Help with custom SNES case

Discussion of hardware and software development for Super NES and Super Famicom. See the SNESdev wiki for more information.

Moderator: Moderators

Forum rules
  • For making cartridges of your Super NES games, see Reproduction.
Post Reply
alfador
Posts: 5
Joined: Mon Feb 06, 2023 8:37 am

Help with custom SNES case

Post by alfador »

Hello good people,

I have been working on a custom rack mounted SNES case. I would like some suggestions/insight on how to diagnose and solve a hardware problem.

I built a cartridge slot "extender" to allow the cartridge to be inserted from the front of the case. One one end I have a custom PCB with a cartridge slot connector and a ribbon cable connector. The ribbon cable then runs to the SNES motherboard and plugs into the cartridge slot connector on the motherboard through another custom pcb which reproduces the pins on a regular SNES cartridge. I included all 62 pins to support cartridges with enhancement chips.

I have tested many games. Most of the games work without any noticeable problems. But some games have issues or will not start at all. Games that do not work seem to have in common that they use enhancement chips (with one exception)

Here are the games that do not work with the corresponding issues :

Pilotwings
Everything works in the menus until starting levels in Mode 7. The background then appears garbled and the game is unplayable (see pic), but does not freeze. The foreground sprites appear ok. The game responds erratically to controller inputs. The game logic is affected (Ex. on the parachute level, the altitude will go down for a while, then starts going up, then down again)

Super Mario Kart
Very similar issue as Pilotwings. Everything works until the race starts. The mode 7 background is garbled (see pic). Foreground sprites are ok. Game is unplayable but does not freezes.

This seems to be an issue with DSP-1 games. To confirm, I purchased Michael Andretti's Indy Car Challenge, and the game indeed suffers from the same issues as Pilotwings and SMK.

Super Mario World 2: Yoshi's Island
Game starts ok and is playable until you hit the first expansion block. Then the games freezes (my guess is that the games calls the Super FX GSU-2 at this point). (Note on the Super FX chip : Starfox 1 works perfectly. No issues at all)

Rockman X2 (Super Famicom version)
Black screen. Game does not load.

Tetris Attack
Tetris Attack has issues with colors. The game is playable, the colors are incorrect and the graphics are simetime garbled (see pic). Also, the defects in the picture vary each time you start a game.

Here is a quick list of games I have confirmed work without any noticeable problems :
- Starfox
- Super Mario World
- Super Bomberman
- Zelda LTTP
- Super Metroid
- Megaman X
- F-Zero
- Mario Paint
- NBA JAM
- NHL 94
- The Adams Family
- Out of this world
- Prince of Persia
- Street Fighter 2 Turbo
- Super Street Fighter
- Killer Instinct
- Donkey Kong Country
- Donkey Kong Country 2
- Star Trek : the next generation
- Star Wars - The return of the Jedi
- Populous
- Sim City
- Desert Strike

Tests I have done so far :
- Check continuity between all pins on the ribbon cable connector. RESULT : All pins have continuity and all pins go where they should.
- Check soldering job. RESULT : Looks ok. Maybe some pins could be retouched. But continuity is good on all pins.
- Thorough cleanup of the game cartridge. RESULT : Problem remains.
- Bypass the ribbon cable connector. RESULT : all games work perfect when plugged directly in the SNES motherboard.
- Try another motherboard. RESULT : I tested my ribbon connector with an Analogue Super NT. The results are exactly the same with all games.

I would welcome suggestion on what to test next.

Here are my thoughts/questions about the next steps :

- Loss of signal in the ribbon cable?
- How can I test beyond continuity?
- How can I test signal quality/integrity?
- Analyse signal on pins with Oscilloscope? What should I look for? For a specific pin, what does a "good" signal look like?
- Since most game work and the problem seems to be related to enhancement chips, there might be some specific pins that have issues? How can I identify the pins I should focus on?

Any help would be appreciated!

Thanks!
Attachments
pilotwings.jpg
smk.jpg
TA3.jpg
TA2.jpg
TA1.jpg
User avatar
TmEE
Posts: 960
Joined: Wed Feb 13, 2008 9:10 am
Location: Norway (50 and 60Hz compatible :P)
Contact:

Re: Help with custom SNES case

Post by TmEE »

This is definitely a signal integrity issue, good enough oscilloscope will be able to show it.

