Jump to content
IGNORED

Atari RGB Light Sixer Repair


ZeroPage Homebrew

Recommended Posts

Hey everyone! As you have heard me talk about on the show a number of times, my RGB light-sixer has been down for the count for a few months now. I've done all I can with the knowledge and skills I have, which consist of poor soldering skills that didn't help anything.

 

Thankfully @Nathan Strum has graciously offered his help to see if he's able to diagnose the issue and get it up and running again! In the mean time @Albert has sent me his S-Video modded four switch that we used for the first time on the most recent show, which was a HUGE step up in quality of the composite Atari Jr we've been using for a while now. Thank you so much Al!!

 

Here's the VCS all packed up nicely and ready to be shipped to Nathan, it was sent on May 22 and just arrived to Nathan today, hooray! I'll let Nathan take over from here on the incredible saga of the RGB Light-Sixer repair. Thank you so much Nathan!!

 

- James

 

image.thumb.png.3278fe550a1e666dbfe3b3b03f564c14.png

  • Like 5
Link to comment
Share on other sites

Hi all!

 

James' 2600 did indeed arrive Monday. I had to clear off my workspace and set up my gear so I could find the top of my desk get to work on it.

 

This is my workspace (pictured below). The 2600 guts you see here are parts donors. I bought them as populated boards only (no case) from Best Electronics a few years ago to troubleshoot my own 2600. I still have enough spare parts to keep it running, so I can use it as a test platform. It has new caps, a power jack and voltage regulator. It still needs some work, because when I press a joystick button connected to it, the picture gets dimmer (it's done that since I bought it). So, there's that. But it's had a tough life. Maybe after I'm done with James' I'll restore this one properly. It's running a new-old-stock Atari power supply from Best, and the photo shows it running the Colors binary off of a Harmony Cart. Plus, I have a Pac-Man cart handy, because that will run on anything, and I don't care if it blows up. Plus there are some spare chips and sockets sitting off to the side.

 

Other stuff includes a Hakko 936 soldering station w/907 iron, a Hakko FR-301 desoldering iron, a Sony PVM-14M2U monitor, Sony TU-1041U tuner (to demodulate RF into composite), a MicroImage color bar generator to calibrate the monitor, and the hammer is by Craftsman.

 

I miss Sears...

 

hammer-time.thumb.jpg.fb0f8bb4ba3229957d605db46f568703.jpg

 

The Sony setup makes even RF look good. Although using Clark CV752 instead of a stock RF cable probably helps. Here's a close-up of the display:

 

rf-color-grid.thumb.jpg.5a40100e805dd95dd22a6b81168ad04c.jpg

 

And S-Video NTSC color bars, as a point of comparison:

 

s-video-color-bars.thumb.jpg.8a3928f4e195fa20957532d97138dbfd.jpg

 

So, James' 2600 arrived intact. With plenty of (presumably biodegradable) styrofoam peanuts:

 

zph-box.thumb.jpg.dbd4f7db4b257084be2079931150a5fa.jpg

 

And here it is:

 

zph-2600.thumb.jpg.aa767a777bbe1bea0f385795e1ef742d.jpg

 

There's just one little problem... 

 

mini-din.thumb.jpg.b2453457de3ca3f7b7734ac1ca37fbc4.jpg

 

The RGB mod uses an 8-pin connector, which isn't supported on my monitor (or any pro video equipment I've ever seen). James also sent along an adapter that converts it to SCART (also not supported). So, I have no way of testing the RGB mod at the moment. I'll have to get an adapter cable of some sort. I don't suppose someone would care to loan me one?

 

Meanwhile, I'm going to gut carefully open up the console and disconnect the RGB mod, and see if I can get the 2600 working as a stock RF-only console. Once I get that far and make sure the 2600 itself is sound, then I can worry about putting the RGB mod back.

 

I'll make a guess that the CD4050 hex buffer is toast. That was the first thing on my 2600 that died. Well, the first thing after the Reset switch. And a capacitor wayyyy back in the 80's. I remember my dad driving me out to Martha Lake Electronics (north of Seattle) to see if they could fix my console. I think they were the nearest Atari-authorized repair place. I was afraid my 2600 was dead permanently, but they fixed it! I thought that was the coolest thing in the world.

