I think I’ve killed my 800XL! Any ideas?

[bfollowell]

Ok, I’m starting a new thread on this one since it isn’t directly 1088XEL related. I’d been having video issues with my 1088 build and started wondering if I might have received a defective GTIA or ANTIC chip. I opened my 800XL and swapped the chips out of it. My 800XL’s chips acted the same in my 1088 as the chips I’d purchased for it. My 1088’s chips worked perfectly in my old 800XL. Ok, one potential problem ruled out.

 

So, I swap the chips back out, install the old chips back into my 800XL, and discover that I’ve killed the 800XL I’ve had for the past 35 years! I thought I was really careful and I have no idea what I may have done to it. Now, when I turn it on, I get a solid yellowish green image, no text, no audio, nothing. Just the green screen. I’ve attached an image below, so you can see what I see.

 

I’ve confirmed I installed the chips in the correct sockets. I’ve pulled and reseated the chips numerous times. I’ve verified proper voltage on all the pins coming out of the power supply. I’m not sure where else to check out or what to try.

 

Has anyone ever experienced a similar issue? Any idea what may have caused this or what I may have done.
 

Thanks.

[xrbrevin]

it may be worth reflowing the solder on the chip sockets in case the disturbance has upset a weak contact

[manterola]

Use a multitester to check continuity between every "pin shoulder" and the next component point (like other chip or resistor). My guess is that the antic or gtia socket died, i.e. the socket is not making contact with one chip foot. To discover which one, you just need to go one by one , using the schematics as a guide to find the next component.

once to discover the foot, try to fix it with contact cleaner or something or just install a good new socket.

[Nightengale]

Saw this green screen the other day. I wired the antic wrong when installing incognito into A800.  Easily fixed in my case...wire correctly and fixed straight away.

 

 

 

[bfollowell]

Ok, I’ll check the sockets out, starting with the ANTIC socket.

 

 Thanks guys.

 

[shoestring]

So many possibilities ( mmu, cpu, osrom, ram, 74LS logic ). Maybe even a single bent pin hiding between the top of the socket and bottom of the chip.

 

Do you have any carts to try in the machine ? ( star raiders ).

[Kyle22]

5 minutes ago, shoestring said:

So many possibilities ( mmu, cpu, osrom, ram, 74LS logic ). Maybe even a single bent pin hiding between the top of the socket and bottom of the chip.

 

Do you have any carts to try in the machine ? ( star raiders ).

Not likely in this case.

 

Check that existing [cheap] socket contacts were not damaged when you removed the chips. Check for bent under pins, etc.

 

Good luck.

 

Edit: This should be an easy one. You'll get it.

 

 

Edited April 7, 2020 by Kyle22

[bfollowell]

Well, just to be safe, I pulled and reseated both GTIA and ANTIC again. Then I went through and verified all 40 pins on each chip. I have continuity from the shoulder of each of the pins, to the solder joint on the bottom of the board, and on to the next component. Unfortunately in this case, everything checks out good, so I'm not sure where to turn next. Since the only thing I did was remove and reinstall those two chips, it makes more sense than not that it would be something related to  those chips or their sockets, but it isn't looking like it.

 

I'll keep looking but I'm kind of flying blind, so any ideas to help me narrow things down would be greatly appreciated.

 

Thanks!

 

[_The Doctor__]

make sure the chips weren't hit by static or damaged by whatever other device you might have put them in, a second known working good old Fashioned Atari can be a life saver here. The chips normally won't kill anything when bad, but bad wiring can kill chips. Please send pictures of whatever pcb's the chips have been in, so visual tracing of circuit boards, orientation, solder bridges might be seen. Maybe a sys check or someone who has one can be some advantage or use here as well.

Edited April 7, 2020 by _The Doctor__

[bfollowell]

I only my 1088XEL and my 800XL. I have nothing else to swap them to except back and forth. Upon my initial swap, the old chips from my XL worked in my 1088 and vice versa. It wasn't until I swapped them the second time, to get things back where the belonged, that things got a little flakey.

 

I guess I could try my XL's chips in the 1088 again, just to rule out one or both of them being bad.

 

Edited April 7, 2020 by bfollowell

[_The Doctor__]

I'd make sure of that 1088 (pictures) before plopping new chips in there but you've go nothing to loose in regard to the old chips now...

Edited April 7, 2020 by _The Doctor__

[manterola]

13 minutes ago, bfollowell said:

I only my 1088XEL and my 800XL. I have nothing else to swap them to except back and forth. Upon my initial swap, the old chips from my XL worked in my 1088 and vice versa. It wasn't until I swapped them the second time, to get things back where the belonged, that things got a little flakey.

