Repairing my VCS, help needed.

[vitoco]

18 minutes ago, ChildOfCv said:

You want a 9VDC, 500mA or better adapter.

My bad! You are right. I meant 9V DC for the required specs...  (9V AC was from the 1050's power adapter specs)

 

When I get one, I'll let you know if that solved the waves problem.

 

[vitoco]

I made some more experiments with power sources and always got waves. I thought that it depended on the type of the adapter how long it takes the waves to appear. But using a regulated adapter seemed to be the best choice. During these tests, I touched every component to check for the temperature trying to find hot ones, and suddently found that waves appeared when I pressed the side walls from the RF module: it seemed that something inside it is making some interference... mud?

 

So, I decided to start with the mod ASAP.

 

On 10/4/2020 at 8:59 AM, Cmherndon79 said:

This is the guide I used when I got started:

 

https://vintagegamingandmore.com/installation-guide-4switch/

 

It doesn't mention removing R222, R209 and C209 which I also do

I checked that guide and many other onesm and combined them into my own solution based on the components I coulf find. I couldn't find the RCA female connectors and the 820 Ohm resistor in my neighborhood, and I cannot move because we are partially locked down (covid).

 

I removed the whole RF module, all 3 caps, both resistors and the transistor. Unfortunately, a couple of vias were also removed in the process.

 

The .22 uF caps were big polyestirene type ones, and C242 fitted just besides the power connector, but I couldn't put C241 through the vias, so I soldered it directly to the traces in the back of the board.

 

Then, I made a small board for the AV mod including a 75 Ohm resistor . Instead of installing RCA connectors, I soldered an A/V cable directly to this board: instead of a coaxial cable, an A/V cable goes out of the case.

 

 

 

When connected with to the CRT TV and powered on with a game, I got nothing!!! Neither video nor audio. I checked the voltages in the regulator, I didn't find shorts, all the lines were connected, I checked for the welding in my board and the motherboard. I reconnected this spawn again to the TV and powered it on again, and it worked. The next time it didn't. It seems that its a problem with the cartridge port that I have to check later.

 

Well, when I could start the game, I wait for the waves... just in case!!!!  Of course none of them appeared on the CRT. Neither the snow. It was an almost clear image with a bit of garbage and artifacts. Then I moved to a Full HD smart TV, and I could play a game after I set up the AV input and aspect ratio. I tried to disable any image processing for the upscaling, but I couldn't find a setup for that.

 

Anyway, I'm happy!!!! Thanks to all of you...

 

Now I need to go for the Harmony Cart or the PlusCart. I never thought I could return to the fun of the VCS (and to program my own games).

 

But I have a clue: should the metalic shield be put again before closing the case?

 

[+Al_Nafuur]

1 hour ago, vitoco said:

Well, when I could start the game, I wait for the waves... just in case!!!!  Of course none of them appeared on the CRT. Neither the snow. It was an almost clear image with a bit of garbage and artifacts. Then I moved to a Full HD smart TV, and I could play a game after I set up the AV input and aspect ratio. I tried to disable any image processing for the upscaling, but I couldn't find a setup for that.

 

Anyway, I'm happy!!!! Thanks to all of you...

Congratulations

 

 

1 hour ago, vitoco said:

Now I need to go for the Harmony Cart or the PlusCart.

If your soldering iron is still hot, I can send you one of the last 4 DIY Kits for the PlusCart.

 

[vitoco]

1 hour ago, Al_Nafuur said:

If your soldering iron is still hot, I can send you one of the last 4 DIY Kits for the PlusCart.

That would be great!

 

I'll PM you... Thanks!

 

 

[vitoco]

I discovered that I lost my console's screws. When I opened it some years ago after I discovered the waves, I never put them back and stored them somewhere. I searched in every place I could think where I have retro stuff, unsuccessfully. Then, I tried every screw I found, including the ones from the paddles and joysticks, also without success. In the service manual they are not listed.

 

As I don't want to break the guides, can someone post a photo of any screw from a 4-switches VCS (2600A rev 13 preferred)??? Any spec is also welcome.

 

Thanks!

 

[Cmherndon79]

@vitoco  Took a picture.  Sorry, these are Imperial measurements in the picture, but that's the ruler I have on hand in my shop.  Then again, at the time these were built everything was measured in inches, so period accurate  haha.

