Capacitor c55 issue

[Rastamafugg]

Just received a 7800 that seems to play 2600 games ok, but I only got a 7800 cartridge to work well enough to play a game once.  When I examined the board, I noticed that the leg of the capacitor labeled c55 wasn't connected to the board. I'm planning on reconnecting it, but seeing as this is my first time using a 7800, I figured it's a good idea to check here first, to make sure it's not just a factory bodge.  I'm going to do a second cleaning of the connectors first, but after a post-clean test, that's the next step as far as I can see.

[+-^CrossBow^-]

Are you sure it was labeled as C55? Can you take a picture of it for us?

 

 

[Rastamafugg]

5 hours ago, -^CrossBow^- said:

Are you sure it was labeled as C55? Can you take a picture of it for us?

 

 

It's a little difficult to get a good picture, but here are a couple of angles

[+-^CrossBow^-]

That is an NTSC model yes?

 

I can tell you it isn't normal for that cap to be undone. Heck that entire section looks to have been redone based on the soldering? Even an odd section of solder at the bottom lead of the FB just before the main filter cap?

 

That cap should be attached.

 

[marauder666]

Some of the resistors have been changed for 0.5W ones, look to be the correct value though.  Don't see any reason for it.

I'd clean up that area and make sure they arent any bad connections.  That part of the circuit is related to the soft power up.

 

[+-^CrossBow^-]

4 hours ago, marauder666 said:

Some of the resistors have been changed for 0.5W ones, look to be the correct value though.  Don't see any reason for it.

I'd clean up that area and make sure they arent any bad connections.  That part of the circuit is related to the soft power up.

 

My theory on this is that a power mod was attempted in the past on this console and the wrong polarity was used and blew some of the components. I'm stating that because of the solder blob on the ground trace and at the base of the FB. That is actually where I solder my wiring for power mods as well or have in the past. I used to attach pigtail cables from those points that had a standard barrel jack on them so you could use a Genesis power supply. Possible someone did something similar and the wrong PSU was plugged in. Just a guess though.

 

[+Stephen Moss]

-^CrossBow^- is correct that cap should be attached, the unattached end should connect to pin 4 of the 4013. So if you have trouble getting it into its original hole you could flip it over the 4013 and attach it to R47 (Brown, Black, Green, between the two caps) which also goes pin 4.  

 

When switching on the connected power unit C55 couples a transient pulse to pin 4, resetting the latch setting the Q output low, preventing Q9 and thus Q10 from conducting and the incoming power for getting to the voltage regulator setting the unit to it off state.

However, if you have a multi-meter it may be worth checking it first. Although its non attachment may have been an oversight, one reason for leaving it unattached is that it has gone short circuit. If that were the case the latch would be held in permanent reset and you could never turn the unit on.

 

The unit should still work without C55 being attached, however when the power unit is switched on without the reset pulse the Q output of the latch could randomly be high or low and so you may be plugging in the first cartridge with the unit in its on state. 

  

[Rastamafugg]

20 hours ago, -^CrossBow^- said:

My theory on this is that a power mod was attempted in the past on this console and the wrong polarity was used and blew some of the components. I'm stating that because of the solder blob on the ground trace and at the base of the FB. That is actually where I solder my wiring for power mods as well or have in the past. I used to attach pigtail cables from those points that had a standard barrel jack on them so you could use a Genesis power supply. Possible someone did something similar and the wrong PSU was plugged in. Just a guess though.

 

It's a good guess.  Someone was definitely in this machine before me and that would explain the solder work.  Unfortunately, I'm new to owning this machine, so I don't know for sure.  It was working when my sister picked it up for me, so I'm hoping the fix is relatively simple.  I also got a 5200 this weekend, complete with 4 controllers and a trackball, and was playing with that instead!  I'll try to repair the 7800 later this week.

4 hours ago, Stephen Moss said:

-^CrossBow^- is correct that cap should be attached, the unattached end should connect to pin 4 of the 4013. So if you have trouble getting it into its original hole you could flip it over the 4013 and attach it to R47 (Brown, Black, Green, between the two caps) which also goes pin 4.  