 

Anyway, I'll remove and socket each of the chips (if they aren't already), and test them in the donor 2600. James also sent along some spare parts including a new power jack, which I'll also install.

 

And for the record, James' Reset switch isn't broken.

 

But it sure looks tired. ;) 

 

- Nathan

  • Like 5
  • Haha 1
Link to comment
Share on other sites

Australian.

 

Near as I can tell, the mini-DIN 8 connector came about as a result of gamers using the Framemeister, which is Japanese. The Framemeister used it as a proprietary connector because... it was small? Cheap? I guess?

 

Anyway, it looks like the adapter may not actually be SCART, it may be RGB21. What I'll probably have to do is just make a cable.

Link to comment
Share on other sites

6 hours ago, Nathan Strum said:

Anyway, it looks like the adapter may not actually be SCART, it may be RGB21. What I'll probably have to do is just make a cable.

You got it! We wired it as JP21 as that's what the Framemeister takes natively to keep the number of adapters down when I plug it in. In the package I also included the JP21 (SCART looking) to Framemeister 8 pin mini DIN adapter (as seen in the picture) just in case you needed it.

 

Also, all the RF video out is intact and "functional" as we used it for troubleshooting when installing the RGB mod. Of course nothing appears functional right now but the RF was when it was working.

 

9 hours ago, Nathan Strum said:

And for the record, James' Reset switch isn't broken.

But it sure looks tired. ;)

It's VERY well used. ?

 

- James

Edited by ZeroPage Homebrew
Link to comment
Share on other sites

I just ordered an 8-pin mini-DIN serial cable and a VGA+Sync to BNC cable from Amazon. They should arrive Tuesday, when I will dissect and reassemble them into a horrible amalgamation of wires, solder and heat-shrink tubing. It won't be pretty, but it should let me see your RGB mod on my monitor.

  • Like 1
Link to comment
Share on other sites

Welcome to Chapter 2!

 

Well, time to rip this sucker apart!

 

After a little hammering, I was able to get the 2600 apart. I suppose a screwdriver would've been easier, but not nearly as much fun! ;) 

 

First discovery: Dirt, dirt, dirt! And more dirt! But at least I haven't found much cat hair.

 

so-much-dirt.thumb.jpg.538a40b3f3190e7ad043a181a516e30d.jpg

 

I'm not sure if cleaning this out will devalue it or not. Do collectors want original 40-year-old dirt left in their consoles? I never understood collectors. Anyway, maybe after I write my name in all of that dust with my finger I'll clean it off before sending it back to James. :D 

 

(I'm totally kidding about the dirt, of course. There's not that much dust in it. But he's always talking about how clean his 2600 is, so I have to give him some grief, right? ;) )

 

Anyway, onto the inspection. First thing I noticed is that the purple cable going to the "palette switch" (description here) is pinched between the board and shield. Not a big deal, since this is a single conductor cable, and I checked it for continuity and it's still fine.

 

pinchy.thumb.jpg.0bc04ff65cee7a9f5a0f16e77b68fc20.jpg

 

The next thing I did was check the parts James had already replaced on the switchboard assembly to see how those were soldered. The capacitors looked fine, but when I checked the fancy voltage regulator, there was a short across Ground and Output. So... that wasn't going to work.

 

So first things first - remove the voltage regulator and see if it's the cause of the short.

 

removed-regulator.thumb.jpg.15f6dd87f24d69749d702592d1114384.jpg

 

Nope! When disconnected it from the 2600, I checked the pins and the regulator was fine.

 

I also checked the one James had removed before, and it did have a short - across Ground and Input. In fact you can see a solder bridge there. But in all fairness, that could've happened when he removed it. I haven't tried it, but this regulator may be perfectly fine once the solder is cleaned up.

 

shorted-regulator.thumb.jpg.c8656be6a2ad0d21f84abaed6416e866.jpg

 

Anyway, with the regulator removed, I started checking the 2600 to see if the short was still there...

 

regulator-removed.thumb.jpg.2f91b150269a3404e82767fe67a6112a.jpg

 

Unfortunately, it was. So the regulator was not the cause of the short, nor was it at the pads beneath the regulator. That meant it was somewhere else. Time to do some digging.