 

I guess I could try my XL's chips in the 1088 again, just to rule out one or both of them being bad.

 

yes, that is a good idea. At the same time, since you might have removed the chips, inspect the sockets very carefully, check if one of the contacts looks a bit misaligned or something. Or if any foot of the atari chips are bent.

[bfollowell]

1 hour ago, manterola said:

yes, that is a good idea. At the same time, since you might have removed the chips, inspect the sockets very carefully, check if one of the contacts looks a bit misaligned or something. Or if any foot of the atari chips are bent.

 

Like I said a couple of posts ago, I've already verified continuity between the shoulders of all the pins to the solder joints on the bottom of the board and on to the next component. I really don't see how it could have anything to do with the socket when continuity is good from the chip on out to the other components.

 

Also, I've verified again that the old chips work fine in my new 1088. It's only in  the 800XL that I get the green screen.

 

[Sugarland]

Have you tried the hot chips test? Leave it on for a few minutes and see if any RAM or other chips get abnormally hot.

[_The Doctor__]

check for debri and old solder flakes etc. you might have knocked, dropped, or had a hitchhiker when transferring parts around.

then once the machine warms up a little, use the alcohol, aeroduster(freeze spray upside down), touch method to ferrit out shorted chips.

the same method can show chips that never change temperature either(open).

so make sure nothing got on, in or under your xl's pcb. nothing bent or bridged. I still think another know working Atari would be helpful. Your 1088 was having issues wasn't it? or is it still?

[manterola]

40 minutes ago, bfollowell said:

Like I said a couple of posts ago, I've already verified continuity between the shoulders of all the pins to the solder joints on the bottom of the board and on to the next component. I really don't see how it could have anything to do with the socket when continuity is good from the chip on out to the other components.

I did not read that post you mentioned that. The only thing i can think of, is to keep checking the other sockets. Another one, is if you have acces to a logic probe, use that to go thru the pins of all major chips. Finally, and this is ideal case, is to get an syscheck.

I've been in your position before, in which a working system stop working after some unrelated modification. I wasted weeks, testing, getting new chips, changing eproms, , to find out that sockets were the problem, or in other cases it was memory or the delay line. 

I wish I would have a syscheck at that time, but I considered too expensive for dealing with a single problem. Now, I am going to get one as soon as I have the opportunity, since I plan to keep playing and messing around with Atari circuits and modifications. Not sure if that is your case.

Anyway, I am sure you will eventually find the problem even if you are going "blind", with the help of the forum and by going thru the testing and schematics very thoroughly.

 

 

[bfollowell]

48 minutes ago, Sugarland said:

Have you tried the hot chips test? Leave it on for a few minutes and see if any RAM or other chips get abnormally hot.

No, I haven't tried that, but this issue happens instantly upon power up, so I'd think it's not related to a chip heating up over time. It would be easy enough to try though. If I don't find anything else obvious, I'll give that try. Sort of like, if you keep blowing a fuse, keep putting in bigger fuses until the bad component completely burns up, then you'll know where the problem was! I know guys at work that feel that is valid troubleshooting method!

 

39 minutes ago, _The Doctor__ said:

check for debri and old solder flakes etc. you might have knocked, dropped, or had a hitchhiker when transferring parts around.

then once the machine warms up a little, use the alcohol, aeroduster(freeze spray upside down), touch method to ferrit out shorted chips.

the same method can show chips that never change temperature either(open).

so make sure nothing got on, in or under your xl's pcb. nothing bent or bridged. I still think another know working Atari would be helpful. Your 1088 was having issues wasn't it? or is it still?

I haven't noticed any sort of debris, but I haven't been specifically looking for any either I'll have to check it out. A good spray with some compressed air wouldn't hurt either. Yes, my 1088 is still exhibiting some minor issues, like hanging on boot until I press a key, but more or less fine after that. The main thing is, my old chips work in it and don't exhibit any of the same behavior they do in my 800XL, so I wouldn't think they were bad. Likewise, since I have good continuity from the pin shoulders on down to the next component, I wouldn't suspect a socket issue.

 

I get what you're saying on having another known good Atari, but it just isn't an option for me. I'm a old user, not a collector. I have no desire to ever be a collector. I don't have 37 other Ataris to check with. I just have my original 800XL and my 1088, that's it. I'm certainly not planning to buy another one just to have it around for troubleshooting the one I have acting up now. Honestly, when I get my 1088 going well, and I'm confident it's at 100%, I'd planned to sell off my 800XL.

