Super Pro System controllers won't work only on some games

[MonsterSky]

Hello everyone, I have i weird situation to expose to you.

Both the hand controllers of a Super Pro full TutorVision system work only with some games and won't work at all with other games. If I plug the BurgerTime or Frogger cartridges I'm able to to both controllers flawless, instead if I plug NHL Hockey or Horse racing the controllers won't work.

Initially I tested the console with the same NHL Hockey cartridge and also with FW Diagnostic cart I recentely purchase and both worked fine. I tested both controller with the diagnostic cartridge and the disc and all the buttons worked fine. Fews days passed and now they present this weird behaviour.

Last, if I swap repeatly BurgerTime and FW Diagnostic cartriges both controllers always work with BurgerTimer and never work with FW Diagnostic.

This situation is driving me nuts, do you have any idea/suggestion to try to solve it?

Thank you!

Edited March 24, 2019 by MonsterSky

[mr_me]

Maybe the fifth pin (diagonal) is shorted in one of the controllers and burgertime/frogger ignores that signal. Try unplugging each controller to see if it helps with nhl or horse racing. Does horse racing still play the title screen music?

[MonsterSky]

Horse racing plays the title screen music both with the two controller connected and with only one connected (I tried first the left one connected then the right one connected).

Then I did the following test, all with no positive results:

• disconnected left controller, right one connected
  
• disconnected right controller, left one connected
  
• disconnected both controllers (a silly test I know but I wanted to try it anyway)
  
• connect the left controller to the right port (only this connected)
  
• connect the right controller to the left port (only this connected)
  

Later I'll check the continuity and the shorts of both cables as you suggested to me.

By the way I tried also to use the MTE Test cartridge v3 but it doesn't start at all, I think this should be normal on a TutorVision board.

[mr_me]

It would be strange for both controllers to have the same problem. The problem could be with the controller circuitry on the mainboard.

 

Edit: I would expect an mte cartridge to run but don't know for sure. Do you have more game cartridges to test. You have two 4-way games that work and two cartridges that use exec titlescreens that don't.

Edited March 25, 2019 by mr_me

[MonsterSky]

Yes, It sounds strange to me too. My Hope is find a short somewhere in the controller cables, otherwise I have to study the 74HCTLS257 on U14 and U15

Edit: Yes, I have more cartridge to test, I'll pick some of them randomly with both EXEC titlescreen and not.

Edited March 25, 2019 by MonsterSky

[MonsterSky]

Some little progress (maybe) achieved.

Tested both controller cables: no shorts on pins and continuity is ok.

Tried more games: Poker & Blackjack, Space armada, loco-motion, Tower of Doom, Bump 'n Jump, Chess, Pole Position. Only Pole Position worked fine, all disk directions worked fine.

So I passed to check the two groups of eight 47 kOhm resistors connected directly to each controller port, as stated in the Joe Zbiciak "TutorVision / INTV88 Reverse Engineering" document and all of them misured correctly 47 kOhm minus the R27 one, which misure 2.5 kOhm.

I'm not and electronic technician but I think that that resistor have to be replaced.

Before I start to desolder the bottom metal shield of the board do you have any advice to approach best and safe this operation? I'm confortable with solderer but this is a precius board to work on and I want to avoid stupid mistakes.

Thank you for your advices.

Edited March 25, 2019 by MonsterSky

[mr_me]

I wonder if the cartridges that don't work are dependent on exec routines that the cartridges that work are not.

[MonsterSky]

I wonder if the cartridges that don't work are dependent on exec routines that the cartridges that work are not.

 

Before this problem the NHL Hockey played fine, do you mean that now there could be a problem on the EXEC ROM also?

[+Lathe26]

The bottom half of the Tutorvision's EXEC is the entire Mattel EXEC. They just doubled the ROM size and added more code and splash screen data. Intellivision games that use the EXEC should be fine.

[mr_me]

With pole position, frogger, and burgertime; we don't know if the fifth disc signal (diagonal) is working. I would have thought burgertime is an exec based cartridge but don't know for sure. I'm just thinking of possible explanations; there could be others.

Edited March 25, 2019 by mr_me

[MonsterSky]

At the end I found some weird contacts between resistor R27 and pin 1 on both controller connectors.

 

So I desoldered the R27 resistor, checked it again and it misured the 47 kOhm, so the resistor is ok

 

Made a test soldering again the resistor on the board and... it worked!

 

So now the Intellivision has come be to full working order, luckily the controllers IC we're not broken so the defect probably was caused by some false contact solved when removing the RF shield and/or the resistor.

 

Inviato dal mio Mi A1 utilizzando Tapatalk

[intvnut]

IIRC, some games' debouncing logic doesn't like stuck pins. I believe it's the debounce logic in the EXEC itself.

 

In some of my programs, I do a "stuck bit" test to know what bits to ignore on each controller. I don't always bother, but it's probably a good idea.

 

Basically, it looks for what bits are stuck-at-0 at start, and ignores them in debounce and key-tap logic. If you ever see those bits return 1, remove them from the "ignore" set.

[MonsterSky]

In some of my programs, I do a "stuck bit" test to know what bits to ignore on each controller. I don't always bother, but it's probably a good idea.

 

Interesting to know. There could be stuck bits on the controller and this can be considered just "normal" and ignore them?

 

AFAIK the stuck bits always indicades hardware problems and can not be ignored, I would like to know more on this.

[intvnut]

 

Interesting to know. There could be stuck bits on the controller and this can be considered just "normal" and ignore them?

 

AFAIK the stuck bits always indicades hardware problems and can not be ignored, I would like to know more on this.

 

Stuck bits are indeed faulty hardware. But, if you have one bad controller and one good controller, and you can still play most games with the remaining good controller, what should a game do?

 

• Play anyway?
• Hang?
• Drop into a diagnostics routine that helps you diagnose and fix your unit?

 

I've sometimes chosen the first bullet, and sometimes chosen the second bullet. As these systems get older and crankier, I sometimes wonder whether we should start choosing the third bullet.

Edited March 29, 2019 by intvnut

[mr_me]

I don't think you should program a game to work around faulty hardware this way. If the program hangs than that is feedback that something is wrong. Of course, if it's a diagnostic program than it should be programmed not to hang.

 

So would all exec-based cartridges hang in a situation like this? I guess burgertime is a little different.

[intvnut]

So would all exec-based cartridges hang in a situation like this? I guess burgertime is a little different.

 

 

BurgerTime directly reads the controllers, at least for moving the chef around:

.

        MVI     .PSG0.lft_hand,R0               ; 6FAC   0280 01FF
        B       L_6FB2                          ; 6FAE   0200 0002

L_6FB0:
        MVI     .PSG0.rgt_hand,R0               ; 6FB0   0280 01FE
L_6FB2:
        ANDI    #$001F, R0                      ; 6FB2   03B8 001F
        CMP     G_01ED, R0                      ; 6FB4   0340 01ED
        BEQ     L_6FBC                          ; 6FB6   0204 0004

        MVO     R0,     G_01ED                  ; 6FB8   0240 01ED
        B       L_6FE6                          ; 6FBA   0200 002A

.

Not sure what it does for the action buttons. The ANDI above strips away the action button bits. I do see it set various EXEC dispatch tables for the controller in my quick, less than 5 minutes analysis.

 

 

So would all exec-based cartridges hang in a situation like this?

 

Not sure, actually. It may depend on which bit's stuck, too.