 

The longer screws go into the holes near the front, the ones that go in straight.

 

The shorter screws go into the holes in the rear that go in at an angle.

[vitoco]

Wow, that was fast....

 

2 hours ago, Cmherndon79 said:

@vitoco  Took a picture.  Sorry, these are Imperial measurements in the picture, but that's the ruler I have on hand in my shop.  Then again, at the time these were built everything was measured in inches, so period accurate  haha.

 

The longer screws go into the holes near the front, the ones that go in straight.

 

The shorter screws go into the holes in the rear that go in at an angle.

Exactly... the thread spacing matches the marks in the ruler (1/16, i.e, 16 threads per inch), so those are Imperial. As we use the metric system, it might be hard to find exactly the same here. For instance, I found one screw with 16 threads per inch, but the diameter seems to be 3/20" and did not enter. So, VCS's screwas are slimmer, I guess 3/16" or 2/20"=1/10". BTW, why do they mix */16" and */10" measurementes in the same piece? I'm sorry, I'm not used to Imperial system and I don't know when to "simplify" fractions.

 

I measured one from the paddles: length=12/20"=6/10"=3/5", thread=1/20" (20 per inch), diameter=1/8" (mixed in the other way!). That's why they didn't fit in the console.

 

Thank you!

 

[Cmherndon79]

And, here are the screws shown on a metric ruler.

 

I'm not sure what size they are as I've never have to buy any.  Unfortunately, I get enough consoles that were poorly packed with damaged cases

 

We use a mismash of systems in the states and local hardware stores will usually have both, especially since building materials are always measured in imperial.  They have metric as well, because, well everything we buy new, even most manufactured in the US use metric fasteners now.  

 

IMO, very worst case you can get screws that are similar in length and if can't get a spot on match the diameter, get ones that are slighter more thick in diameter.  Use a drill bit for the new size screw and use a piece of painters, electrical tape etc to mark the maximum depth on the bit and drill pilot holes for the new screws.  There is enough material in the shaft to accommodate a screw that is fractionally larger than the original and in fact I have done that with larger screws in the past when I receive a case that the screw threads are stripped out on.

[vitoco]

It is strange that the heads of both pairs are different between sizes. Might it be that one of the pairs are not original ones?

On 10/11/2020 at 8:15 PM, Cmherndon79 said:

IMO, very worst case you can get screws that are similar in length and if can't get a spot on match the diameter, get ones that are slighter more thick in diameter.  Use a drill bit for the new size screw and use a piece of painters, electrical tape etc to mark the maximum depth on the bit and drill pilot holes for the new screws.  There is enough material in the shaft to accommodate a screw that is fractionally larger than the original and in fact I have done that with larger screws in the past when I receive a case that the screw threads are stripped out on.

I thought about this, but I'll wait a little more. I'm sure that mine will appear just after I drill the case!!!

 

Now, I have another problem. Some of my games start OK at power on, but some of them do not, or just sometimes. From 17 cartridges, 3 of them (Warlords, Math Gran Prix and Championsip Soccer) start with this screen:

 

(the upper border dances from side to side)

 

Other few games start with partial game bitmaps or solid P/M columns.

 

I thought that it could be some dirt in the cartridge or connector, but I drop that idea when I found that pattern (3 games with the same problem). Now I think that there is something wrong with the power switch, because when I turn it on by plugging the power adapter instead of using the switch, the games start almost all the times.

 

Could it be a power switch failure or that it still has some mud or dirt inside of it?

 

Thanks!!!

[GENZZA]

5 hours ago, vitoco said:

Now, I have another problem. Some of my games start OK at power on, but some of them do not, or just sometimes. From 17 cartridges, 3 of them (Warlords, Math Gran Prix and Championsip Soccer) start with this screen:

 

 

I think this is more likely to be the contacts of your cartridges or the cartridge slot on the Atari, try cleaning all the pins on the cartridges with alcohol. Atari cartridges rarely break, most of the time they just need some cleaning and that is all you don’t even need to take a part the cartridge just use some Q tips and contact cleaner or isopropyl alcohol and rub the golden pins on the edge connector of every faulty cartridge. You might need a flat blade screwdriver to push the release the sliding cover that is fairly easy there are videos on you tube how to do it if you struggle.

