nesdev.org

tepples wrote:If it doesn't handle carts that write to ROM, how does it handle carts that write to SRAM?

I was just speculating, but to speculate further: they probably expected the snes to write to sram. You can sense the difference, so they could have easily allowed sram to work.

How can the chip tell the difference? Only HiROM games have SRAM at $0x6000-$0x7FFF. As I understand it, LoROM games tend to have it in $700000-$7DFFFF, which overlaps where 32 Mbit HiROM games have ROM.

You could watch for which one of those areas the SNES wrote to first and then assume if it was Hi/LoROM based on that and continue to operate with that assumption. If they did something like that it would give even more reason as to why flash carts don't work.

Again all of this is only speculation, but they are things that could have happened that would break current flash carts.

Who knows perhaps the both the PAR and flash carts use the nn6000h-nn7FFFh expansion, something like that would also cause issues.

Hey IFL

Good work!!! If you are accepting orders (or preorders), I'd like to be on that list.

Please let me know.

Thanks!
Mark

Awesome project! Will certainly buy one at some point. And for your information, there are new SNES cart case available. A reproduction seller called Pachichi (you might remember him posting basic questions about repro and looking for programmer for random vapor-ware project) had mold made from the official cart and is now selling blank cases for 6$, maybe less if you buy in bulk. I'm definitively not affiliated with him but always welcome the use of new parts over destroying carts. http://store.retroquestgames.com/

I must say that your different projects are becoming more awesome every days!

Thanks guys.

Markfrizb wrote:If you are accepting orders (or preorders), I'd like to be on that list.

Not formally, There will be plenty to go around. I'll put you down on the list so you can be one of the first to get one. I'll post here when they're ready.

SkinnyV wrote:And for your information, there are new SNES cart case available.

I had heard about this, but I didn't have all the details so thanks for sharing. I just ordered a pair of his cases to see how they are. All I knew is someone supposedly had them and they were identical to the standard carts. So that's part of the reason I went with the original PCB form factor. They fit in the retrousb cases as well, but they are also not doing SNES repros anymore so I'd have to see if they still have cases around they'd like to part with.

Nice and compact, looks excellent paul!

I would imagine the PAR would work. I have a PAR MK3 I could try.

So every game up to 2MB in size, without using any special chips should work without a hitch?

mog123 wrote:So every game up to 2MB in size, without using any special chips should work without a hitch?

16MB, I made the same mistake. Those are 8 megaBYTE ROM chips. And it depends on your definition of "without a hitch" since some games have copier protections and won't work on this because of the larger RAM size (Super Metroid, for one), but uCON64 is able to patch most of those...

The hardware on the cart could support SRAM size limiting. Only INL knows for sure.

Yeah we'll have to make sure someone with a PAR gets one in their hands to test em out when they're ready for full release. I appreciate the volunteers!

mog123 wrote:So every game up to 2MB in size, without using any special chips should work without a hitch?

That's the goal, but actually all the way up to 128Mbit as qwertymodo pointed out. Which is somewhat of a lie, There's only ~90+ some Mbit addressable to the cart at one time due to the SNES memory map. But I can put a unique bit in every last bit of addressable ROM. To answer your question though, anything that works on a HiROM or LoROM board should be as 'hitchless' as possible.

As for the issue of some games having 'flash cart' protection like metroid, I can pretty much make any of the memories look as small as desired using the CPLD. So if something breaks with more than 64Mbit of SRAM I can 'trim' the SRAM down to 64Mbit to trick the rom. I can do similar things with the flash (or put a smaller flash chip on the board) if desired as well. Although choices like that may have to be made at time of purchase, not sure how configurable they'll be afterwards. I did add a slide switch to the final PCB that should allow for switching back and forth between Hi and Lo ROM on a single cart. There's enough space in the CPLD to support that.

In other news I did fix my bug with the programmer and have successfully tested out both Hi and Lo ROM games thus far. Still have some things to work on before I can test out 6MBYTE games, but they shouldn't be much trouble. I'm ordering the first 'official' batch of boards this evening, along with a big batch of kazzos to go along with em

Additionally I got a pair of the SNES cases, thanks for pointing them out to me SkinnyV. The good news is they're pretty awesome and I'm pretty impressed by the 'snap' open/close action of the cases. Honestly I wasn't expecting them to work well, but they're pretty great in practice. I'm actually planning on having a batch of those when the boards are ready to release so it can be a package deal with case and all right out the gate!

Would it be any cheaper to use an 8 MiB chip and a 4 MiB chip instead of two 8 MiB chips?

tepples wrote:Would it be any cheaper to use an 8 MiB chip and a 4 MiB chip instead of two 8 MiB chips?

Yeah it would. I don't think there is any one game/hack that uses that much memory at the moment anyways though. Either way all the address lines for a 8MByte chip was already there so it only took another /CE pin on the mapper to expand up to 16MByte.

I'm not sure how many different flavors of memory sizes I'll sell it in, it'll kinda depend on what people request.

If you really wanted to test the full range of the mapping, the No-SDD1 Star Ocean hack would be just the thing...