Atariman

Only just saw this. Thanks for the update! Got YM2020 ordered.

Thanks for sharing all of this great stuff! Will definitely be digging through it over the holidays.

That is super cool... The few pieces of equipment in his lab that I do recognize and have interacted with are ungodly expensive (and I suspect that those components are a drop in the bucket compared to the stuff I HAVEN'T used). What a cool setup!

Two of my favorites that I don't think I saw in this thread... Beetlejuice Taboo: The Sixth Sense Worth getting the games for the music? Probably not these days. I'm pretty sure these were $1 carts I bought 20 years ago...

Just now saw this and tried to order it, but got an error message that said that the store isn't taking orders right now. I'll try again soon, but was bummed since I, too, ordered the YM2017 compilation cart in the past! Definitely looking forward to (hopefully) getting this.

When it comes to 3D, I think it is pretty cool what some folks have come up with. Sure, the Jaguar isn't a 3D powerhouse, but I even appreciate when some 3D elements are thrown in alongside stuff that is predominantly 2D... For example, although I know it certainly isn't a homebrew game, doesn't Tempest 2000 have some polygon-based elements? Definitely correct me if I'm wrong...

Well, I will keep experimenting. I have some hardware coming in within the next month that should help me try out a variety of different concepts and if it garners any interest, cool. I just don't want to be accused of not sharing stuff that I find.

Clint and Chris... Out of curiosity, is there opportunity to merge RAM and bankswitching into the same concept? Like... say a bankswitching scheme that incorporates a 2 (or 4MB) window that can be switched along with the remainder of the address space dedicated to RAM? Or is that basically what the GD does already? I've got some spare cart PCBs that would be good for experimentation for this kind of concept and would be willing to give it a try at some point, I just need some guidance as to what would be most useful. EDIT: I think I answered my own question, but I'm not 100% certain. Looks like the Game Drive is what I would have expected it to be, largely there to switch between ROMs loaded onto an SD card... Not sure whether there are clever things you could do in a specially developed game to utilize the hardware any differently from that...

I fail to see how this detracts from any other thread in the Jaguar forum... ?

Another sidetrack, but what ever happened the ambitious stuff that Buddy was working on? I thought there was a project to work on the JagDuo and maybe some other stuff, but lost track.

I actually don't mind creating something that won't be widely used. If it adds functionality that's interesting to somebody or allows someone to do something that has never been done before, I'm game to give it a shot... It might just take awhile to get to every concept.

Yeah, I know that this isn't largely practical. It's one of the reasons why I was leaning toward not pursuing this as an upgrade since existing software wouldn't benefit from it (unless there would be a performance boost from being able to use faster RAM, which... there isn't) and it would require specific efforts to benefit from having more RAM. At this point I'm pretty much doing it to do it and share it with anybody else that wants it. I expect that as I work my way through concepts some will be more palatable than others. Adding 8MB of RAM does get me closer to a hazy goal in the future of something that I'd like to port to it, but it's such a pipe dream at this point that I'm hesitant to even tell anybody what it is for fear of sounding like an overly-optimistic noob... ...and no, it's not Area 51 or any of the CoJag games.

Minor update: I'm working on a few different concepts, but as for a memory upgrade, I started drawing up a schematic for an 8MB board. It's incomplete, but I've got it down to about 11 ICs and a 3.3v regulator. General strategy is to demux the address coming from Tom and use SRAM. It's a little pricey, but considering the premium and/or difficulty in getting drop-in replacements, I thought it was worth a shot. The other ICs are bi-directional level shifters since 5V parts, at least in my search, have been a little rare. My goal is to create something with parts that don't appear to be nearing their end-of-life and seem to be available from multiple suppliers. Once I'm working on layout, I'm going to try to line up the board/pins with the existing pads that they'll need to be soldered to since my thought would be that it would require removal of the existing RAM and then soldering pins to some of the pads. It would also require a jumper or two to some test points on the board. I still need to go through and double-check how I think it would work, but what I have so far seems feasible and wouldn't be a total nightmare to solder since it would require minimal jumper wires. I think it ought to work with any Fast Page Mode accesses as well since that would just be updating part of the demuxed address and there shouldn't be any performance issues jumping from one address to another. If anybody is interested in the parts that I'm thinking about using, I'm willing to share that information. Once I've put together a couple prototypes and modified the boot rom, I should be able to test it out and see if it works. First test will just be trying out an existing game to make sure that it still works, then I can move on and try my hand at creating something that would access the additional memory. Next steps after that are unknown. I've been pursuing some other leads on CoJag information that could lead to other concepts, but wanted to share this for now. Thoughts? Comments? I know this isn't as easy as piggybacking parts and running jumper wires, but I'm hoping that this could be beneficial for more than just me.