[ChildOfCv]

On 10/11/2020 at 3:41 PM, vitoco said:

Wow, that was fast....

 

Exactly... the thread spacing matches the marks in the ruler (1/16, i.e, 16 threads per inch), so those are Imperial. As we use the metric system, it might be hard to find exactly the same here. For instance, I found one screw with 16 threads per inch, but the diameter seems to be 3/20" and did not enter. So, VCS's screwas are slimmer, I guess 3/16" or 2/20"=1/10". BTW, why do they mix */16" and */10" measurementes in the same piece? I'm sorry, I'm not used to Imperial system and I don't know when to "simplify" fractions.

 

I measured one from the paddles: length=12/20"=6/10"=3/5", thread=1/20" (20 per inch), diameter=1/8" (mixed in the other way!). That's why they didn't fit in the console.

 

Thank you!

 

They can't be 16/inch.  That would be a #10 screw, which is too big.  #8 screws are 18 threads per inch, which would track closely with the 16th marks for a ways.  #6 screws are 20 threads per inch, which would lose sync far more quickly.  So I'd bet that it's #8 by 5/8" and #8 by 3/4".  The screws are self-tapping kind.  "Pan head" or "truss head" means that the head is larger, so avoid that.

Edited October 18, 2020 by ChildOfCv

[vitoco]

On 10/18/2020 at 4:29 PM, ChildOfCv said:

 So I'd bet that it's #8 by 5/8" and #8 by 3/4".

I found some screws labeled as #8 3/4 in its package at a local store and tried them. They didn't fit, and I guess that they actually are 4mm because of the metric system we use, and the labels are just to name them.

 

I'm still searching for the original ones in my stuff.

 

[ChildOfCv]

I suppose metric is possible, but I doubt it, simply because they were designed in America, manufactured by the millions, and it's simply cheaper to just use the same designs and hardware as much as possible.

 

I looked at a size table again, and it says that a "type A" self-tapping screw of #6 size has an 18-thread pitch too.  "type AB" has the finer threads (20 for #6 or 18 for #8) and won't work.  It also seems, though, that "type A" is being phased out.  You can certainly still find them at screw/bolt specialty shops though, even if the hardware store doesn't carry them.

[0078265317]

https://www.ebay.com/itm/264860934203

 

A little expensive.  But maybe keep the screws and sell the bottom.

[vitoco]

On 10/18/2020 at 3:46 AM, GENZZA said:

I think this is more likely to be the contacts of your cartridges or the cartridge slot on the Atari, try cleaning all the pins on the cartridges with alcohol. Atari cartridges rarely break, most of the time they just need some cleaning and that is all you don’t even need to take a part the cartridge just use some Q tips and contact cleaner or isopropyl alcohol and rub the golden pins on the edge connector of every faulty cartridge. You might need a flat blade screwdriver to push the release the sliding cover that is fairly easy there are videos on you tube how to do it if you struggle.

This is weird... Even when I cleaned the cartidge slot and the cartridges, nothing was solved, and as I had some problems with the PlusCart (it resets itself after some seconds at power up), it retuned to my mind that the 2600 board is still faulty. I also found that the mod is not displaying white but yellow.

 

It seemed that there was a power failure somewhere, the same that caused the waves with the RF module that I later removed when I changed it by the AV mod.

While checking voltages and reviewing traces, with a magnifying glass I found a bit of dirt between components, so I washed the board again using baking soda and vinegar.

 

Now some behaviours changed: the startup pattern changed for the cartridges that do not start, and one of them (Warlords) changed the color for one of the elements of the screen (when started using the power plug method), the upper score digits should be of the same colors of the players (yellow and blue) instead of red:

 

I took this picture before todays wash:

 

 

Another strange behavior is that even when I don't have paddles connected to the console, both players from the bottom reacted when I shook the 2600 board or pressed the cartridge over the slot.

 

What could be the reason for these behaviors? Might @Cmherndon79 know about this?

 

On 10/22/2020 at 9:23 PM, 0078265317 said:

https://www.ebay.com/itm/264860934203

A little expensive.  But maybe keep the screws and sell the bottom.