 

The short showed up everywhere else on the switchboard, until I unplugged it from the main 2600 board. The short was gone! So that meant the switchboard wasn't the cause.

 

board-disconneted.thumb.jpg.6d837010dbed6c46b9c6c571d1b03cb0.jpg

 

Unfortunately, that meant the short was somewhere in the main board of the 2600 itself. Yeeeesh.

 

Checking the main board showed the short was indeed there. I colored the photo below to show where the short is on the other end of the board connector. It shorts between red (+5 v) and bright green (ground).

 

tracing-the-short.thumb.jpg.a24c9712a870c8fa0a008afbd5d2b012.jpg

 

So, time to start digging into that. Removing the switchboard gave me a look at the first thing I needed to fix on the topside:

 

bad-wire-no-biscuit.thumb.jpg.59668d238af9ebb1e3cca46b3726cc0e.jpg

 

The board connector wire on the far right was hanging on by a thread, and some wires were hanging loose. Was this the cause of the short? Well, no. The short would have actually been at the opposite end of that connector - the red wire (+5 v) and the beige wire to its left (ground).

 

To get a better idea of where the short might be happening, I flipped over my donor board, and used a multimeter to trace where everywhere the +5 volt line connected. I stuck some tape down to give me a visual reference that I could use while working on the ZPH board.

 

Basically... it goes everywhere.

 

tracing-on-the-donor.thumb.jpg.05e13446ac952ae3f109d4024480e61c.jpg

 

It was at this point that I decided the RGB mod board would have to come off. If the short was there, that would be the fastest way to find out.

 

the-mod.thumb.jpg.c34a8d8621d147d5848bc7b0b6f61e54.jpg

 

Fortunately, the mod itself is a plug-in module. The whole thing comes out without having to desolder anything.

 

the-mod-removed.thumb.jpg.9a6e3bde1e84b8531bfd177f8998262b.jpg

 

Okay, two things... the wires going up to Select and Reset for the so-called "Extra button". The idea is, by connecting these (and some other wires) to various points on the main board, you can add a button to a stock 2600 joystick that will allow you to use Select and Reset (and change palettes) remotely. However, the rest of the wires were never connected.

 

more-stuff-to-remove.thumb.jpg.ffc7b0c6f23d504514cb7edfdb3399f0.jpg

 

Anyway, so those had to come off. But that also gave me the chance to fix that one sketchy board connect wire while I was in the neighborhood.

 

fixed-bad-wire.thumb.jpg.370ce74b80403d2d068a5ba26386b6ed.jpg

 

With the mod removed, I tested the board again, and the short was gone! I popped the TIA off of the mod, and put it back in the 2600. The short was still gone! So that meant the short wasn't in the TIA itself, nor anywhere else on the stock 2600 board. Good news!

 

Checking the mod by itself showed that it's indeed where the short is - present between +5v and Ground. Now, where on the mod the short is... will have to wait for another day. First, there's a 2600 to get working.

 

So, with the TIA removed from the mod, and restored back to its rightful place, would the 2600 work? Would it? WOULD IT?!?

 

tia-restored.thumb.jpg.31040c0963f771c398c2cbe160c63337.jpg

 

 

Stay tuned for Chapter 3! :D 

  • Like 3
Link to comment
Share on other sites

5 minutes ago, Nathan Strum said:

the wires going up to Select and Reset for the so-called "Extra button". The idea is, by connecting these (and some other wires) to various points on the main board, you can add a button to a stock 2600 joystick that will allow you to use Select and Reset (and change palettes) remotely. However, the rest of the wires were never connected.

I wonder why James never used the feature to get rid of his reset problem. And the pause function seems like another nice to have on some games.

Link to comment
Share on other sites

2 hours ago, Thomas Jentzsch said:

I wonder why James never used the feature to get rid of his reset problem. And the pause function seems like another nice to have on some games.

Probably because it would restrict him to always having that modified joystick plugged in, to be able to use it.

Link to comment
Share on other sites

 

9 hours ago, Thomas Jentzsch said:

I wonder why James never used the feature to get rid of his reset problem. And the pause function seems like another nice to have on some games.

