Controller Port Problem/Unwanted Autofire

[thenickman100]

No matter what game I try or which controller I use, my 7800 will now think I am rapidly pressing both of the fire buttons.  Is something wrong with the controller port?  I'm hoping whatever it is, it will be an easy fix, but I dread that I might need to crack it open and operate on the circuit board.

Edited September 20, 2019 by thenickman100

[boggis the cat]

Does it do this with no controllers attached?

[thenickman100]

1 hour ago, boggis the cat said:

Does it do this with no controllers attached?

No. Just opens up the intro screen for the game.

[+SmittyB]

What sort of controllers are you using; official 7800 ones?

[boggis the cat]

1 hour ago, thenickman100 said:

No. Just opens up the intro screen for the game.

If you have a two-player game then try connecting a joystick to only port 2.  If port 2 functions correctly then it’s likely a fault on the board.  If port 2 also misbehaves then it might be a bad RIOT chip.

 

If it’s the chip and your 7800 has socketed chips then that’s fairly easy to fix, provided you can get your hands on another (search for “MOS 6532”).  If it’s a board fault then you’d be best to let someone look at it who is comfortable troubleshooting these old systems.

Edited September 21, 2019 by boggis the cat
Clarification.

[+DrVenkman]

19 minutes ago, boggis the cat said:

If you have a two-player game then try connecting a joystick to only port 2.  If port 2 functions correctly then it’s likely a fault on the board.  If port 2 also misbehaves then it might be a bad RIOT chip.

 

If it’s the chip and your 7800 has socketed chips then that’s fairly easy to fix, provided you can get your hands on another (search for “MOS 6532”).  If it’s a board fault then you’d be best to let someone look at it who is comfortable troubleshooting these old systems.

 

RIOT handles directional signals on the controller ports; TIA handles the FIRE button signals, run through some resistors. There are 470pF capacitors on each joystick FIRE button as well. If one or both of those are shorted, that might be the cause of your issue. Or you might actually have a bad TIA chip. Check capacitors C33 and C34 for the Left and Right ports, respectively, then resistors R34 and R35 for the resistors on the same trigger lines. These lines lead to TIA pins 36 and 35, respectively.

 

Good luck.  

[thenickman100]

3 hours ago, SmittyB said:

What sort of controllers are you using; official 7800 ones?

I have tried it with the original 7800 controllers and with Sega Genesis controllers.  Same effect occurs.

[thenickman100]

3 hours ago, DrVenkman said:

 

RIOT handles directional signals on the controller ports; TIA handles the FIRE button signals, run through some resistors. There are 470pF capacitors on each joystick FIRE button as well. If one or both of those are shorted, that might be the cause of your issue. Or you might actually have a bad TIA chip. Check capacitors C33 and C34 for the Left and Right ports, respectively, then resistors R34 and R35 for the resistors on the same trigger lines. These lines lead to TIA pins 36 and 35, respectively.

 

Good luck.  

So I tried combat with player 2's controller port, and the same effect occurs (just with player 2 now).  The problem is definitely with the Atari and not my controllers.  Should I replace my TIA chip?  What exactly do I check R34, R35, C33, and C34 for?  I have a multimeter.

[thenickman100]

I checked R34, R35, C33, and C34 with a multimeter, and they all seem to be functional.

[boggis the cat]

9 hours ago, thenickman100 said:

I checked R34, R35, C33, and C34 with a multimeter, and they all seem to be functional.

Did you check continuity through to the TIA chip?  (It is the TIA as DrVenkman pointed out, not they RIOT as I incorrectly suggested.)

 

TIAs are harder to get hold of, and you will need the correct version (they are for NTSC or PAL, and the pinouts differ), but it’s the same approach.  If you have socketed chips then it’s fairly simple — if they’re soldered, then it’s a PITA.

 

Once you know the culprit, perhaps you can put up a post on the buy/sell to see if anyone nearby has a suitable spare.

[thenickman100]

1 minute ago, boggis the cat said:

Did you check continuity through to the TIA chip?  (It is the TIA as DrVenkman pointed out, not they RIOT as I incorrectly suggested.)

 

TIAs are harder to get hold of, and you will need the correct version (they are for NTSC or PAL, and the pinouts differ), but it’s the same approach.  If you have socketed chips then it’s fairly simple — if they’re soldered, then it’s a PITA.

 

Once you know the culprit, perhaps you can put up a post on the buy/sell to see if anyone nearby has a suitable spare.

How do I check "continuity"? 

I did a little bit of googling earlier, and saw that 2600s use TIAs too.  Are the two chips interchangeable as long as they are both NTSCs?

 

Also, I really appreciate all of the helpful advice so far

[+-^CrossBow^-]

10 hours ago, thenickman100 said:

How do I check "continuity"? 

I did a little bit of googling earlier, and saw that 2600s use TIAs too.  Are the two chips interchangeable as long as they are both NTSCs?