 

When switching on the connected power unit C55 couples a transient pulse to pin 4, resetting the latch setting the Q output low, preventing Q9 and thus Q10 from conducting and the incoming power for getting to the voltage regulator setting the unit to it off state.

However, if you have a multi-meter it may be worth checking it first. Although its non attachment may have been an oversight, one reason for leaving it unattached is that it has gone short circuit. If that were the case the latch would be held in permanent reset and you could never turn the unit on.

 

The unit should still work without C55 being attached, however when the power unit is switched on without the reset pulse the Q output of the latch could randomly be high or low and so you may be plugging in the first cartridge with the unit in its on state. 

  

Thanks for the details!  I'll test it out with my multimeter before I turn it on again.

[Rastamafugg]

So I've reattached the leg and tested around with my multimeter before turning it on.  Everything seemed fine, but when I powered it on I got the same results.  2600 cartridges work fine, 7800 ones don't.  Examining the board some more, I think this has been recapped at some point, and one chip looks to have been replaced/resoldered.  I'm attaching picks of the parts of the board that looked worked on.  Any further troubleshooting suggestions are welcome!

 

 

[Rastamafugg]

One more addition: 

On 10/4/2021 at 4:21 AM, Stephen Moss said:

-^CrossBow^- is correct that cap should be attached, the unattached end should connect to pin 4 of the 4013. So if you have trouble getting it into its original hole you could flip it over the 4013 and attach it to R47 (Brown, Black, Green, between the two caps) which also goes pin 4.  

 

When switching on the connected power unit C55 couples a transient pulse to pin 4, resetting the latch setting the Q output low, preventing Q9 and thus Q10 from conducting and the incoming power for getting to the voltage regulator setting the unit to it off state.

However, if you have a multi-meter it may be worth checking it first. Although its non attachment may have been an oversight, one reason for leaving it unattached is that it has gone short circuit. If that were the case the latch would be held in permanent reset and you could never turn the unit on.

 

The unit should still work without C55 being attached, however when the power unit is switched on without the reset pulse the Q output of the latch could randomly be high or low and so you may be plugging in the first cartridge with the unit in its on state. 

  

I think this chip has also been replaced.  Is this the 4013 chip?  The chip is labelled a TI 33A41D8 (The 8 may be a B, it's hard to tell). I can't find any info online about the chip, and I don't see any chip on the board with 4013 in the name.  I see the 4013 in the schematics, but I find it hard to map schematic details to the pcb.  Couldn't find a picture of a de-populated 7800 board online either.  That means I can't easily figure out where R47 is located on the board, either. 

[+-^CrossBow^-]

Here you are:

 

[Rastamafugg]

Ok, I tested and I have continuity between that end of the capacitor and both R47 and pin 4 of the 4013.  Not sure what to try next.

[+-^CrossBow^-]

13 hours ago, Rastamafugg said:

Ok, I tested and I have continuity between that end of the capacitor and both R47 and pin 4 of the 4013.  Not sure what to try next.

You already stated earlier that you are able to play 2600 games so the issue isn't the system not power up. I assume you are just getting a black screen when trying to play 7800 games? Do you even see the atari rainbow loader when trying to start up with a 7800 game?

 

BTW...where are you located roughly?

 

[Rastamafugg]

8 hours ago, -^CrossBow^- said:

You already stated earlier that you are able to play 2600 games so the issue isn't the system not power up. I assume you are just getting a black screen when trying to play 7800 games? Do you even see the atari rainbow loader when trying to start up with a 7800 game?

 

BTW...where are you located roughly?

 

I was able to get one 7800 game to play long enough for a short game, and I think I got another to display the rainbow loader, but then it went to black screen.  Everything else has been black screen.  I do have a DragonFly, although I haven't been successful in getting it to load a game, either.  I believe there are some utilities that I might be able to use, if I could get them to load.  I've just been cautious with cartridges until I know the system won't cause any damage, given I don't know what is wrong.

 

I live in the Maritimes on the east coast of Canada.

[+Stephen Moss]

I am sure someone will correct me if I am wrong but if I recall correctly when a 2600 game is detected Maria is disabled and TIA enabled, when a 7800 game is detected it is the opposite way round. 