The wiring for the joystick extra button cause problems with some controllers (notably, trackballs and mice) or when using the left port as output.
A workaround is to install a switch to disable the extra button functionality when you're using an incompatible controller.


The existing "palette switch" can be used to pause a game instead of the extra button.

Link to comment
Share on other sites

On with Chapter 3!

 

Can you stand the suspense? Are you sitting on the edge of your seat? I'm leaning back in mine. Way more comfortable. Plus I can read more easily through my progressive lenses this way.

 

Anyway... when we last left the ZPH 2600, we were about to find out if it worked!!

 

Well, did it? DID IT?!

 

Well, nope. Mostly, I'd get this:

 

fuzzy-gray.thumb.jpg.1cb5538965ee8074bfcf736e8394ca02.jpg

 

Sometimes, I'd get this:

 

rolling-glitch.thumb.jpg.629b75c27a05bf7d7c426dda3f8ac4b3.jpg

 

That's Pac-Man, in case you were wondering.

 

Okay. So, now where's the problem?

 

Well, the first thing to suspect was the switchboard, because I'd reinstalled the fancy voltage regulator, rather than a stock one.

 

voltage-regulator-restored.thumb.jpg.8e661f90b7971531c210c60cb71a5714.jpg

 

But I did check it, and it was outputting a steady 4.95 volts. So that should be fine. But maybe something else was wonky. So I temporarily installed my donor's switchboard on top of the ZPH console, and fired it up:

 

frankensteined.thumb.jpg.d5351f4cc75e1a053cb31ba0e0397e17.jpg

 

Ah ha! So the console worked. So what could be wrong on the switchboard? Well, there was one cap in the recap kit that hadn't been installed yet. So I did that:

 

new-switchboard-cap.thumb.jpg.86a3c12c0318cbd16ae154c1a4d7f8cd.jpg

 

And since there was another spare cap left in the kit, I replaced one of the previously replaced caps again (the white one):

 

other-new-swithboard-cap.thumb.jpg.b05ea3ff7d4d715ee7590173fcea2d8d.jpg

 

The big honkin' blue one in the background had been replaced already, and I didn't have a handy spare, so I left it in place. If the board still didn't work, I'd pull the cap off my donor.

 

So I reconnected the ZPH board, fired it up... and:

 

teaser.thumb.jpg.cff90d38ad8a5e2dc0b072f4c62e6614.jpg

 

Success! Job done! 2600 restored!

 

At least until I turned it back off and on again.

 

Then it started showing the same glitches pictured above. So... now what's going on? It worked literally a minute before.

 

I reinstalled the donor switchboard again, and...

 

zph-board-donor-switches.thumb.jpg.8d7284ae88e3906b30a82864ed766720.jpg

 

It worked! Success! Job done! And such!

 

At least until I turned it back off and on again. Again.

 

I was getting the noise patterns again. So... something on the main 2600 board was still wonky. Pac-Man would show up about one out of every ten power-on attempts. Probably not really desirable for a live stream...

 

To double-check that it was the main board, I installed the ZPH switchboard on my donor's main board. And sure enough, it worked just fine. Repeatedly. No issues.

 

donor-board-zph-switches.thumb.jpg.4fb64ff181ccac29d37b3cccb4b83b34.jpg

 

But the ZPH main board was still not working.

 

At that point (late Saturday night), I decided I was done for the night. I posted the Teaser image, and started writing Chapter 2. I wrote up to the point where I was about to test the console, decided that was a suitably dramatic cliffhanger, and went to bed.

 

 

(Sunday)

 

Well, back to the main ZPH board then.

 

One plus about the ZPH 2600 is that all of its main chips are already socketed. So I wouldn't have to do that.

 

All of them except the hex buffer, that is:

 

old-buffer.thumb.jpg.892e298663099ef9e486bd995ae0cc9d.jpg

 

So I yanked carefully desoldered it, and socketed a fresh CD4050 chip in there. James had sent a spare one along, but I had just gotten some brand new ones from Digi-Key, so I decided to use one of those, not knowing if James' spare had any history to it.