How long is that extender ? You likely need to actually shield your cables, i.e grounded aluminium or copper foil around the ribbon cable. You may need to add fat GND and VCC wires and also some bulk capacitance at the cartridge end to help with the voltage drop from the cables and counter some of the excess inductance that the long cable adds.
alfador
Posts: 5
Joined: Mon Feb 06, 2023 8:37 am

Re: Help with custom SNES case

Post by alfador »

TmEE wrote: Thu Feb 23, 2023 10:08 am This is definitely a signal integrity issue, good enough oscilloscope will be able to show it.

How long is that extender ? You likely need to actually shield your cables, i.e grounded aluminium or copper foil around the ribbon cable. You may need to add fat GND and VCC wires and also some bulk capacitance at the cartridge end to help with the voltage drop from the cables and counter some of the excess inductance that the long cable adds.
The extender is about 20 in long. 28 AWG.

I was thinking about getting an oscilloscope. I guess this is the perfect opportunity. :-)

That makes me realize I haven't checked voltage level on VCC on the Cartridge side. I'll start there.

Thank you.
User avatar
TmEE
Posts: 960
Joined: Wed Feb 13, 2008 9:10 am
Location: Norway (50 and 60Hz compatible :P)
Contact:

Re: Help with custom SNES case

Post by TmEE »

20" is almost 51cm, that is a very long cable. You absolutely need shielding then and even then it may not work in the end. This amount of cable will round off signal edges quite a lot and there will definitely be power issues assuming it is standard ribbon cable seen in old PCs, those skimpy wires are not up to the task at this distance. You definitely need some bulk capacitance at the end of such a cable, few hundred µF may be enough.
alfador
Posts: 5
Joined: Mon Feb 06, 2023 8:37 am

Re: Help with custom SNES case

Post by alfador »

I just checked checked voltage at the cartridge slot and I get 5.03V (I tested without a cartridge in the slot)

The bulk capacitance purpose would be to stabilize voltage coming from the console?

I'll try to make some tests with a shorter ribbon cable. I could eventually modify my design and move the motherboard closer to the cartridge and use a shorter cable.

Also Maybe try to find a beefier ribbon cable.
User avatar
TmEE
Posts: 960
Joined: Wed Feb 13, 2008 9:10 am
Location: Norway (50 and 60Hz compatible :P)
Contact:

Re: Help with custom SNES case

Post by TmEE »

Multimeter is unable to show the small wiggles that happen when signals change state, capacitor is able to help provide the instantaneous current needs. Oscilloscope is able to show a whole new world lol
alfador
Posts: 5
Joined: Mon Feb 06, 2023 8:37 am

Re: Help with custom SNES case

Post by alfador »

Thanks for the info. Any idea why cartridges with enhancement chips seem to be having more issues?
lidnariq
Posts: 11432
Joined: Sun Apr 13, 2008 11:12 am

Re: Help with custom SNES case

Post by lidnariq »

Their power draw is much higher (and more variable), so it's likely that there's invisible brown-out.
alfador
Posts: 5
Joined: Mon Feb 06, 2023 8:37 am

Re: Help with custom SNES case

Post by alfador »

Update on my investigation :
  • Tried shorter ribbon cable (6 in). Problems still present.
  • Tried FFC cable instead of ribbon cable. Problems still present.
  • Tried shielding ribbon and FFC cables by tightly wraping them in grounded aluminum foil. Problems still present.
  • Very interesting observation : I tested all the problematic games with the FX Pak Pro and everything works perfectly. That leads me to believe that this is not a cable-signal-integrity issue, but something more along the lines of the cartridges not getting enough power? Or clock issue perhaps?
  • Tried to add bulk capacitance in parallel to the 5v supply on the cartridge, but the console will not boot when the caps are present. Is there something I didn't understand correctly here?
  • I tried hooking the cartridge directly to a bench supply to provide stable 5V. Problems still present.
  • I got myself an oscilloscope. I probed pins in both configurations (with extension cable and direct connection) but I haven't noticed significant differences between the different observed value so far. I was mainly looking at noise (Peak-to-Peak voltage variations).
For my next step I was thinking of systematically probing each pins and building up a file to compare all the values for each configuration.

Any suggestions would be welcome!
User avatar
Memblers
Site Admin
Posts: 4044
Joined: Mon Sep 20, 2004 6:04 am
Location: Indianapolis
Contact:

Re: Help with custom SNES case

Post by Memblers »

FX Pak Pro has series resistors, those could be terminators for the reflections on that long run. Also I'd imagine the output of the '245 bus transceivers on there would look different on the scope, compared to a normal cart.
Post Reply