The Goal: System Card 4.0

[OldRover]

This is something tomatheous and I (and a few others as well) had talked about on IRC awhile back, but never really did anything with. So, I'll outline some of the ideas that we came up with as well as some of my own that I considered afterwards.

THE PLAN:
A new system card, revision 4.0. Designed to play all existing PC Engine/TurboGrafx CDROM software as well as new software built specifically for the new card.

WHAT IT WOULD HAVE:
Switchable rev 1.0, rev 3.0, and Games Express BIOS code + normal RAM
Full ACD support
New 4.0 BIOS, supporting the modern CDROM standard
Onboard 4-channel Vorbis decoder with 8MB buffer and mixer (would all but replace the built-in ADPCM circuit)
tomatheous also suggested an onboard coprocessor, possibly an NEC V810, but this would probably be very difficult to do

Like the original cards, two versions would be available; one for the normal hardware and one for the Duo.

Most of the card would likely be handled with an Actel flash-based FPGA chip.

Any suggestions, comments, criticisms, cash offers for development, etc. are welcome.

[nat]

Since my programming skill level is undoubtedly inadequate to help the project in any significant capacity, I'd be willing to offer to help back the project financially. Although I guess I'd probably need some kind of assurance that some fruit would come of the labor so I'd know I wasn't just blindly dumping money into a black hole.

All that said, it'd be really cool to see some kind of 32-bit CPU on the card that new 4.0 games could utilize.

[Turbo D]

I think that its a great idea! I would definitely buy them from you!

[esteban]

This is a very kool project. One suggestion, though: instead of a Vorbis decoder, use WMA. ;0

[SignOfZeta]

This is a very kool project. One suggestion, though: instead of a Vorbis decoder, use WMA. ;0

Better yet, Real Media!

Anyway, do you really need to make one for the Duo and one for non-Duos? Unless I'm missing something the original reasons for there being "Pro" and "Duo" versions of the original AC (the cost of memory, mainly) have long past, and making two versions of a card now would just increase the price. Am I missing something?

Since an Arcade Card Pro will work in any PCE, it seems like you'd just want to make a card like that.

[OldRover]

tomatheous would know better than I would about the possibility of a single design.

I hope esteban was kidding about WMA. It's a proprietary format and Microsoft would throw a shit fit.

nat: I'll probably support the project financially (it's what I do, hehe) so no problem, but if anyone else wants to contribute, it'd be best to wait until there's at least a solid plan, perhaps a prototype or two.

[nat]

Sure, sounds great. You've got my word if you reach that phase I will be more than happy to contribute.

[nat]

As long as I get one of the prototypes after you're done with it. 

[PCEngineHell]

Youve got my word if it does not take more then 2 years to do,Ill buy one,esp if you trow in free signed copies of Neutopia 3

[chop5]

wow, i would buy one!

[spenoza]

I'm going to add in here that the best system card you can make is one that is fundamentally simple and doesn't add too much extra functionality (at the risk of over-reaching).

That said, it'd be nice to have some expanded battery backup (save) capabilities. Though I suspect it to be near-impossible, save-state capabilities would rock, too.

[OldRover]

Save state probably isn't possible, but I don't know for sure. Battery backup was something that crossed my mind (built-in Tennokoe bank, perhaps). Adding too many capabilities wouldn't be a good thing; the original reason for doing a new system card was to standardize CDROM access so new discs would be much easier to produce.

[MissaFX]

This all sounds great:
Switchable rev 1.0, rev 3.0, and Games Express BIOS code + normal RAM
Full ACD support
New 4.0 BIOS, supporting the modern CDROM standard
Onboard 4-channel Vorbis decoder with 8MB buffer and mixer (would all but replace the built-in ADPCM circuit)
Built in Tennokoe Bank

Can you elaborate more on what an onboard coprocessor, possibly an NEC V810 would allow homebrew programmers to do?  Also, like someone else said, couldn't you just make 1 version of this card for all PC-E systems, Duo and non duo?  Or would it actually save you money to make separate versions? 

Is there any way to make a device that takes an internal screenshot and saves it to battery memory on this card or something?  Then you could save your best 10 scores or whatever on it and show them off/view them again?