 

new-buffer.thumb.jpg.7a38be127df320d4fde2fac54dc36167.jpg

 

Incidentally, this would all go a lot faster if I wasn't stopping to take pictures all the time. :roll:

 

Elsewhere on the main board, this cap hadn't been replaced yet:

 

old-main-cap.thumb.jpg.d457ca5f8fc2b5b27e06aad11b3bab06.jpg

 

So I did that, too:

 

new-main-cap.thumb.jpg.941f318564840bb0f7d1e32d7f7c663c.jpg

 

And while I was at it, I replaced the power jack. The old one (on the right) worked, but they do get dirty and intermittent after awhile, so why not?

 

new-power-jack.thumb.jpg.6ff60529d9558553070d8ec8d4d598fc.jpg

 

So, with a fresh hex buffer and new cap... would the 2600 finally work? Would it? WOULD IT?!?

 

Stay tuned for Chapter 4!!

 

 

 

 

 

 

Nah, I'm just messin' with ya'. It works.

 

recapped-zph-2600.thumb.jpg.56dc2f7338a764b94a7fcf9da4b36ab5.jpg

 

I've power-cycled it a few-dozen times, and it comes up every time now. My Harmony Cart works, too. I'm going to let it sit powered up for a few hours, and see how it's doing later. But right now, it seems happy.

 

Up next: Troubleshooting the RGB mod.

  • Like 3
Link to comment
Share on other sites

12 minutes ago, Thomas Jentzsch said:

Good job!

 

After all the fixes at once, will we ever know what was wrong?

 

Well, the mod itself created a short across +5 v and Ground. That's certainly the main issue. If I can't find a physical cause for it on the mod (solder bridge, bad wire), then it's a component failure. What might've caused that? Don't know. I seem to recall James had a power supply fail on him, and that could have caused it.

 

As for the console itself, I could put the old hex buffer back in the socket and test it again. Although with the console presently working, I'm not inclined to. But maybe I'll throw it in the donor and see what it does. It had to be that or the cap though, since I've changed nothing else on the main board.

 

I still have an issue on my own 2600 to fix. It won't work unless an AtariVox is plugged into the second joystick port. Guessing the RIOT chip is bad.

  • Like 1
Link to comment
Share on other sites

3 hours ago, Nathan Strum said:

Nah, I'm just messin' with ya'. It works.

 

I've power-cycled it a few-dozen times, and it comes up every time now. My Harmony Cart works, too. I'm going to let it sit powered up for a few hours, and see how it's doing later. But right now, it seems happy.

 

Up next: Troubleshooting the RGB mod.

 

This is incredible news and worth the cliffhanger wait from last night! Thanks so much Nathan!

 

That great that the basic VCS is still functional and hopefully the fault in the RGB components can be easily found and reinstalled. Looking forward to Chapter 4: Revenge of the MOD!

 

- James

Link to comment
Share on other sites

8 hours ago, alex_79 said:

 

The wiring for the joystick extra button cause problems with some controllers (notably, trackballs and mice) or when using the left port as output.
A workaround is to install a switch to disable the extra button functionality when you're using an incompatible controller.


The existing "palette switch" can be used to pause a game instead of the extra button.

 

I think I read about the incompatibility issues with installing the extra reset switch and I also didn't want extra cabling hanging out of my 2600 at all times. I was trying to keep the exterior looking as clean and original as possible. Getting up to press reset is pretty annoying sometimes though, especially when I'm going for a high score, hahah.

 

I don't think I ever thought of using the "palette switch" for a pause so far. I totally forgot about that until it was mentioned in this thread actually!

 

- James

Link to comment
Share on other sites

1 hour ago, ZeroPage Homebrew said:

Getting up to press reset is pretty annoying sometimes though, especially when I'm going for a high score, hahah.

 

Yeah, but it's part of the 2600 experience. Of course, the 2600 was never designed to be very far from the player. That's why there's that crazy-long RF cable with it.

 

I'm enjoying playing on a CRT again. This particular one is a loaner from work (where we still have leftovers from our analog production days). I may have to see if they'll let me borrow it permanently. :) 

 

The 2600 has been on all afternoon, and is still working and powering up just fine. I've unplugged it from power now, and will let it sit overnight and see how it starts up tomorrow. Meanwhile, I'm desoldering wires from the mod, to see if the short is in the wires, connector, or the mod itself. If it's a component failure, you're going to have to get it replaced. I can't repair surface mount stuff.

