S-Video mod.

[Albert]

My only prob is that the cables are so short... Anyone know where I could get an S-video cable with a female end or some other adaptor to alow me to plug another S-video cable into the S-video cable from the mod?

 

I agree, the cables are far too short. For the five systems I modified yesterday, I only used the audio cable that came with the mods. I used a six foot a/v cable instead of the s-video cable to separate out the chroma and luminance outputs, since I am using these systems with Commodore 1702 monitors. These monitors have RCA inputs for Chroma and Luma on the back. Fortunately, they have an audio jack on the front that continues to work even if you are using the back video inputs.

 

Radio Shack sells (or at least used to sell) a little adapter that lets you bridge two s-video cables, which would allow you to use a longer cable. When I have time, I am going to install panel jacks in these systems so I can use audio and video cables of any arbitrary length.

 

..Al

[Rockin' Kat]

Why do you guys even bother with the RF shielding if you are no longer using RF?    

 

I didn't put them back on.

 

..Al

 

And neither did I. I also didn't bother to try getting the RF sheild back on when I installed the 5200 video mod in my 5200.

[cwilkson]

I ran into one of those sockets too. Those are the best sockets money can buy. Isn't it funny (more like sad) how much variation there was in not only the component count, but the quality of the components in the various 2600 models?

 

The easiest way I could think of to install the board into that socket was to do the following:

 

1) place a towel on your work surface

2) remove the motherboard from the 2600 and place it face up on the towel

3) align the CyberTech socket pins with the holes in the 2600 TIA socket

4) using you palm, slowly (= CAREFULLY) but inexorably add your weight on top of the CyberTech socket.

 

It helps if you get one end started first. You have to essentially reshape the plastic of the mb socket when you're doing this and it's hard to do it to 40 of them at once.

 

If you get it reseated all the way, it should fit fine with the RF shield in place. You will need to cut holes for the cables if you reinstall the shield, but the TIA should fit fine. Atari2600.com could have used sockets with a different pin length than I did. If you're certain that the pins are fully seated, and the RF shield doesn't fit, you can remove the CyberTech board, trim the pins, and reinstall. If you do this be sure to cut the pin at an angle so it will be easy to reinstall (think sharpened stick). And don't cut them too short! There has to be a space between the 6507 on the mb and the bottom of the CyberTech board.

 

-Chris

[atarian1]

So I guess I have one of these super strength sockets, eh?

 

Well, I'll try just pushing it in as much as I can and hope for the best. I wonder what the screen will look like if the board doesn't seat properly? I hope it doesn't affect the quality of the display...

[cwilkson]

So I guess I have one of these super strength sockets, eh?  

 

Well, I'll try just pushing it in as much as I can and hope for the best.  I wonder what the screen will look like if the board doesn't seat properly?     I hope it doesn't affect the quality of the display...

 

Sounds like it. If it goes in a little but pops back out, it's not seated all the way. In fact, if it's that hard to put in it'll take a concentrated effort to remove a properly seated board.

 

What happens if it's not all the way in? Well...bad case is that you don't get any video at all. Possibly worse is that you get great video. And you're about to beat Todd Rogers' Barnstorming score when POP! out comes the board, and you die. :/

 

If you get it in a little bit, just keep pushing. You should feel it when you've bottomed out.

 

-Chris

 

I know...get your mind outta the gutter!

[atarian1]

Thanks guys for all your help.

 

The board does stay in for a while, but one bump of the console and it pops out again. Now when I look at the socket, it looks like the "hooks" are bent out of shape from constantly pushing the board back in after it pops out. I think these bent "hooks" are not creating a good connection anymore because the screen now displays a blue ghost of all the images, and the whites are now pink. (Example: the black line blocking the entrance to the hidden room in Adventure now displays a blue ghost line next to the black line) When I first installed the board, the system displayed a great sharp picture.

 

I hate to desolder the old socket from the motherboard, but it looks like the only option I can think of now. Or maybe something else?

 

The system still works if I connect the TIA back into the original board socket though. Just that I get the old RF quality picture.

[cwilkson]

Thanks guys for all your help.