[OldRover]

An onboard coprocessor would give the system a major processing boost, as a lot of tasks could simply be offloaded to the coprocessor, similar to how it's done on the Sega CD. As for a screenshot function...I don't know about that one, the hucard port would have to have video in pins, and I don't think it does...

[MissaFX]

An onboard coprocessor would give the system a major processing boost, as a lot of tasks could simply be offloaded to the coprocessor, similar to how it's done on the Sega CD. As for a screenshot function...I don't know about that one, the hucard port would have to have video in pins, and I don't think it does...

I was thinking of a video ram dump to the card, even if it might be a little cluttered.  Maybe that cannot be done though, I have no idea.

Could a coprocessor be somthing like a mpeg4 video decoder?  This would allow mp4 video to be used in new PC-E games since mp4 can be low bitrate and high quality, it sounds like the perfect format for a single speed cd-rom.

[OldRover]

A VRAM dump to the card *might* be able to be done (sequential buffering to main RAM -> upload to memory bank on card...I think it could be done), but it really wouldn't help much since this is a planar tile + sprite system, not a linear pixel system.

[Kitsunexus]

You should add an FPGA that simualtes a SNES.


In all seriousness, this is freaking awesome.

[BigT]

This all sounds great:
Switchable rev 1.0, rev 3.0, and Games Express BIOS code + normal RAM
Full ACD support
New 4.0 BIOS, supporting the modern CDROM standard
Onboard 4-channel Vorbis decoder with 8MB buffer and mixer (would all but replace the built-in ADPCM circuit)
Built in Tennokoe Bank

Can you elaborate more on what an onboard coprocessor, possibly an NEC V810 would allow homebrew programmers to do?  Also, like someone else said, couldn't you just make 1 version of this card for all PC-E systems, Duo and non duo?  Or would it actually save you money to make separate versions? 

Is there any way to make a device that takes an internal screenshot and saves it to battery memory on this card or something?  Then you could save your best 10 scores or whatever on it and show them off/view them again?

Sounds like a fun project, but there is always a danger of creating a system within a system that does not really represent the magic and the spirit of the PC-Engine.

For Homebrew projects, ISO 9660 support would be nice to allow for much easier mastering of the CDs.  Extra built in memory banks also seem feasible.

Adding coprocessors and codecs may get complicated.  There's relatively little space to work on using a HuCard form factor, so it may be hard to hack together with a soldering iron and off the shelf components.  Plus, someone would need to synchronize this with the native hardware.  The easiest thing to do would be a video codec whose output would be overlayed over the native output (but I'm not even sure if the hucard has all the pins necessary for this.

Other features like screenshots and save states are theoretically possible, but are much easier to implement on an emulator.

[spenoza]

I guess for those developing and for those of us reading, we must recognize where desires and practical reality split.

I would be most supportive of a simpler design that keeps costs down. More memory is great and support for more standard CD formats (like ISO 9660) might also be very welcome. I would be very concerned about including additional CPUs, DSPs, etc... as that could raise the cost dramatically. What, essentially, is the best compromise between simplicity, features, and cost?

[Turbo D]

Is it going to be a region free hucard? I'd love to play my acd games without using a kisado for the system card.

[SignOfZeta]

I'm sure it would be dual region, maybe even auto switching. That sort of thing is peanuts compared to the borderline impossible stuff other people are talking about (ie: MP4 decoding).

[OldRover]

The idea for an onboard sound processor wasn't mine originally (someone wanted to do mp3, but that's not a good idea), but that could easily be removed, as it's mainly a "perk feature". Again though, the main idea for this was mainly to develop a new card that brought in the standard CDROM format so new software would be easier to produce (this might also lead to re-pressing of older games). Making it ACD-able is also important, since it's an existing technology and would greatly help any new game development efforts. I certainly don't want to add a whole mess of features, as it's really outside the scope of the project. As far as region-free goes...don't know about that, tomatheous would know more. But really, a lot of people are having region mods done lately it seems, so I don't see that as an immediate issue. We'll see what happens in the next couple of months.