 

Also, I really appreciate all of the helpful advice so far

Most multimeters feature a continuity mode. Usually shared with the diode mode on a few or at least it always was with the meters I've owned. So yeah...continuity is the beep mode also. 

 

Your other question about the TIA in 2600s. Yes, you can use those just fine. That is one of the main reasons the TIA and RIOT are in the 7800 was to maintain the 2600 compatibility in the console but then they also happen to use them for other functions for the 7800 as well. But yes, the TIA in both 2600s and 7800s are the same and it is junked 2600s that I salvage the TIAs and RIOTs from.

[thenickman100]

53 minutes ago, -^Cro§Bow^- said:

Most multimeters feature a continuity mode. Usually shared with the diode mode on a few or at least it always was with the meters I've owned. So yeah...continuity is the beep mode also. 

 

Your other question about the TIA in 2600s. Yes, you can use those just fine. That is one of the main reasons the TIA and RIOT are in the 7800 was to maintain the 2600 compatibility in the console but then they also happen to use them for other functions for the 7800 as well. But yes, the TIA in both 2600s and 7800s are the same and it is junked 2600s that I salvage the TIAs and RIOTs from.

I checked the TIA, RIOT, and MARIA chips for continuity.  To check them, I made contact on the top side of the pins--the red contact on the ground, and black contact on each of the other pins.  The TIA and RIOT gave a non-zero and non-beep for each pin.  The MARIA gave a 0 read-out on pin 5 and a beep on pins 25 and 46.  I don't really know anything about whether that indicates a problem with the MARIA.  I assume it is fine since that was not one of the chips that you all thought would be relevant.

MARIA pinout: https://www.atarimax.com/jindroush.atari.org/7800cmar.html

RIOT pinout: https://www.atarimax.com/jindroush.atari.org/2600crio.html

TIA pinout: https://www.atarimax.com/jindroush.atari.org/2600ctia.html

 

Just for good measure, I checked the bottom solder points for the TIA and RIOT and everything was non-zero and non-beep too.

[boggis the cat]

12 hours ago, thenickman100 said:

Also, I really appreciate all of the helpful advice so far

Just to add that you can search for the NTSC version of the TIA by the part number CO10444.  “UM6526” is the chip designation: but “UM6526P1” is the PAL version, which won’t work, and you often see PAL versions incorrectly labelled as just UM6526 on e.g. AliExpress.  (The PAL TIA is part number CO11903.)

 

So far as I know, all UM6532 chips (the “RIOT” chip) and UM6507 should work on either NTSC or PAL VCS / 2600s; but the 6502C in the 7800 is a custom 6502, and standard ones won’t work.  I’ve not seen the custom video chip from the 7800 for sale.  Not enough 7800s scrapped to have a pool of recycled chips, I guess.

[+DrVenkman]

8 minutes ago, thenickman100 said:

I checked the TIA, RIOT, and MARIA chips for continuity.  To check them, I made contact on the top side of the pins--the red contact on the ground, and black contact on each of the other pins.  The TIA and RIOT gave a non-zero and non-beep for each pin.  The MARIA gave a 0 read-out on pin 5 and a beep on pins 25 and 46.  I don't really know anything about whether that indicates a problem with the MARIA.  I assume it is fine since that was not one of the chips that you all thought would be relevant.

MARIA pinout: https://www.atarimax.com/jindroush.atari.org/7800cmar.html

RIOT pinout: https://www.atarimax.com/jindroush.atari.org/2600crio.html

TIA pinout: https://www.atarimax.com/jindroush.atari.org/2600ctia.html

 

Just for good measure, I checked the bottom solder points for the TIA and RIOT and everything was non-zero and non-beep too.

That’s not really checking continuity; you’ve merely verified that they’re not shorted to ground somewhere. To check continuity, start at the TIA pins suggested and then measure from the chip to the first component leg or via along the trace. If you get a beep, that indicates the signal path is intact and the trace isn’t bad or shorted. Then go from that component to the next, and so on until you reach the end of the signal chain - in this case, the pins where the joystick ports are soldered to the board. When you come across capacitors or resistors, just measure then individually. For a resistor, check the schematic for the correct value (the schematics are on AtariAge somewhere). Set your meter to resistance measurement and range (if your meter isn’t auto-ranging) and measure the resistance - if it’s within about 10% of the stated value it’s fine. For capacitors it’s a bit trickier - most meters can’t measure caps accurately in-circuit. But you can at least see if it’s shorted to ground.