 

The board does stay in for a while, but one bump of the console and it pops out again.  Now when I look at the socket, it looks like the "hooks" are bent out of shape from constantly pushing the board back in after it pops out.  I think these bent "hooks" are not creating a good connection anymore because the screen now displays a blue ghost of all the images, and the whites are now pink.  (Example: the black line blocking the entrance to the hidden room in Adventure now displays a blue ghost line next to the black line)   When I first installed the board, the system displayed a great sharp picture.  

 

I hate to desolder the old socket from the motherboard, but it looks like the only option I can think of now.    Or maybe something else?

 

The system still works if I connect the TIA back into the original board socket though.  Just that I get the old RF quality picture.    

 

Hmmm. Where are you in norcal? I'll be interviewing in Folsom (just east of Sac) sometime in June. If you want to hook up, I can give you some personal tech support while I'm there. Together, I'm sure we can get it working reliably.

 

BTW, if the picture is fine (albeit RF) when you plug the TIA into the mb socket, then you'll get great contact with the wirewrap socket in the CyberTech board. As someone else mentioned, the pins are *much* larger than standard chip pins.

 

-Chris

[atarian1]

Hi Chris,

 

I'm in the SF Bay Area, so Folsom is a little too far.

 

Anyway, more on today's adventures...

I went to the electronics store to pick up some stuff, I decided to fiddle with the 40-pin sockets at the store by trying to stick a socket like the one on the Cybertech board into a socket like the one on the 2600 motherboard. I ended up successfully doing this by inserting the pins starting from one side to the other.

 

So I tried doing this with the board at home. This time, instead of placing the board with all the pins lined up and pushing the board into the socket at once, I pushed the right pins into the socket first, the moved to the left. After this, I got a solid connection. I bumped the unit and tried pulling the board out with just a little strength, but it didn't pop out! I pushed the board in until the "crinking" stopped, so I think the connection is solid now. The RF shielding still won't fit though, so I left it out.

 

I tested it out and.....it still looks the same. Colors are still off (white looks blueish and the blue ghost image line on the right side of the screen) I double checked the pins, and they were all lined up in the socket. My soldering looks ok. I'm really puzzled now.

 

Here's something wierd. I tried plugging in an S-Video to Composite adapter and connected it up to my TV, and got a black and white image instead! I don't know if that helps any.

 

Any ideas?

[cwilkson]

I tested it out and.....it still looks the same.      Colors are still off (white looks blueish and the blue ghost image line on the right side of the screen)  I double checked the pins, and they were all lined up in the socket.   My soldering looks ok.  I'm really puzzled now.    

 

Here's something wierd.  I tried plugging in an S-Video to Composite adapter and connected it up to my TV, and got a black and white image instead!  I don't know if that helps any.

 

Any ideas?

 

Wow...this is a tough one. The color circuitry is so tight to get the good quality you initially saw....any of a hundred things could be wrong. What kind of test equipment do you have access to? And which board rev do you have? (is it the rectangle or the L-shaped one?)

 

It sounds like you aren't getting any chroma at all. It could be a bad or broken solder joint (yours or the factory's). It could be a blown chip on the CyberTech card. Or you could just be getting a lousy connection between the TIA and the CyberTech socket. (unlikely)

 

Let me know if you have test equipment available (particularly an oscilloscope). And also if you have a Rev 1.0 board or the earlier rectangular board. Also what kind of 2600 are you using? I need the case style and the motherboard revision number (printed on the board).

 

With the composite adapter, is the black and white image good quality?

Does the line still appear on the right side?

 

-Chris

[cwilkson]

Hmmm...a long shot. Try adjusting the color delay pot on the mother board. I don't think it's causing the problem, since you initially saw a good picture. But maybe you bumped it during all the banging around?

 

It's worth a try. I know if that pot is misadjusted too far the color disappears except for one color [pale greens?] which might look blue or purple....

 

-Chris

[cwilkson]

One more thing...just to reassure you: it sounds like you have the board in all the way now.

 

-Chris

[Mike OC]

Here's something wierd. I tried plugging in an S-Video to Composite adapter and connected it up to my TV, and got a black and white image instead! I don't know if that helps any.

