News: 11 March 2016 - Forum Rules

Author Topic: slidelljohn (a.k.a.[J]) snes projects page  (Read 55692 times)


  • Sr. Member
  • ****
  • Posts: 351
    • View Profile
Re: slidelljohn (a.k.a.[J]) snes projects page
« Reply #120 on: July 07, 2021, 05:58:04 pm »

Just popping in to mention that the latest development build of bsnes-plus now fully supports the 128kb/256kb VRAM expansion (currently only in "accuracy" builds, though this may change later...) Both the emulator itself and the built-in debugging tools can view and use all available expanded VRAM.

October 12, 2019, 05:40:27 pm - (Auto Merged - Double Posts are not allowed before 7 days.)
Also, since I've still got my head in the VRAM expansion stuff, could you clarify something?

In the GitHub issue a while ago, you mentioned that you were inverting both of the CPU I/O pins that the expansion uses - am I correct in understanding that writing all zeroes to $4201 will select the second 128kb VRAM bank?

If so, then I'm unfortunately finding a lot of games that break with this behavior, especially ones that include support for multitaps and possibly other peripherals, since they write 0x80 and 0x00 to the I/O port and sometimes do so after it's already started loading stuff into the other VRAM bank, causing some pretty serious visual glitches.

Another good example is Nintendo's burn-in test carts, since those also write the same values while the screen and VRAM contents are already in use:

e: made a more detailed (but probably not complete) list of games that have obvious issues with my 256k VRAM emulation:

Most of the offending games seem to involve multitap support, some of them behave worse with the 5 player multitap plugged into port 2.

I updated the burn-in test cart to check for the extra vram. It currently only checks for 64kb and 128kb vram sizes. The other sizes will be added at a later date. It also tests the extra vram to make sure its storing the bytes correctly. Here is the patch for it:

Im removing the I/O port feature that enables/disables the mod chip. The new rom chip will be activated by the switch on the back of the console. The I/O port feature that enables/disables the mod chip was suppose to help certain games work correctly like super Mario World 2 Yoshi's Island but there is still just too many games not working correctly. As soon as I update the mod chip Ill upload the new schematics with a part list so anyone would be able to build it if they would like to. I plan on building a custom test cart that shows off the features of the extra vram.
« Last Edit: July 07, 2021, 06:07:27 pm by slidelljohn »


  • Full Member
  • ***
  • Posts: 206
    • View Profile
Re: slidelljohn (a.k.a.[J]) snes projects page
« Reply #121 on: July 09, 2021, 07:38:10 pm »
Will the newer mod still use the I/O port for bankswitching larger amounts of VRAM?

I can update the implementation in bsnes-plus, just let me know how it should work.


  • Sr. Member
  • ****
  • Posts: 351
    • View Profile
Re: slidelljohn (a.k.a.[J]) snes projects page
« Reply #122 on: July 12, 2021, 01:07:29 pm »
Yes, the I/O port will still be used for bank switching the different sets of 128kb vram. Im going
to make it to where the switch on the back of the console enables the modchip, the I/O ports will
switch between the 128kb sets of vram, but Im also going to have another switch on the modchip itself
that can disable the I/O ports if only the expanded 128kb vram is to be used. I think channel 3 will
disable the modchip and channel 4 will enable it. Now how the modchip is currently set up I believe
the I/O ports have a not gate on them because of trying to get games like yoshi's island to work
without having a switch. These not gates will be removed on the new modchip. Im also going to make a
new pcb for the controller connector so I can put a dual color (red/blue) led on it so when its red
it will be like the original console and when its blue it will have the modchip activated.

Let me know what you think about the new setup for the modchip.

Also the patch I posted doesn't use the I/O ports its just for 128kb vram testing. The hidden feature
in higan (tested on v100 - v106) to activate the 128kb should display the 64kb and 128kb correctly.
The patch was made to test the switch on the back of the console. The switch was wired to cpu pin 19
position on the modchip and cpu pin 20 was disconnected on the modchip to test it and it works perfectly.
Here is a picture of where I connected to the RF board on the snes:

When I get the new modchip made I would like to send you one of the consoles (for free) for research purposes
if you would like one to run test on. I still have to design the new pcb's and run test on them to make sure they
work properly so its probably going to be a couple of months before its ready.