  • Like 1
Link to comment
Share on other sites

1 hour ago, Nathan Strum said:

The 2600 has been on all afternoon, and is still working and powering up just fine. I've unplugged it from power now, and will let it sit overnight and see how it starts up tomorrow. Meanwhile, I'm desoldering wires from the mod, to see if the short is in the wires, connector, or the mod itself. If it's a component failure, you're going to have to get it replaced. I can't repair surface mount stuff.

That's excellent the burn in test is going smoothly, hopefully not a burn in for the CRT!

 

If the RGB mod itself has failed I'll order a new one, no problem. ?

 

- James

Link to comment
Share on other sites

14 minutes ago, ZeroPage Homebrew said:

That's excellent the burn in test is going smoothly, hopefully not a burn in for the CRT!

I shut the monitor off during the test. :) 

14 minutes ago, ZeroPage Homebrew said:

If the RGB mod itself has failed I'll order a new one, no problem. ?

And with that...

 

Chapter 4!

 

(Yeah, that intro does not bode well.)

 

So, the first thing to check on the mod (being the simplest to fix) was the mini-DIN connector used for the video output. The short appeared across +5 v and Ground there too.

 

mini-din-front.thumb.jpg.a2a038a09c5660b58e433f60989e54c7.jpg  mini-din-back.thumb.jpg.8f845037273936b5687f3b25c51a63c8.jpg

 

The easiest thing to do would've been to cut the wires, but I desoldered them instead.

 

mini-din-desoldered.thumb.jpg.a7dbe8cad9b9f8c9882f644e6393a0d4.jpg

 

After that, there was no short present on the connector. Too bad... that would've been an easy fix. So, back to the mod itself.

 

mod-overview.thumb.jpg.8d0d0a1ff5b21b98358fd5e0247d24b3.jpg

 

Checking +5 v and Ground, the short is still there. So now what? Well, about the only thing I can troubleshoot at this scale are the wires. Maybe there was a tiny little chunk of solder between pads somewhere.

 

So I desoldered the +5 v wire, as well as the ones adjacent to it, just in case.

 

I also removed the wires for the joystick "Extra" button. James wasn't using them anyway, and didn't seem inclined to.

 

mod-desoldered.thumb.jpg.e3a540908bd938116e2082b286d26074.jpg

 

Unfortunately, the short is still present on the board. I also pulled the only socketed chip from the board. No luck.

 

I had considered desoldering the TIA socket and adapter board just to be thorough, but then realized the pins for +5v and Ground are literally at opposite ends of the TIA. So a soldering mistake couldn't cause a short there. (At least I didn't see any blobs of solder 2" long.)

 

rgb-mod-stack.thumb.jpg.99364a145267d5a808b34c2838e4efa4.jpg

 

It doesn't appear any work was done to the mod itself after initial installation. So for some sort of physical short to just develop is really unlikely.

 

So at this point (unless someone else has a suggestion) I'm declaring it a component failure on the mod. Backing that up, is that another AtariAge forum member recently posted that their 2600RGB mod had failed in the exact same manner.

 

So James - looks like you need a new mod. :( Unless the guy who makes these also does repairs. Either way, this one is done.

 

If you want to have it shipped directly to me, I'll install it. Maybe also get the component board? Not sure if that works with the Framemeister though.

 

Anyway... thus ends the saga for now. When I have a new mod to install, we'll pick it up again.

 

(I wish that Chris Wilkson had finished his CyberTech RGB mod. I've been incredibly happy with my S-Video one from him.)

  • Like 1
Link to comment
Share on other sites

It has a PF color timing glitch and maybe (not sure) a PF timing glitch too. The color timing glitch can be seen in Quick Step! (see left edge of 3rd platform column) and the PF timing glitch in Pesco (see PF column left of score). James was wondering if that might have been a result of the mod.

 

Examples (I hope Stella emulates the glitches correctly):

Quick Step! (1983) (Imagic).png

Pesco (Ebivision).png

Edited by Thomas Jentzsch
Link to comment
Share on other sites

  • Recently Browsing   0 members

    • No registered users viewing this page.
×
×
  • Create New...