 

When you change from S-video to composite, most newer TV's require you to either flip a switch or adjust settings in the menu for the apropriate signal source. I had the same issue and found that there was a small switch on the back of my TV that was near the S-Video input. The switch allows the user to choose S-video/Composite. Hope it helps.

[atarian1]

 

Wow...this is a tough one.  The color circuitry is so tight to get the good quality you initially saw....any of a hundred things could be wrong.  What kind of test equipment do you have access to?  And which board rev do you have?  (is it the rectangle or the L-shaped one?)

 

It sounds like you aren't getting any chroma at all.    It could be a bad or broken solder joint (yours or the factory's).  It could be a blown chip on the CyberTech card.  Or you could just be getting a lousy connection between the TIA and the CyberTech socket.  (unlikely)

 

Let me know if you have test equipment available (particularly an oscilloscope).  And also if you have a Rev 1.0 board or the earlier rectangular board.  Also what kind of 2600 are you using?  I need the case style and the motherboard revision number (printed on the board).

 

I don't have any complex equipment like that. I'm not really an electronics guy as the only thing I can do is solder this to that. I have a multimeter which I have no idea how to use.

 

I have revision 1.0 of the board (the "L" shaped one). I also have a 4-switcher. I believe I have Rev G of the motherboard, but I'll have to check.

 

With the composite adapter, is the black and white image good quality?

Does the line still appear on the right side?

 

Yes, the line still appeared except it was gray instead.

 

Hmmm...a long shot. Try adjusting the color delay pot on the mother board. I don't think it's causing the problem, since you initially saw a good picture. But maybe you bumped it during all the banging around?  

 

It's worth a try. I know if that pot is misadjusted too far the color disappears except for one color [pale greens?] which might look blue or purple....

 

I tried all the pots on the board and on the motherboard with no good results.

 

BTW, I don't think this helps, but I didn't use the cable that came with the board. I drilled holes into the 2600 and installed S-Video and RCA audio ports instead. I checked the solder connections between the board and the jacks, and they look pretty solid. No wires touching each other either.

[cwilkson]

Ok...without an o'scope I can't think of any reasonable way to remotely debug your issue.

 

How about a cuttle cart? Can you load arbitrary code into your 2600? If so, we can try some analog tricks

with your multimeter.

 

Actually, scratch that. Here's a sanity-check test.

 

Turn the machine off and disconnect the s-video cable. Remove the TIA from the CyberTech board and set it aside. Then solder a test wire (#22 gauge solid) to pin 6 of the wirewrap socket, between the CyberTech board and the motherboard. Next add a wire to pin 9. These will let us simulate digital signal levels. FYI, pin 1 is at the end of the socket, right next to the missing pin 2. Pin 20 is in the same row of pins, at the other end. Now connect both wires to pin 1 (= ground). These are temporary connections. Finally, connect the negative lead of your multimeter to pin 1 and set the meter to read DC Volts from 0V to +5V. Now we're ready.

 

Test 1:

Turn the 2600 on and measure the voltage on pins 1 and 20. They should be 0V and +5V respectively.

 

Test 2:

Measure the chroma voltage. This will be a little tricky. There is a convenient test point - a hole right next to the place you soldered the chroma wire to. But this is capacitively coupled and will only allow you to measure machine generated signals. The capacitor is a surface mount component connected to the test point. Take your measurement at the other end of that cap. Be careful to not let your probe slide across the board by accident! The voltage should be roughly 2.05V.

 

Test 3:

Connect pin 9 to pin 20 (instead of pin 1) and measure chroma. The voltage should be roughly 2.95V.

 

Test 4:

Connect pin 6 to pin 20 (instead of pin 1) and measure chroma. By turning the SAT pot on the CyberTech board, you should be able to vary chroma between 3.75V and 2.6V.

 

Test 5;

Connect pin 9 to pin 1 (instead of pin 20) and measure chroma. By turning the SAT pot on the CyberTech board, you should be able to vary chroma between 1.25V and 2.4V.

 

Turn off the 2600.

 

Let me know what you find.

 

-Chris

[gulag picture radio]

BTW where are you getting your S-video female sockets from? I've always had a bear of a time finding those. I found some at a mail oder supply, but they're PC mount an I'd rather have some sort of chassis mount types. In aperfect world, I'd love to find some that have a not around the connector itself to mount into a case (sort of like some RCA connectors... if that made any sense)

 

I've searched for it as a 4 pin mini-din as well. Don't know why these are so hard to find! Thanks for any help.