[Kitsunexus]

The Old Rover: Forget all the extra features everybody wants and just make the basic universal System Card thing. Some people might bitch and moan about X feature might not be included, but it would be better to have a solid new System Card than some mediocre jack-of-all-trades "console enhancer".

Seriously, with all the features you guys want to add, why not just call his card the Aladdin Turbo Edtion.

[guyjin]

the aladdin was awesome. Wish I had one.

[OldRover]

Yeah Kitsu, I tend to agree with that train of thought. Making a universal card is pretty much my own personal goal for this project; a card that will play all PCE/TG16 software regardless of format, including that one title (Altered Beast CD, was it?) that only plays on the system 1.0 card.

[CrackTiger]

If you do consider some sort of save file storage, I think that it would be worthwhile to do if it could copy and manage single save files. Otherwise, a Tennokoe Bank style bank swapping feature isn't a worthwhile complication.

[Ceti Alpha]

I like the title of this thread. 

...best topic evar...

[Kitsunexus]

Hey TOR, while I think the sound processor is pointless, it does bring up a fascinating question:

Was the OGG format chosen simply because the Turbo cannot do floating-point maths, or were there other reasons?

[OldRover]

I chose a Vorbis player because it's patent-free and has better compression and quality than MP3.

[Arjak]

Thank you, Old Rover, Thank you! My day started out great, recently took a slight turn for the worst, and now, reading this topic has made it better than ever! I would definitely support this in any way I can! (Which at the moment limits me to moral support but hey, you never know!)

One possible feature instantly comes to mind: No Turbo/PCE games up to this point have ever been more than one disk. I was wondering if there was any way to implement this as is. If not, this would be a great feature for your new system card!

*Returns to bouncing off the walls in excitement!*

[MissaFX]

The biggest question on my mind is:  Will you be able to get a really good artist to do the label for this card?  I'd love to see one of those lables which is plastic and you turn the card back and forth to see it animate

[Turbo D]

that would be bad ass. It could be a hologram like my pokemon cards 

[OldRover]

One possible feature instantly comes to mind: No Turbo/PCE games up to this point have ever been more than one disk. I was wondering if there was any way to implement this as is. If not, this would be a great feature for your new system card!

No game has ever been more than one disc because no game has ever required more than one disc. Multi-disc games are very possible; in fact, I've already done some experimenting with disc swapping techniques as proof-of-concept. It's just that there's never been a need, and I would imagine that back-in-the-day, since discs were expensive enough to produce, there was likewise no desire to make a game that required multiple discs. Times have changed...

Missa: We'd probably be able to convince CrackTiger to do the artwork for it, hehe. His artwork is fabulous, and I'm sure he'd enjoy the opportunity.

[Joe Redifer]

Does the HuCard slot even have a "sound in" pin?  From what Bonknuts (I think) was saying, you can really only add RAM via the card slot.  It can't be used to upgrade the video or audio in any way. 

Would this be made to work with Jap or US machines?

Also I am considering moving this thread to the development forum unless I hear any convincing arguments against (and they had better come soon).

[nat]

Also I am considering moving this thread to the development forum unless I hear any convincing arguments against (and they had better come soon).

Someone already moved it! Is that a good argument?

[Joe Redifer]

I'm moving it back to the old forum and then back here just so I can say I moved it!

[OldRover]

Does the HuCard slot even have a "sound in" pin?  From what Bonknuts (I think) was saying, you can really only add RAM via the card slot.  It can't be used to upgrade the video or audio in any way.

The hucard slot has an "audio in" line, but it's monaural and I'm not exactly sure where it enters the system (mixed in with the ADPCM circuit maybe?). It lacks any kind of video lines though.

Would this be made to work with Jap or US machines?

It could probably be made to work with both.

[Keranu]

I like the title of this thread. 

...best topic evar...

Agreed!

I support this idea and think it's really neat. However as BigT said, I hope this doesn't take away from the spirit of the Turbo/PCE.

[ParanoiaDragon]

Thank you, Old Rover, Thank you! My day started out great, recently took a slight turn for the worst, and now, reading this topic has made it better than ever! I would definitely support this in any way I can! (Which at the moment limits me to moral support but hey, you never know!)