 

 

[thenickman100]

3 minutes ago, DrVenkman said:

That’s not really checking continuity; you’ve merely verified that they’re not shorted to ground somewhere. To check continuity, start at the TIA pins suggested and then measure from the chip to the first component leg or via along the trace. If you get a beep, that indicates the signal path is intact and the trace isn’t bad or shorted. Then go from that component to the next, and so on until you reach the end of the signal chain - in this case, the pins where the joystick ports are soldered to the board. When you come across capacitors or resistors, just measure then individually. For a resistor, check the schematic for the correct value (the schematics are on AtariAge somewhere). Set your meter to resistance measurement and range (if your meter isn’t auto-ranging) and measure the resistance - if it’s within about 10% of the stated value it’s fine. For capacitors it’s a bit trickier - most meters can’t measure caps accurately in-circuit. But you can at least see if it’s shorted to ground.

 

 

I'm not quite sure how to do what you are describing.  Where should I put the red contact and where should I put the black contact?  

[thenickman100]

47 minutes ago, DrVenkman said:

That’s not really checking continuity; you’ve merely verified that they’re not shorted to ground somewhere. To check continuity, start at the TIA pins suggested and then measure from the chip to the first component leg or via along the trace. If you get a beep, that indicates the signal path is intact and the trace isn’t bad or shorted. Then go from that component to the next, and so on until you reach the end of the signal chain - in this case, the pins where the joystick ports are soldered to the board. When you come across capacitors or resistors, just measure then individually. For a resistor, check the schematic for the correct value (the schematics are on AtariAge somewhere). Set your meter to resistance measurement and range (if your meter isn’t auto-ranging) and measure the resistance - if it’s within about 10% of the stated value it’s fine. For capacitors it’s a bit trickier - most meters can’t measure caps accurately in-circuit. But you can at least see if it’s shorted to ground.

 

 

Edit: I have found schematics of controller ports and circuit board to follow where the wires lead on the board.  Am I trying to check continuity to pins #5 & 6 on the controller port from the TIA chip pin that supposedly should lead there?

Here is the diagram of the port that I found: 

 

[thenickman100]

2 hours ago, DrVenkman said:

That’s not really checking continuity; you’ve merely verified that they’re not shorted to ground somewhere. To check continuity, start at the TIA pins suggested and then measure from the chip to the first component leg or via along the trace. If you get a beep, that indicates the signal path is intact and the trace isn’t bad or shorted. Then go from that component to the next, and so on until you reach the end of the signal chain - in this case, the pins where the joystick ports are soldered to the board. When you come across capacitors or resistors, just measure then individually. For a resistor, check the schematic for the correct value (the schematics are on AtariAge somewhere). Set your meter to resistance measurement and range (if your meter isn’t auto-ranging) and measure the resistance - if it’s within about 10% of the stated value it’s fine. For capacitors it’s a bit trickier - most meters can’t measure caps accurately in-circuit. But you can at least see if it’s shorted to ground.

 

 

Using the schematics I found (in the link below), I followed all of the paths from pin #6 in the controller port for players 1 and 2.  I marked the paths green to indicate that I checked them for continuity and resistance, when applicable.  Everything appears to be okay?  Any leads on what I should check next?

https://atariage.com/7800/archives/schematics_ntsc/images/Schematic_Atari7800_NTSC_4000.jpg

[thenickman100]

I have a weird turn of events that do not really make sense.

->Removed TIA and pushed back into place

->Tested combat, game works perfectly now

->Tested Xevious, the game that I had tried playing when the problem first appeared

->The problem came back

->Tested Galaga, no problems

->Tested Xevious, problem came back

->No problem with any games except for Xevious

 

I had just purchased the copy of Xevious, and it seemed like a very obvious repro with how glossy the label was and the spring-loaded dust cover inside.  Is it common for repros to cause damage to Atari consoles?  I'm surprised that the problem was temporary the second time and did not become permanent after trying Xevious again.  What would removing and reinserting the TIA do to fix a problem like this?

[+Mitch]

For Xevious try checking your difficulty switch.

 

Mitch

[+-^CrossBow^-]

8 hours ago, Mitch said:

For Xevious try checking your difficulty switch.

 

Mitch

Does Xevious have an auto fire mode? I thought the switches only changed whether the buttons worked independently or would fire and drop bombs simultaneous with either button for single button use?

[thenickman100]

12 minutes ago, -^Cro§Bow^- said:

Does Xevious have an auto fire mode? I thought the switches only changed whether the buttons worked independently or would fire and drop bombs simultaneous with either button for single button use?

I do not think it has an autofire mode.  When controllers are connected, it skips right through the title screen, indicating that the fire buttons are "held down".

 

[+DrVenkman]

Xevious does not have an auto fire button but then difficulty switches control whether it will fire guns and bombs with a single button. Curious however - if you set the switch for separate firing, do both guns and bombs still fire at once without the buttons being held?

 

And lastly, have you confirmed this behavior with multiple controllers? 

[+Mitch]

I thought that I remembered that certain third party controllers auto-fired when the difficulty switch was set in one position. I don't remember which way caused the issue anymore.

 

Mitch

[Trebor]

Thread here mentions: 

 

"...when I start up Xevious it goes through the player select screen and immediately starts firing until I hit my buttons..."

 

It may be worth looking over.