[bjk7382]

@gulag - I just pulled This up on www.froogle.com I don't know if it is what you are looking for, but froogle works pretty good.

[+Nathan Strum]

BTW where are you getting your S-video female sockets from?  I've always had a bear of a time finding those.

 

One that I like, is Mogami's TCS-415P (about halfway down the page). It's actually a pass-through connector, so you can just plug the short S-VHS cable from the video mod into the backside of the connector. No extra soldering.

 

CalRad also has some. You have to download their PDF catalog, and look on pages 46 (S-VHS bulkhead pass-through) and 40 (chassis-mount S-VHS jack).

 

But since neither place sells directly, you'll have to go to their vendors pages to find someone who does. But once you find one, you're all set. (I'm in California and use Pacific Radio.)

 

Also, S-VHS jacks require a pretty huge hole (1/2" or larger), so you may need to invest in a real big drill bit or a hole saw. Or lots of quality time with a Dremel.

[atarian1]

Ok...without an o'scope I can't think of any reasonable way to remotely debug your issue.

 

How about a cuttle cart?  Can you load arbitrary code into your 2600?  If so, we can try some analog tricks

with your multimeter.

 

Actually, I do have a Cuttle Cart.

 

Actually, scratch that.  Here's a sanity-check test.

 

Turn the machine off and disconnect the s-video cable.  Remove the TIA from the CyberTech board and set it aside.  Then solder a test wire (#22 gauge solid) to pin 6  of the wirewrap socket, between the CyberTech board and the motherboard.  

 

I couldn't get my soldering iron between the Cybertech board and the socket, so I just pushed a wire into the pin socket. Is that ok?

 

Next add a  wire to pin 9.  These will let us simulate digital signal levels.  FYI, pin 1 is at the end of the socket, right next to the missing pin 2.  Pin 20 is in the same row of pins, at the other end.  Now connect both wires to pin 1 (= ground).  These are temporary connections.  Finally, connect the negative lead of your multimeter to pin 1 and set the meter to read DC Volts from 0V to +5V.  

 

I set my multimeter to "20" in the "DCV" setting. I don't see a setting on my meter that is between 0 and 5 in the DCV setting area. It seems to work in the first test.

 

Now we're ready.

 

Test 1:

Turn the 2600 on and measure the voltage on pins 1 and 20.  They should be 0V and +5V respectively.

 

This checks out.

 

Test 2:

Measure the chroma voltage.  This will be a little tricky.  There is a convenient test point - a hole right next to the place you soldered the chroma wire to.  But this is capacitively coupled and will only allow you to measure machine generated signals.  The capacitor is a surface mount component connected to the test point.  Take your measurement at the other end of that cap.  Be careful to not let your probe slide across the board by accident!  The voltage should be roughly 2.05V.

 

I don't see a "hole", but I do see a blob of solder connected to a line between the Chroma pin and capacitor C6. This "hole" may have been filled with solder somewhere along the way.

 

I placed one probe on the solder blob and one on the right end of capacitor, but my multimeter registers nothing. Just 0V.

 

Test 3:

Connect pin 9 to pin 20 (instead of pin 1) and measure chroma.  The voltage should be roughly  2.95V.

 

Test 4:

Connect pin 6 to pin 20 (instead of pin 1) and measure chroma.  By turning the SAT pot on the CyberTech board, you should be able to vary chroma between 3.75V and 2.6V.

 

Test 5;

Connect pin 9 to pin 1 (instead of pin 20) and measure chroma.  By turning the SAT pot on the CyberTech board, you should be able to vary chroma between 1.25V and 2.4V.

 

Tests 3, 4 and 5 registered 0V on my multimeter. (Assuming the solder blob is the "hole" you're referring to.)

[cwilkson]

I couldn't get my soldering iron between the Cybertech board and the socket' date=' so I just pushed a wire into the pin socket. Is that ok?[/quote']

It'll work, but it could spread the pins of the socket if the wire is too big. Just be gentle with it. Use some telephone wire if you have it. It's only #24 gauge.