I had meant to respond to the concept of adding RAM to the DSP connector earlier. I was scratching my head, especially when I actually looked on the schematic and was met with essentially what I expected... just... DSP stuff. Might be nice for interfacing with a MIDI device, but... Erm... Or was I missing a joke somewhere?

? Yes, RAM bank switching sounds just fabulous... but I do see where I went wrong in my interpretation of the memory maps. Yes, I see that 8MB is the maximum based on the memory map - I suspect that if I were to devise a way make that happen it might make a few people happy, though. Let me see what I can cook up, though!

Speaking of RAM... Although no current games could use it, do you think that a RAM expansion design would be welcomed to the Jaguar? The CoJag schematics have made me think about what it would take to interface "modern" SDRAM with the DRAM interface of TOM and I'm thinking about what signal manipulation it would take to bridge that gap. SRAM looks like it's still too expensive (which, too, would require more hardware to make it work, anyway) and DRAM seems like it's just getting harder and harder to acquire in any size. SDRAM is plentiful, cheap, and fast at least. If I can figure out a strategy for tackling the signal translation, I would be willing to develop something on an FPGA and lay out a PCB for the upgrade that would attempt to reduce the amount of pain introduced from installation. On a similar note, maybe my calculations are incorrect or perhaps I haven't found the right errata information yet, but shouldn't TOM be able to address 16MB of RAM? I've seen 8MB thrown around and that IS that maximum that was accounted for on the CoJag, but I was wondering how much of that is a limitation of TOM and how much of it came from the fact that they used the same RAM in the CoJag as they did the Jaguar which looks like it required some finagling to get it to work. The datasheet says RAM can be 64-bits wide (which, when all of the current RAM ICs are strung together, it currently has) and up to 2M columns, calculating out to my 16MB value. Although the last address pin for RAM isn't connected to anything, it does appear to go to a test point which would be easier to solder a wire to than a pin on the chunky QFP that is TOM. Not sure how excited software folks would get about that, but it seems like that would be awfully nice to be able to have everything in ROM transferred to RAM, decompression be damned. At this point I guess the biggest benefit would be for future homebrew CD titles? Anyway, just one of the things I've been chewing on for the last few days.

So I'm leaning heavily on the quotations around "practical". My goal is to explore things that can be done to the Jaguar that at very least don't create incompatibilities with existing software and/or increase performance where software suffers from the bottlenecks of the architecture that was chosen. Yes, hacking in a 68EC020 wouldn't be a quick solder job, but I think I've got the ability to do it (and maybe it would give me a reason to break out my surface mount rework station). The big kicker from what I can tell would be the race conditions that were mentioned by Zerosquare - I was really hoping that those didn't exist, but... :) As for doing it "just for myself", I'm definitely open to creating a design that others could benefit from if something pans out. I just don't want anyone to get the impression that I'm one of those folks who swagger into the Jaguar forum proclaiming to have worked out all of the steps to unleash some unknown, unseen power of the Jaguar and for the low, low price of $X. If there's opportunity to improve something in a way that is desirable to others, I'm more than willing to share what I've done and would even entertain the idea of trying to produce something, but until then I'm just trying to take small steps to understand what has been done before and where the dead ends are. I know there are people on this forum that have done a LOT of work with the Jaguar, so I'm really trying to step carefully. And curse you, @Clint Thompson, because I've had bankswitching stuff starting to kick around in my head now, too! ? But seriously, thanks to everyone who has responded so far. I'm digging deeper.

Out of curiosity, what have you found with the peripheral HW bits? I have scrolled past some of that in the manual, but haven't totally tried to get my head wrapped around that yet. Always curious about what others have discovered (and I will tell you - I have read a LOT of Jaguar hardware posts over the last week or so). I've been digging around in the CoJag schematics a bit and haven't gotten the hard drive interface figured out quite yet. I've mostly been looking at the hardware configuration differences on TOM, but the hard drive interface has been at least in the back of my mind.