It is possible there is a problem with 7800 cartridge recognition and/or the TIA/MARIA selection which is resulting in your blank screen but I don't really know enough about who that works to suggest a good test procedure.

 

[+-^CrossBow^-]

5 hours ago, Stephen Moss said:

I am sure someone will correct me if I am wrong but if I recall correctly when a 2600 game is detected Maria is disabled and TIA enabled, when a 7800 game is detected it is the opposite way round. 

It is possible there is a problem with 7800 cartridge recognition and/or the TIA/MARIA selection which is resulting in your blank screen but I don't really know enough about who that works to suggest a good test procedure.

 

Pretty close although the TIA is still in use for controller fire button inputs and likely some other logics when 7800 games are played. 

 

@Rastamafugg have you closely examined the cartridge port? Specifically the extra sets of pins on the right and left?

[Rastamafugg]

@-^CrossBow^- I did take a card with some ipa-wet paper and put it into the slot and used Q-tips to clean the visible parts of the pins.  Not much room to get deep into the extra 7800 pin at the edges, so it might still be a corrosion issue, I suppose.  Although I tried a few cartridges last night and when I put a 7800 cart in, it would stop powering on at all (even after going back to a known working 2600 cart), which means something is shorting out, doesn't it?  I tried again this morning to similar results, so maybe reconnecting that cap is causing the short.  Wish I had an oscilloscope, or even a logic probe, I suppose.  All I have is a multimeter for testing.

[+Stephen Moss]

17 hours ago, -^CrossBow^- said:

Pretty close although the TIA is still in use for controller fire button inputs and likely some other logics when 7800 games are played. 

Thanks for the clarification, I probably could have been a little clearer in my post.

 

What I was trying to say although probably in an over generalised way was that (again I may be wrong but) I believe that the TIA is responsible for generating the video for 2600 games and MARIA for doing the same for 7800 games.

If that is the case then as the system reportedly plays 2600 games without issue it is probably safe to assume that elements common to both, i.e. Power On/Off latch, power supply, clock, 6502, 6532 and RF modulator are all OK. Therefore it would be logical to conclude that the problem is likely to lie with either MARIA, 7800 cart detection/7800 only cart connections or the logic that determines whether MARIA or TIA is responsible for video generation (maybe BIOS and U11/U12?).

In which case as the video output of both appear to be ORed together by U3 I would speculate that the video output for the TIA would be logic 0 if MARIA alone is generating the video and vice verser, therefore if there is no video output for 7800 games either something is stopping MARIA from outputting video signals or stopping them from reaching U3 thus causing the blank screen.  

 

11 hours ago, Rastamafugg said:

I tried again this morning to similar results, so maybe reconnecting that cap is causing the short.  Wish I had an oscilloscope, or even a logic probe, I suppose.  All I have is a multimeter for testing.

If there is a short on the power supply a muilti-meter is all you need. Try measuring the input voltage across the big 2200uF cap (C62?) and between the middle pin and the two outer pins of the 7805 voltage regulator, when...

a) The power is off

b The power in on 2600 game inserted and 

c) The Power on 7800 game inserted.

 

Also thinking about what I said above try measuring the voltage on pins 1 & 2, 4 & 5, 9 & 10 and 11 & 12 of U3 with a 7800 game inserted and a 2600 game inserted, values will either be 5V, 0V or an indeterminate intervening voltage as they change state between 5V and 0V for the video signals.

 

If you have any 7800 games that are known to auto run an attract screen with sound once powered up try one of those to see if the audio is also affected or if it is just video.

 

You could always build a logic probe if you really want one there, are various schematics on line.    

[Rastamafugg]

5 hours ago, Stephen Moss said:

Thanks for the clarification, I probably could have been a little clearer in my post.

 

What I was trying to say although probably in an over generalised way was that (again I may be wrong but) I believe that the TIA is responsible for generating the video for 2600 games and MARIA for doing the same for 7800 games.

If that is the case then as the system reportedly plays 2600 games without issue it is probably safe to assume that elements common to both, i.e. Power On/Off latch, power supply, clock, 6502, 6532 and RF modulator are all OK. Therefore it would be logical to conclude that the problem is likely to lie with either MARIA, 7800 cart detection/7800 only cart connections or the logic that determines whether MARIA or TIA is responsible for video generation (maybe BIOS and U11/U12?).

