There was the "Dr. PC Jr" many years ago, but I'm not holding out much hope for anything like that to show up. The games you listed all use obscure software tricks, and almost certainly are unrelated to why the PowerPak doesn't work on clone systems. It's a hardware issue, an unsolved one.

Ah, too funny. The only software you really need is a 6502 assembler, same as it always was and always will be (yes, even if you code in C). You definitely don't need a PowerPak (it's kind of sub-optimal for development use anyways, it's just not designed for that in particular). You can write all the tools you want on top of that, if it outputs something the assembler accepts. If you're actually serious, the only way to make something happen is to do it yourself. Sorry but you just won't find someone else to program your ideas, regardless of it being a good idea or even being possible. There are no shortcuts. I didn't know how to program when I started writing NES stuff, it's not that hard!

I've never seen any kind of complete set of docs about it, or even a list of it's opcodes. But it is a general-purpose CPU. I'd imagine it has some instructions that are useful for doing some 3D calculations, and a way to DMA it's own memory into the SNES's Video RAM. Surely it could have been used for more than that, but no one would have because of the added cost of using it. Probably a chip like the SA-1 was more attractive to developers for general-purpose use (though it too was very rarely used).

I haven't checked nesdev in a while, but "memblers industries" had a NES PCB ready. No clue about the specs/mappers it supported though. IIRC it came with a socket for the CIC, so you could either cannibalize some or require your customers to cut a certain pin somewhere inside the NES. It's UNROM (mapper #2, up to 256kB ROM, 8kB RAM for graphics). I've still got a whole bunch of them, $5 for one if anyone wants to try it (be aware it's all surface-mount except for the ROM and CIC though). And there was at least one NES homebrew release on cart, albeit a limited one: http://nesworld.com/r_garage.htm