I hear you there. Most of what I've done so far is just think about some of this stuff. Once temperatures start to drop here and I can't justify getting work done outside anymore, I'll be spending more time working on this stuff. I'm going to try something... and I'll share the results once I've either had success or given up! (but I'll tell you that I don't know enough about developing new bankswitching schemes (yet?), so it won't be that (again, yet?) )

To elaborate the reason I ask is because I noticed that XMA[5] is tied to ground in the schematic and the MEMCON1 register seems to have the 32-bit Micro bit set to 0. That to me would indicate that it would require a change to the boot rom, but not to how XMA[5] is pulled.

Thanks for the response, Zerosquare - I'm amazed that it is already at its most aggressive timing considering the complaints I have read about Atari's decision to use "slow" RAM. Knowing this, I'm definitely not as keen on trying to replace it, not even with something that didn't require the DRAM refresh since I didn't figure that alone would be much of a benefit. I had been looking at the Cojag schematics and wondered why they seemed to use the same RAM and configuration (just more of it), so your answers explain that as well. Any thoughts on the lengthy and arduous task of hacking in a 68020? I've looked less into this aside from understanding that it should be generally backwards compatible with the 68k. The thing I'm less certain of is why the bitness of processor is seemingly selected via an input pin on TOM in addition to Bit 14 in MEMCON1. Is one to indicate the width of the interface to the micro and the other is indicative of the bitness of the micro core? (and for some reason I can't seem to find the 68020 in the Cojag schematics. I must be missing something.) Thanks for the responses to my n00b questions. I know I'm not the first to ask about adding a 68020, but I didn't find these specific questions being asked. Also, I'm most definitely interested in the FPGA implementations of the Jaguar. It has been quite some time since I've written anything in a hardware description language, but it's something I'd like to get back into at some point. I'll plan on reaching out to SCPCD at some point to see if there is anything that could be shared...

So it has been a long time since I've actually posted, but there was something I have been mulling over ever since I read through the TOM manual... I have an embedded hardware and software background and I don't think I ever realized just how configurable the Jaguar chipset is until now. Looking at the schematics and trying to get my head wrapped around the documentation, I started thinking about "practical" upgrades, meaning hardware improvements that could be made that wouldn't require software to be rewritten. While admiring the architecture laid out in the manual, I did see that the chipset can support a relatively wide variety of RAM options. This left me with some thoughts and questions... 1. How much is RAM speed contributing to the bottlenecks in the Jaguar? If it had RAM available to it that was...say 50ns or faster (versus the 80ns in it now), would there be ANY performance changes, noticeable or not? 2. How widely available is documentation pertaining to the boot ROM? Based on the TOM manual, there would be some changes that would have to happen to MEMCON when the system starts up. I think I have my head wrapped around how these should be configured for the most part, but I'm not sure where to start for looking at changing the ROM. 3. Whether due to age of the parts or rarity, I'm not seeing a lot of drop-in replacements that are available. There are some that are close... but where the current DRAM has signals for Upper/Lower write enable and a single CAS input, the ones I've been finding have had a single write enable with upper and lower CAS inputs. This would presumably add a need to modify the board to accommodate it. 4. I also see that the manual mentions SRAM, but with the multiplexed addressing used by DRAM, I'm not 100% certain how this would be implemented. I'm guessing that it would have to be demuxed and I've read more about hardware solutions for that, but I wasn't sure if I'm right. I'm thinking that SRAM is the use case for the option to turn off refresh entirely. So... thoughts? I have some spare Jaguar hardware and the ability to make modifications (and lay out PCBs if necessary). I'm also trying to steer clear of overclocking and other "easy" modifications that have been proven to reduce software compatibility (in addition to adding stress to the hardware). Hopefully this is a change of pace from the usual 'what if' topics that get kicked around regarding Jaguar hardware. I'm not thinking that this would be leading to some marketable upgrade or anything like that, merely thinking about what someone could do to get better performance out of the hardware with as few modifications as possible. I was also thinking about the modifications that would be necessary to add a 68020... but that's a mess that I'm not quite ready to wade into (although the TOM manual did get those gears turning in my head as well). Try and cut me some slack, Jaguar HW gurus... I'm doing my best to get up to speed and not be totally ignorant. ?

I always enjoyed reading his posts as well... Condolences to his family and those who knew him.

I second what was said about walking. I went down two notches on my belt when I was there for two weeks. Totally wonderful experience and I have never felt so safe in an unfamiliar place.

I had some of that experience, too. I think without knowing the language, there are going to be times that this is unavoidable. Restaurants usually weren't too bad (and I was a big fan of the ones where you ordered at a kiosk and brought the ticket to the counter) but there were times where unless I had been totally fluent, it was not possible to effectively communicate. I learned to actively avoid anything that required any amount of instruction.