nesdev.com

What symptoms do you see?

Are you sure Mr. Gimmick can start without the IRQ?

squall926 wrote: ↑

Sun Apr 11, 2021 7:52 pm

-- PRG-ROM Max 128kB 8kB window (A16, A15, A14, A13)
type cpu_t is array(0 to 3) of unsigned(3 downto 0);

But Mr. Gimmick has 256KB PRG?

The other things you ask about tentatively look ok

wait until (falling_edge(romsel)); [...] when 0 to 7 => ppu_bank(reg_sel) <= unsigned(cpu_data_i); Data bus is not valid on the falling edge of /ROMSEL! You must wait for the rising edge, or approximately 70ns after the falling edge. See: https://forums.nesdev.com/viewtopic.php?p=244126#p244126

During access to addresses $21xx, the B bus is a copy of the bottom 8 bits of the A bus. During DMA it's driven according to the DMA registers. The rest of the time, it does random weird things. I don't think it's ever undriven (although looking at the die shot would let us tell), but it seems to ge...

Unfortunately, I promise, it's a problem with just how the NES generates video. Certain TVs might deal better with edges between different brightnesses of grey, but vertical lines between colors will always have that appearance. NTSC NESes have a different artifact - the pattern happens every 3 scan...

The 8-bit depth is pretty relevant here. I don't think it's fair to say that it means "almost nothing"? While I was definitely being flip, my point was more that focusing on a single bit of difference in bit depth is why it's misleading. The GBA has the problem that its output DAC always resamples ...

So why is the NES with 7-bit samples and a 1.79 MHz CPU sound better than the GBA with a MUCH better CPU, and 8-bit samples, and if the GBA could sound better, is there a example without the "scratchy" sound of the GBA? Because bit depth means almost nothing. https://en.wikipedia.org/wiki/1-bit_DAC

2. You mean like use its mapper emulation to intercept $4016/$4017 access kind of like a Game Genie? Or patch the ROM on the fly? The Everdrive does have mapper 99 listed so maybe it does something hacky with these games. I have no idea if mapper 99 was listed years ago when I tried VS SMB though. ...

um... 1- Vs. SMB was only released for the 2C04-0004, so if you played it on a home console on a flashcart, the colors would all be Very Wrong. If you don't remember them being Very Wrong, you most likely played a version that was patched for home consoles. very-wrong-colors.png 2- The Everdrive can...

The Vs. System is not the same hardware as the NES, and you shouldn't have $4016 and $4017 behaves the same for both.

It is absolutely correct that if you put Vs. SMB on a NES cart it would perceive the coin drop as continuously activated.

If you don't know that the compiler is doing the slow thing, then it's definitely premature to try to "help" it.

This is true for all optimizations ever, honestly. If you don't know what's taking up time, your assumption will probably be wrong and will make things worse instead.

Do you know that SDCC is doing the slower thing?

You could get one of the very cheap 8/16-channel ones. Search AliExpress for "fx2lp". You'll need all the probes, too, so it may not work out to a win to buy things piecemeal, though.

It's limited - only 12MSa/sec at 16 channel, but it's what I used for the PPU logic analyzer traces.

There are rather few mappers that explicitly can provide 4 nametables: NES-commerical era: mapper 4/206 = MMC3 or its simpler predecessor Namco 108 mapper 77 Games released for the "Vs System" arcade cabinets (and not at home) Modern homebrew era: mapper 30 = UNROM512 (INL 4sc variant) mapper 111 = ...

Is the S-CPU well enough understood Probably? a new-production chip Unaffordable. You can't get custom silicon without an initial outlay on the order of $80k+ in a pin-compatible package I've seen clever things with castellated PCBs and mezzanine connectors, such as the UltraPIF or fancy flex PCBs,...