Now, I'm thinking about to buy a "new" one.

 

Edited November 4, 2020 by vitoco

[ChildOfCv]

Sounds like you need to press on local parts of the board in search of possible broken traces or lifted pads.  Don't forget that there are traces on both sides of the board.

[vitoco]

On 11/4/2020 at 6:36 PM, ChildOfCv said:

Sounds like you need to press on local parts of the board in search of possible broken traces or lifted pads.  Don't forget that there are traces on both sides of the board.

In the search of that, I removed the cartridge slot and unmonted it to check for the internal pads. As a result, I lost one of the vias, so I'll have to put a bridge at mounting time. But I haven't found a single broken trace. Anyway, tomorrow I'll wash the board again to remove the dirt (mud) that it was under the cartridge slot and also inside it, and let it dry at the sun...

 

BTW, that reminds me about this: when I did the last wash, some water remained inside the cartridge port, and I think that it was the reason why I got this:

On 11/4/2020 at 2:31 PM, vitoco said:

Another strange behavior is that even when I don't have paddles connected to the console, both players from the bottom reacted when I shook the 2600 board or pressed the cartridge over the slot.

and also this:

On 11/4/2020 at 2:31 PM, vitoco said:

the startup pattern changed for the cartridges that do not start, and one of them (Warlords) changed the color for one of the elements of the screen (when started using the power plug method), the upper score digits should be of the same colors of the players (yellow and blue) instead of red

 

Well I'll discover if I can reassemble the VCS and solve all this issues tomorrow.

 

But I found a thing that it is also disturbing me: I got yellow instead of white. Could this be something introduced by the AV mod?

 

[vitoco]

On 11/6/2020 at 9:51 PM, vitoco said:

I removed the cartridge slot and unmonted it to check for the internal pads. As a result, I lost one of the vias, so I'll have to put a bridge at mounting time. But I haven't found a single broken trace. Anyway, tomorrow I'll wash the board again to remove the dirt (mud) that it was under the cartridge slot and also inside it, and let it dry at the sun...

After a deep clean, I resoldered the cartidge port to the board and tested it... the result is ame as before: the cartridges that do not start keep showing the same pattern on screen, and the ones that run at power up, can be played as expected, and white color is still displayed as yellow. Well, at least it isn't worst than before.

 

My VCS was bought brand new in the US by those days, but in my country we use 220V instead of 110V, so I always used a 110V-220V autotransformer in the 80's. When I wanted to use the VCS in the recent years, I was using a 220V multi voltage adapter set in 9V (+), but as I found that the voltage was going down after some minutes of use, I changed it to a regulated power adapter and that solved the fall of the voltage. Few days ago, I found the original 110V power adapter of my VCS, so I tried it using an autotransformer. Result: nothing changes, except that the input voltage (with load) in the internal regulator it's 10.6V, but the output keeps in 4.87V, even after many minutes.

 

Also, the whites are still displayed as yellow/orange.

 

AAAAAAAAAaaaaaaaaaargh!

 

Some other idea for any of these problems? Tips?

 

Thanks in advance!
 

[Mr SQL]

On 11/6/2020 at 8:51 PM, vitoco said:

I got yellow instead of white. Could this be something introduced by the AV mod?

Yes an AV mod can change the display characteristics because the Atari leaks chroma which makes colors brighter and more vibrant looking.

This effect is even more pronounced on older Atari models:

https://en.wikipedia.org/wiki/Helmholtz–Kohlrausch_effect

 

AV mods generally have a cleaner signal so some colors may change or look washed out like on the 7800.

There is also a potentiometer to change the colors you may need to adjust.

[ChildOfCv]

8 hours ago, Mr SQL said:

Yes an AV mod can change the display characteristics because the Atari leaks chroma which makes colors brighter and more vibrant looking.

This effect is even more pronounced on older Atari models:

https://en.wikipedia.org/wiki/Helmholtz–Kohlrausch_effect

 

AV mods generally have a cleaner signal so some colors may change or look washed out like on the 7800.

There is also a potentiometer to change the colors you may need to adjust.

Shades of white and black do not have a color component though, so the color correction pot does not affect them.  There's a YT video where someone demonstrates this with Space Invaders.  Neither the background nor the white invaders change color while the rest of the board does.