 

Test 1:

Turn the 2600 on and measure the voltage on pins 1 and 20. They should be 0V and +5V respectively.

This checks out.

Good. That means the board is getting power.

 

Test 2:

Measure the chroma voltage. This will be a little tricky. There is a convenient test point - a hole right next to the place you soldered the chroma wire to. But this is capacitively coupled and will only allow you to measure machine generated signals. The capacitor is a surface mount component connected to the test point. Take your measurement at the other end of that cap. Be careful to not let your probe slide across the board by accident! The voltage should be roughly 2.05V.

I don't see a "hole"' date=' but I do see a blob of solder connected to a line between the Chroma pin and capacitor C6. This "hole" may have been filled with solder somewhere along the way.

 

I placed one probe on the solder blob and one on the right end of capacitor, but my multimeter registers nothing. Just 0V.

[/quote']

Oh right. They were wave soldered. The hole would have been filled in. The 0V is expected on that blob. That's what I meant about only being able to measure the machine generated signals. The capacitor blocks every thing at low frequencies...no DC will get through it and everything will be 0V.

 

Test 3:

Connect pin 9 to pin 20 (instead of pin 1) and measure chroma. The voltage should be roughly 2.95V.

 

Test 4:

Connect pin 6 to pin 20 (instead of pin 1) and measure chroma. By turning the SAT pot on the CyberTech board' date=' you should be able to vary chroma between 3.75V and 2.6V.

 

Test 5;

Connect pin 9 to pin 1 (instead of pin 20) and measure chroma. By turning the SAT pot on the CyberTech board, you should be able to vary chroma between 1.25V and 2.4V.

[/quote']

 

Tests 3, 4 and 5 registered 0V on my multimeter. (Assuming the solder blob is the "hole" you're referring to.)

 

You want to measure at the other end of the capacitor. Try tests 2-5 again, measuring at that point and let me know. Leave the negative lead of your meter on pin 1 for all of the tests. E.g. for Test 2, one lead is on pin 1 and the other is on the end of the chroma capacitor that is farthest from the board's chroma output.

 

-Chris

[atarian1]

Ok. Let's try this again.

 

 

Turn the machine off and disconnect the s-video cable.  Remove the TIA from the CyberTech board and set it aside.  Then solder a test wire (#22 gauge solid) to pin 6  of the wirewrap socket, between the CyberTech board and the motherboard.  Next add a  wire to pin 9.  These will let us simulate digital signal levels.  FYI, pin 1 is at the end of the socket, right next to the missing pin 2.  Pin 20 is in the same row of pins, at the other end.  Now connect both wires to pin 1 (= ground).  These are temporary connections.  Finally, connect the negative lead of your multimeter to pin 1 and set the meter to read DC Volts from 0V to +5V.  Now we're ready.

 

Test 1:

Turn the 2600 on and measure the voltage on pins 1 and 20.  They should be 0V and +5V respectively.

 

Test 2:

Measure the chroma voltage.  This will be a little tricky.  There is a convenient test point - a hole right next to the place you soldered the chroma wire to.  But this is capacitively coupled and will only allow you to measure machine generated signals.  The capacitor is a surface mount component connected to the test point.  Take your measurement at the other end of that cap.  Be careful to not let your probe slide across the board by accident!  The voltage should be roughly 2.05V.

 

This came out 2.45V.

 

Test 3:

Connect pin 9 to pin 20 (instead of pin 1) and measure chroma.  The voltage should be roughly  2.95V.

 

This also came out 2.45V.

 

Test 4:

Connect pin 6 to pin 20 (instead of pin 1) and measure chroma.  By turning the SAT pot on the CyberTech board, you should be able to vary chroma between 3.75V and 2.6V.

 

I wasn't sure which was the SAT pot, so I tried all of them. It looks like the far left one by itself is it since it's the only one that changes voltage when I turn it.

 

Here I got a range of 1.75V to 2.45V.

 

Test 5;

Connect pin 9 to pin 1 (instead of pin 20) and measure chroma.  By turning the SAT pot on the CyberTech board, you should be able to vary chroma between 1.25V and 2.4V.

 

Hmm. I got the same range again - 1.75V to 2.45V.