One possible feature instantly comes to mind: No Turbo/PCE games up to this point have ever been more than one disk. I was wondering if there was any way to implement this as is. If not, this would be a great feature for your new system card!

*Returns to bouncing off the walls in excitement!*

Nod/Rover probably forgot, but, we do have a game that Lord Cack & I have come up with, that will probably require more then 1 disc.  But, this game is pretty far away from completion, seeing as how we haven't even released Mysterious Song yet, & we're seeing how long these projects can take.


The main thing I would like to see, is more ram or whatever to load stuff in a level, so levels can be longer, feature more paralax, more enemies(rather then just 1 or 2, with some color swap's)etc.  Basically, you have a Hucard, which can load anything, anytime, but the Super CD, can only load 2mb's, & then to load anything beyond that requires more disc access.  The Arcade card doesn't rememdy this, I forget how it works, people keep explaining it to me, & it never sinks in.  The Sega CD, IIRC has 6mb's to tool around with.  I would just love 6, 8, etc. mb's to for us to play with, to do more in a level or whatever.  Get my drift?

All these other things sound great as well though.  But, more then the standard 2mb's is the main thing I want.  I want to eliminate that limit.

[Kitsunexus]

The biggest question on my mind is:  Will you be able to get a really good artist to do the label for this card?  I'd love to see one of those lables which is plastic and you turn the card back and forth to see it animate

that would be bad ass. It could be a hologram like my pokemon cards 

Thirding the lenticular label art.

[GUTS]

I don't get why, if you guys want to do something like this, don't you just program for a more powerful system that already HAS these features?  It seems completely preposterous that not only are you going to design and manufacture a System Card 4.0, but also manage to sell more than the 4 or 5 to people on this board.  Especially when you consider that import adapters for the Turbo sell for so much money, yet no one has designed (or hell just ripped off existing designs) and manufactured a new batch in like 15 years, so the logical conclusion is that it just isn't possible or cost effective to manufacture something that will fit in the hucard slot anymore.

[MissaFX]

Especially when you consider that import adapters for the Turbo sell for so much money, yet no one has designed (or hell just ripped off existing designs) and manufactured a new batch in like 15 years, so the logical conclusion is that it just isn't possible or cost effective to manufacture something that will fit in the hucard slot anymore.

I'm willing to pay a good deal of money for one as long as it has features like built in bank, the fact it has the arcade hardware as well adds a good amount of value.  If it opens up cheaper or more available homebrew I am all for it.  I figure a lot of junk hucards could be collected.  The pin part and the main body of the card could be re-used and a new PCB would be attached to the pins.  It might need a bubble back part of the card, but that could be manufactured or an existing part could be purchased that could be easly modified to fit.

[Kitsunexus]

I don't get why, if you guys want to do something like this, don't you just program for a more powerful system that already HAS these features?  It seems completely preposterous that not only are you going to design and manufacture a System Card 4.0, but also manage to sell more than the 4 or 5 to people on this board.  Especially when you consider that import adapters for the Turbo sell for so much money, yet no one has designed (or hell just ripped off existing designs) and manufactured a new batch in like 15 years, so the logical conclusion is that it just isn't possible or cost effective to manufacture something that will fit in the hucard slot anymore.

I will buy one, and I don't even have a Turbo.

[ParanoiaDragon]

I was thinking, maybe, if this gets made, it should be included with whatever particular game is the first to utilize it.  However, even that, I don't know if that would work.  But, it would force people to buy one particular game(Neutopia 3) to play other future games...  Just an idea.

[MissaFX]

I was thinking, maybe, if this gets made, it should be included with whatever particular game is the first to utilize it.  However, even that, I don't know if that would work.  But, it would force people to buy one particular game(Neutopia 3) to play other future games...  Just an idea.

Maybe it could come with a unique digital coupon code to get a discount on a future homebrew game which supports it.