In which case as the video output of both appear to be ORed together by U3 I would speculate that the video output for the TIA would be logic 0 if MARIA alone is generating the video and vice verser, therefore if there is no video output for 7800 games either something is stopping MARIA from outputting video signals or stopping them from reaching U3 thus causing the blank screen.  

 

If there is a short on the power supply a muilti-meter is all you need. Try measuring the input voltage across the big 2200uF cap (C62?) and between the middle pin and the two outer pins of the 7805 voltage regulator, when...

a) The power is off

b The power in on 2600 game inserted and 

c) The Power on 7800 game inserted.

 

Also thinking about what I said above try measuring the voltage on pins 1 & 2, 4 & 5, 9 & 10 and 11 & 12 of U3 with a 7800 game inserted and a 2600 game inserted, values will either be 5V, 0V or an indeterminate intervening voltage as they change state between 5V and 0V for the video signals.

 

If you have any 7800 games that are known to auto run an attract screen with sound once powered up try one of those to see if the audio is also affected or if it is just video.

 

You could always build a logic probe if you really want one there, are various schematics on line.    

Thanks for the detail!  I'll try to get some voltage measurements and update here.  And I ordered a logic probe, so I should be able to do a little more in-depth debugging this weekend.

[+-^CrossBow^-]

5 hours ago, Stephen Moss said:

Thanks for the clarification, I probably could have been a little clearer in my post.

 

What I was trying to say although probably in an over generalised way was that (again I may be wrong but) I believe that the TIA is responsible for generating the video for 2600 games and MARIA for doing the same for 7800 games.

If that is the case then as the system reportedly plays 2600 games without issue it is probably safe to assume that elements common to both, i.e. Power On/Off latch, power supply, clock, 6502, 6532 and RF modulator are all OK. Therefore it would be logical to conclude that the problem is likely to lie with either MARIA, 7800 cart detection/7800 only cart connections or the logic that determines whether MARIA or TIA is responsible for video generation (maybe BIOS and U11/U12?).

In which case as the video output of both appear to be ORed together by U3 I would speculate that the video output for the TIA would be logic 0 if MARIA alone is generating the video and vice verser, therefore if there is no video output for 7800 games either something is stopping MARIA from outputting video signals or stopping them from reaching U3 thus causing the blank screen.  

This has been my thought on his issue as well. It is something specific to the 7800 side only since 2600 games are working we know that his CPU, RIOT, and TIA are not likely the issue.

 

[ChildOfCv]

On 10/11/2021 at 1:55 AM, Stephen Moss said:

I am sure someone will correct me if I am wrong but if I recall correctly when a 2600 game is detected Maria is disabled and TIA enabled, when a 7800 game is detected it is the opposite way round. 

It is possible there is a problem with 7800 cartridge recognition and/or the TIA/MARIA selection which is resulting in your blank screen but I don't really know enough about who that works to suggest a good test procedure.

 

It doesn't have any hardware to 'detect' a 7800 game.  It's that a 7800 game writes to certain registers to enable the MARIA chip and do a couple of other 7800-specific things.  Chips involved in that are the MARIA itself (it apparently stands in front of the TIA anyway and passes the chip-select when appropriate), a couple gates of U12 (74LS32), U11 (74LS174), and a gate of U4 (74LS08).

[+Mitch]

Is the resistor connected to a leg of one of the RAM chips? If so which leg?

 

Mitch

[+Stephen Moss]

12 hours ago, Mitch said:

Is the resistor connected to a leg of one of the RAM chips? If so which leg?

It might help if we know which resistor you are referring to as the only resister referred to so far is R47 which is part of the power on/off latch circuit. 

[+-^CrossBow^-]

6 hours ago, Stephen Moss said:

It might help if we know which resistor you are referring to as the only resister referred to so far is R47 which is part of the power on/off latch circuit. 

I think Mitch is asking about the one installed between the SONY RAM chips.

[+Mitch]

Sorry. Yes, I am referring to the one connected to one of the RAM chips. If there is one.

 

Mitch