 

The numbers looks close to except #4. Hmmm.

[cwilkson]

Ok, now we're getting somewhere. Let me think about this for a while.

 

BTW, you found the SAT pot. I forgot that I didn't have room to label them on the Rev 1.0 silkscreen.

 

-Chris

[cwilkson]

Ok...without an o'scope I can't think of any reasonable way to remotely debug your issue.

 

How about a cuttle cart?  Can you load arbitrary code into your 2600?  If so, we can try some analog tricks

with your multimeter.

 

Actually, scratch that.  Here's a sanity-check test.

 

Turn the machine off and disconnect the s-video cable.  Remove the TIA from the CyberTech board and set it aside.  Then solder a test wire (#22 gauge solid) to pin 6  of the wirewrap socket, between the CyberTech board and the motherboard.  Next add a  wire to pin 9.  These will let us simulate digital signal levels.  FYI, pin 1 is at the end of the socket, right next to the missing pin 2.  Pin 20 is in the same row of pins, at the other end.  Now connect both wires to pin 1 (= ground).  These are temporary connections.  Finally, connect the negative lead of your multimeter to pin 1 and set the meter to read DC Volts from 0V to +5V.  Now we're ready.

 

Test 1:

Turn the 2600 on and measure the voltage on pins 1 and 20.  They should be 0V and +5V respectively.

 

Test 2:

Measure the chroma voltage.  This will be a little tricky.  There is a convenient test point - a hole right next to the place you soldered the chroma wire to.  But this is capacitively coupled and will only allow you to measure machine generated signals.  The capacitor is a surface mount component connected to the test point.  Take your measurement at the other end of that cap.  Be careful to not let your probe slide across the board by accident!  The voltage should be roughly 2.05V.

 

Test 3:

Connect pin 9 to pin 20 (instead of pin 1) and measure chroma.  The voltage should be roughly  2.95V.

 

Test 4:

Connect pin 6 to pin 20 (instead of pin 1) and measure chroma.  By turning the SAT pot on the CyberTech board, you should be able to vary chroma between 3.75V and 2.6V.

 

Test 5;

Connect pin 9 to pin 1 (instead of pin 20) and measure chroma.  By turning the SAT pot on the CyberTech board, you should be able to vary chroma between 1.25V and 2.4V.

 

Turn off the 2600.

 

Let me know what you find.

 

-Chris

 

Ok. I made a mental mistake. The power supply for the chroma section is actually 3.3v, not 5V. Grrr. Amazing what sleeping after finals week will do for the mind.

 

Above, I've quoted my previous post in it's entirety. Because editing is turned off, I'll just summarize here in a truth table.

 

+-------+-------+-----------+----------------+

| pin 6 | pin 9 | sat pot   | chroma voltage | (measured)

+-------+-------+-----------+----------------+

|  0v   |  0v   | no effect | 1.37v          | (1.368v)

|  0v   | +5v   | no effect | 1.93v          | (1.934v)

| +5v   |  0v   | left      | 1.58v          | (1.579v)

| +5v   |  0v   | right     | 0.833v         | (0.866v)

| +5v   | +5v   | left      | 1.72v          | (1.723v)

| +5v   | +5v   | right     | 2.47v          | (2.431v)

+-------+-------+-----------+----------------+

 

Hmmm....that looks like crap. I don't know how to get a monospaced font in here, so you'll have to deal.

 

Anyway, whatt you measure should be very close to the design spec above. When I say very close you should hear "almost exactly...within a few mV". When the pot is turned to the right and pin 6 = +5v, the values could be a few tens of mV away, but no more. This is because pots don't truly zero out sometimes.

 

Some of your numbers are way off. So try repeating the measurements, but this time measure the voltage at the middle pin of the saturation pot. It's the one right beside the big electrolytic cap. For your peace of mind, these are indeed the values you should see. I included the measured values from my partially stuffed test board next to the calculate values.

 

-Chris

[cwilkson]

PS. Please note whether the pot is turned all the way left or all the way right when you take your measurements.

 

-Chris

[cwilkson]

PPS. Turns out my browser fonts were hard coded....the truth table looks fine afterall.

 

-Chris

[cwilkson]

Any luck?

 

-Chris