If white actually shows up as yellow (especially if the pot affects the color of "white"), then it's a problem with the TIA (or possibly communication with the TIA).

 

The likely culprit if some games lock up while others don't but have weird artifacts, is the RIOT chip.  Bad memory locations can cause any number of issues.  A second possibility is a broken trace, but that means one must go through the painstaking process of using the schematic to match pin to pin between all 3 chips and the cartridge port, to make sure they all have continuity where they are supposed to.

Then the other possibility is a short.  The best place to find shorts is where previous solder work was done.

[vitoco]

1 hour ago, ChildOfCv said:

Shades of white and black do not have a color component though, so the color correction pot does not affect them.  There's a YT video where someone demonstrates this with Space Invaders.  Neither the background nor the white invaders change color while the rest of the board does.

When I first saw this problem, I tried the pot, but the "white" didn't change as the remaining colors. Now, the pot is adjusted where all the elements in Warlords has their corresponding colors.

 

In the search for this, I tried many cartridges I have, and it seems that the color depends on the surrounding ones. The worst case is the PlusCart (+) logo, which looks completely yellow in a blue screen. In hounted house, the eyes and the score look orange/brown. In Math Gran Prix, white also look rusty. In Championship Soccer, the lines seem to be light green, and in Video Olympics, the playfield seem to be white, but the borders are yellow (there is a full scan line of this color).

 

1 hour ago, ChildOfCv said:

If white actually shows up as yellow (especially if the pot affects the color of "white"), then it's a problem with the TIA (or possibly communication with the TIA).

 

The likely culprit if some games lock up while others don't but have weird artifacts, is the RIOT chip.  Bad memory locations can cause any number of issues.

I bet it is one of my chips, and I hope I'll discover this during the weekend, because I'll try my chips in one of the consoles of a friend (and vice versa). This would explain why some games don't start when I power on the console using the switch, but run OK when I plug the power supply while the power switch was already on, and why the PlusCart resets or locks itself after some seconds of operation.

 

1 hour ago, ChildOfCv said:

A second possibility is a broken trace, but that means one must go through the painstaking process of using the schematic to match pin to pin between all 3 chips and the cartridge port, to make sure they all have continuity where they are supposed to.

I tried this but aborted when I found that there were many other components in the middle that I couldn't check because I don't know how to do that while they are soldered to the board.

 

1 hour ago, ChildOfCv said:

Then the other possibility is a short.  The best place to find shorts is where previous solder work was done.

I've checked this many times, and every time I move something, there is more to check!!!

 

BTW, I included a 75 Ohm resistor in the AV mod that it was not in the original tutorial I took, but it was recomended in another one. I removed it, but it did not solve the "white" problem.

 

On 10/3/2020 at 12:24 PM, Cmherndon79 said:

you can improve the color saturation on the rev13 board by installing a 820 ohm resistor between pins 6 and 9 on the TIA chip.

This was the only mod I did't include, because I didn't have that resistor by that time. Could this make the white to return?

 

[Cmherndon79]

2 hours ago, vitoco said:

BTW, I included a 75 Ohm resistor in the AV mod that it was not in the original tutorial I took, but it was recomended in another one. I removed it, but it did not solve the "white" problem.

 

This was the only mod I did't include, because I didn't have that resistor by that time. Could this make the white to return?

@vitoco No, unfortunately it won't fix the problem you're having with the color accuracy.  This fix improves color saturation. 

 

Per <Start> Atari Tech Tip #4  11/17/1982.  Subject:  Blanking Resistor.  Description:  Some 2600A PCBs have an 820 ohm 1/4 W 5% resistor installed on the solder side (bottom).  The resistor is located between pins 6 and 9 of A201 (TIA) and improves the color reproduction of the unit.

 

The addition will result in improved color saturation.

 

Rev. 16 PCBs and above will have the resistor incorporated into their design. <End> (Removed bits about Atari PNs for the resistor and expenses of the part, no longer relevant) 

 

I've also seen many RIOT chips cause a no output situation, though every single time I have seen that the behavior is over any title I try.  I've seen others boot up with garbled graphics or strange behavior.  With only 128 bytes of RAM in the RIOT, pretty much every title uses all the memory, all of the time I'd estimate. 