[TurboXray]

 Oh - you made a thread for this 

 I think the most important thing of all, is supporting absolute address booting of the first sector instead of the PCE's index 01 method. Seems the stupid pressing houses can't figure out how to set index 01 to the correct sector    This would allow you to boot from an ISO9660 image.

 Here are the main features that would be great and still have that PCE/TG feel(order of importance):


 - Port based setup like the AC card did would be great. You could have up to 8 ports memory mapped on the hardware bank and could be programmed to point *anywhere* in the hucard address range. There's nothing like saving code/complexity and cpu cycles like free auto increment/decrement I/O port based accessing. That's like having 8 hardware indirect registers but with 21bit address range. Each port should have a Bank number associated with it too like the AC card. Making it compatible with the block transfer instructions.

 - Simple single channel PCM DAC with volume setting. Standard 8bit version. You set the frequency and which buffer to read from. You wouldn't need much room, just maybe two 512kbyte buffers. The output could be on the audio input pin of the hucard port. It'd be great for things like samples or mono style instruments.

  - External interrupt for a finer TIMER value. Maybe something that could even be in sync with hsync (you'd have to reset the TIMER on the first scanline VDC interrupt to resync it).

 - 16bit/16bit -> 16bit MUL and DIV support. Port based of course, just like the SNES one. It help make coder cleaner and more convenient. Something under 10 clock cycle time.


 A Bonus feature could be:

 - Switch into all RAM mode for banks $00 to $7F. You could do this from main ram (bank $F8). Basically you could run any hucard/rom from a CD that's 8megs or less (or an all RAM with the SF2 32megabit address banking system - yeah I know it only *used* 20megabits). This would also allow for different/custom system cards to be dynamically uploaded into RAM(it would be lost on power off).

[nat]

Great stuff there, Bonknuts. Keep it up!

[NecroPhile]

No game has ever been more than one disc because no game has ever required more than one disc.

It's obviously not the same thing, but CD Battle requires a few extra discs to generate character stats.  Are there any similarities between your proof-of-concept program and CD Battle, or does CD Battle not really read much of anything from the extra discs?

I don't get why, if you guys want to do something like this, don't you just program for a more powerful system that already HAS these features?  It seems completely preposterous that not only are you going to design and manufacture a System Card 4.0, but also manage to sell more than the 4 or 5 to people on this board.  Especially when you consider that import adapters for the Turbo sell for so much money, yet no one has designed (or hell just ripped off existing designs) and manufactured a new batch in like 15 years, so the logical conclusion is that it just isn't possible or cost effective to manufacture something that will fit in the hucard slot anymore.

It's not like they're trying to make a 32x for the Duo; adding memory and support for a standard disc image is not going to make games look and feel earth shatteringly different from the old games.  The success of Tototek and Neo Flash has shown that there is a market for new hardware, albeit not a huge one.  I'll bet that the reason there aren't any new converters being made is because there is no source for the HuCard socket.  Any monkey can make a board that'll fit the socket (the flash carts are proof), but it'd be mighty expensive to have a few sockets fabricated.  The cost of new converters would end up being just as high as the old ones, so it's more logical to push for a region switch, which have been made recently.

I figure a lot of junk hucards could be collected.  The pin part and the main body of the card could be re-used and a new PCB would be attached to the pins.  It might need a bubble back part of the card, but that could be manufactured or an existing part could be purchased that could be easly modified to fit.

Cannibalizing old games would be unnecessary and would make little baby Jesus cry, even if they were all copies of Turbo Golf and Keith Courage.  PCBs can be made with pins to fit the HuCard socket, just like the flash carts and converters.

[Kitsunexus]

It's obviously not the same thing, but CD Battle requires a few extra discs to generate character stats. 

Dude, that sounds like Monster Rancher! TELL ME MORE! ^___^

[NecroPhile]

Dude, that sounds like Monster Rancher! TELL ME MORE! ^___^

There's not much more to tell.  After the game starts up, you insert various discs (audio or whatever - I don't think that it really matters), and the game sets your character stats (what you look like, weapons, spells, etc.) based on something from those discs.  You set up three characters, fight three other characters in a Final Fantasy style side view battle, and you win or lose the battle.  That's it.  That's the whole damn game.