rotary switch (not soldered on yet) or the 10-pin header can go to my LCD selector or a remote mounted rotary or dip selector
twistedsymphony's adventures in making PCBs.
- Thread starter twistedsymphony
- Start date
yes
I do want to stress that I'm not expecting to make this particular board into a product. This is just a test board, but if this works out, I have an idea for an updated version of this that could fit a lot more games on a single cart
So despite some mild success I've run into an issue where my implementation of the modules is producing some minor data corruption, I've been banging my head against this for the last few weeks and run dry of ideas to try to fix it. There seems to be some kind of timing issue with the voltage translation buffers that causes the occasional, seemingly random bit-flip.
For now I'm stepping away from this to work on some other projects which means NAOMI work and by Extension the Seta multi are on hold for the time being until I get some new ideas on how to fix this problem.
For now I'm stepping away from this to work on some other projects which means NAOMI work and by Extension the Seta multi are on hold for the time being until I get some new ideas on how to fix this problem.
Ape Cosplay
Grand Master
Do you mean the data on the flashes get corrupted? How do you handle the WP and RST signals ?
Regarding LVC245s the datasheet recommends using a pull-up resistor to Vcc if used in forced A to B translation.
Regarding LVC245s the datasheet recommends using a pull-up resistor to Vcc if used in forced A to B translation.
data does not get corrupted. The Modules themselves I think are fine, I'm using DS's original design unmodified and can write to them and read from them using a PROMAN and NOR flash program with 100% success.
the problem is that once that module is is on my NAOMI PCB I'm trying to read from in my programmer (Xeltek 580U) AND on the NAOMI itself through the ROM test mode. and I can dump 4 times in a row and I get different data every dump. In both cases I'm dumping as a 27c322
out of the 32Mbit of Space I'm trying to dump maybe only 20 or so bytes are wrong, but it's a different 20 bytes nearly every time
those are disconnected on my PCB since I'm not using it to write, but the module itself pulls them up to 3V3.
I didn't catch that, that's an easy fix but I don't think it explains the behaviour I'm seeing (I'm also not seeing pull ups on some of the other multi's that are using this chips
)
Was chatting on Discord with some mates and we all agreed that there needs to be a better solution for the SSV (with all the games).
There’s yet this other one now with another 4 games including Twin Eagle 2 but no Storm Blade.
Just throwing this out there for consideration @twistedsymphony
There’s yet this other one now with another 4 games including Twin Eagle 2 but no Storm Blade.
Just throwing this out there for consideration @twistedsymphony
PASTBLASTER
Champion
^This would be cool. I have the ebay 4 in 1 and while it works, it's not great.
I have 90% of the research done on this one because I worked out conversions for nearly every joystick game on the platform. the problem is that it uses a BOAT LOAD of Graphics ROMs, hence why the existing multi is just a 4 in 1
Storm Blade and Drift out 94 use an extra Math CO-Processor on the ROM board that would need to be implemented, probably why it's not been added to the 4 in 1.
most of the other joystick games don't have anything special. though some of them (like Monster Slider) include a counter chip connected to a header that seems to imply support for a spinner or something similar.
for those interested I have a sheet that breaks down the hardware config for all of SSV games: https://docs.google.com/spreadsheets/d/1wmQPw5pdKulfY4E5g_i6TAsOWyV_jt-WN1agtq4dbTQ/edit?usp=sharing
Last edited:
looking through the datasheet I'm not seeing a pull-up resistor recommendation:
https://www.ti.com/lit/ds/symlink/sn74lvc4245a.pdf
is this maybe recommended on a different brand/model LVC4245
Last edited:
ShootTheCore
Legendary
The Math Co-Processor is the Seta ST010, which Seta also used for the Super Famicom game Exhaust Heat 2 / F1 ROC II: Race of Champions. Supposedly, decapping revealed that it's a rebranded NEC uPD96050 so parts could possibly be sourced.
https://wiki.superfamicom.org/st010
EDIT: The SNES game is $10 - $15 on Ebay so a potential Multi could also just have a socket for it and let procurement of a donor ST010 be left to the end user.
EDIT: Confirmed that the MAME code for the SSV games that use the ST010 is the identical code used in SNES emulation of Exhaust Heat 2.
Last edited:
RepairGuy95
Champion
Was just about to say, I recognize that chip #
Excellent work @ShootTheCore
Excellent work @ShootTheCore
Ape Cosplay
Grand Master
WP is active low, therefore must be tied to ground for write protection.
That's the datasheet from TI I was refering to.
There's also this regarding OE:
Not sure how critical that is, just throwing ideas.
I'm using 74LVC16T245 myself and I learned the control signals must follow the levels on the A side.
I did know about that. What I don't know is if the ST010 is pin-compatible with the NEC uPD96050 or if there are differences that needs to be taken into account.
I think we're talking slightly different devices.WP is active low, therefore must be tied to ground for write protection.
That's the datasheet from TI I was refering to.
There's also this regarding OE:
Not sure how critical that is, just throwing ideas.
I'm using 74LVC16T245 myself and I learned the control signals must follow the levels on the A side.
I'm using the 4245 not the 245, the datasheet I linked doesn't even include a section 10.2 and only has a single OE pin, not 2 of them.