 

I'd say it is possible you've got a TIA problem and if you've got a friend that you can swap TIAs between, it's worth experimenting as at least the Rev. 13s all have socketed chips (Some Rev 16s and all 17s I've seen they're soldered down, as well as the Jr.).  When swapping TIAs it isn't uncommon for the colors to come up differently from one to another and adjusting the trimmer/pot is necessary and I usually use Pitfall! as my reference.  If you're going to swap TIAs make sure you're swapping them that are made for the same region, I think you said yours was NTSC, so you'll need an NTSC TIA - Atari PN C010444.  RIOTs and CPUs are not region specific however.

 

 

[ChildOfCv]

10 hours ago, vitoco said:

I tried this but aborted when I found that there were many other components in the middle that I couldn't check because I don't know how to do that while they are soldered to the board.

I'm not sure what you mean.  If the schematic shows multiple connections, then all of the connections shown are supposed to be connected together.

 

I reduced the most critical tests to a pin connection map though.  You can place a probe on the appropriate pin of the cartridge port and then verify connectivity to each of the other chips at their appropriate pins.

2600Continuity.pdf

[vitoco]

13 hours ago, Cmherndon79 said:

I've also seen many RIOT chips cause a no output situation, though every single time I have seen that the behavior is over any title I try.  I've seen others boot up with garbled graphics or strange behavior.  With only 128 bytes of RAM in the RIOT, pretty much every title uses all the memory, all of the time I'd estimate. 

The common strange behavior of some carts at power on makes me think that there is something wrong with the RIOT. But the fact that all of those games can be started using the 9V plug to power on the console makes me doubt. I'd like to know what do those cartridges have in common.

 

13 hours ago, Cmherndon79 said:

I'd say it is possible you've got a TIA problem and if you've got a friend that you can swap TIAs between, it's worth experimenting as at least the Rev. 13s all have socketed chips (Some Rev 16s and all 17s I've seen they're soldered down, as well as the Jr.).

The other console will be a socketed Jr., and we will do the swap LIVE today at 17:00 GMT-3 on Youtube:

 

If someone wants to help, I'll be checking the chat.

 

13 hours ago, Cmherndon79 said:

When swapping TIAs it isn't uncommon for the colors to come up differently from one to another and adjusting the trimmer/pot is necessary and I usually use Pitfall! as my reference.  If you're going to swap TIAs make sure you're swapping them that are made for the same region, I think you said yours was NTSC, so you'll need an NTSC TIA - Atari PN C010444.  RIOTs and CPUs are not region specific however.

He has a NTSC TIA chip, but my only concern is that it seems the Jr was modded with cables directly to the pins. I hope this won't be an issue during the tests.

 

16 hours ago, vitoco said:

The worst case is the PlusCart (+) logo, which looks completely yellow in a blue screen.

Tonight, @ZeroPage Homebrew showed a LIVE PlusCart firmware update (link). Before the update (ver 0.14.15), his logo was yellow like mine (ver 0.15.0). After the update, he got the brand new animated white logo. What it is strange is that I got the white logo in my RAMBO 2600B clone. BTW, James has some pixel glitches I also have.

 

6 hours ago, ChildOfCv said:

I reduced the most critical tests to a pin connection map though.  You can place a probe on the appropriate pin of the cartridge port and then verify connectivity to each of the other chips at their appropriate pins.

2600Continuity.pdf 61.28 kB · 2 downloads

Thank you... That it's very usefull. I was using the diagram at the last page of the 2600 Field Service Manual.

 

 

[vitoco]

Well, we put all the chips from the Jr into my Woody and the test cartridge did not fail, i.e. it started OK at the power on. Then we restored the TIA chip: didn't fail. Then the RIOT chip: didn't fail... Then the CPU 6507: FAILED!!! That was unexpected...

 

The next test was to try the PlusCart, but it failed. We put all the "new" chips in the board, but it still failed. We tried the PlusCart in a Heavy Sixer, and it also failed, so I have to check the kit I built... once again!

 

Out of camera, I tested all the cartridges using the full "new" chipset and all of them worked as expected, including the ones that always failed. So I have to find a replacement for the 6507, and I think that in the meanwhile I'll use the one from my 1050 disk drive.