 

22 minutes ago, manterola said:

I did not read that post you mentioned that. The only thing i can think of, is to keep checking the other sockets. Another one, is if you have acces to a logic probe, use that to go thru the pins of all major chips. Finally, and this is ideal case, is to get an syscheck.

I've been in your position before, in which a working system stop working after some unrelated modification. I wasted weeks, testing, getting new chips, changing eproms, , to find out that sockets were the problem, or in other cases it was memory or the delay line. 

I wish I would have a syscheck at that time, but I considered too expensive for dealing with a single problem. Now, I am going to get one as soon as I have the opportunity, since I plan to keep playing and messing around with Atari circuits and modifications. Not sure if that is your case.

Anyway, I am sure you will eventually find the problem even if you are going "blind", with the help of the forum and by going thru the testing and schematics very thoroughly.

 

 

I wouldn't mind getting a logic probe. I had one in college decades ago, and have no idea what happened to it. Haven't seen it in 25+ years. I may have to look into another one.

 

As far as Sys-Check, I'm not certain how it would help me since I can't get video out of my machine. I thought it was a cartridge that plugged into the machine and helped in troubleshooting. It seems like it would be sort of a hindrance if you couldn't see the menus or options or feedback or whatever. Maybe I'm confused on how it works though. Also, once I get this initial problem fixed, and get my 1088 going I'll probably sell off my 800XL, and I doubt, but don't know, that the Sys-Check would work with my 1088. Anyway, without being able to see anything, I fail to see how the Sys-Check would be helpful in my current condition.

 

Thanks for the suggestions guys! I'll keep on plugging away at it.

 

 

[Sugarland]

Bad RAM will fail instantly but will take several seconds to get hot. It's not about overheating it's that often when RAM fails it will get abnormally hot due to failure.

[+kheller2]

Ignore me if this is a stupid comment... but if you checked ALL the chips that go into the 1088 by swapping them from the 800XL.. the only thing really left in the XL is OS/BASIC and some glue chips (also assuming RAM isn't the issue).  (Syscheck plugs into the PBI and can take over the OS if need be, btw)  So perhaps one of the other chips is damaged, another thing to test is make sure you test this outside the case, outside the shield on paper or anti-conductive pad.  I've had a few XL's flake out because the solder tails were touching the bottom shield with a missing paper slip.

 

[Keatah]

A 100% faulty chip may take a few moments to heat up. It fails on power up and never works.

[bfollowell]

50 minutes ago, kheller2 said:

Ignore me if this is a stupid comment... but if you checked ALL the chips that go into the 1088 by swapping them from the 800XL.. the only thing really left in the XL is OS/BASIC and some glue chips (also assuming RAM isn't the issue).  (Syscheck plugs into the PBI and can take over the OS if need be, btw)  So perhaps one of the other chips is damaged, another thing to test is make sure you test this outside the case, outside the shield on paper or anti-conductive pad.  I've had a few XL's flake out because the solder tails were touching the bottom shield with a missing paper slip.

 

I didn’t swap out ALL chips. I swapped out the graphics related chips, mainly ANTIC and GTIA.

 

And the 800XL is sitting bare on my work bench at the moment but thanks for the suggestion.

[+bob1200xl]

The screen is green because the GTIA and ANTIC have not been reset and initialized. You'll hear a little 'pop' when the POKEY gets initialized, also. If nothing gets reset, it means that the system is not executing code. CPU, MMU, ROM - basic guts of the system have to work.

 

Computers fails all the time when you turn them on. Yours happened to fail when you were fooling with it. No matter. Swap all the 'big chips'.

 

Unless you have test equipment, you are limited to what you can see. Which isn't much. I will say the most failures are the ICs, but 'most' is just more than half...

 

Bob

[bfollowell]

I did pickup a new logic probe, my first in about thirty years. One thing I did notice that I thought was strange is that all of the data lines are stuck low. I really haven't had the chance to look back any further than that, but it  didn't seem normal.

[_The Doctor__]

I've seen bad memory do that, as well as cpu do that.

[bfollowell]

I'm thinking CPU, but I'm not really basing that on anything. Just a feeling.

 

Something else that I just sort of remembered. My work bench is in the garage. I've got a nice 12-outlet power strip plugged at bench level that is plugged into one of my garage receptacles. Like most newer construction in the last thirty years or so, my garage receptacle circuit is protected by a GFCI, the lead receptacle, not a  GCFI breaker. Right before the problems started, I remember that GFCI tripping a couple of times, and I couldn't figure out why. We built our house and have only been in it five years, so I really didn't suspect a weak or old GFCI receptacle. After a couple of resets and further trips after about 1-2 minutes of power, I unplugged my bench power strip and the GFCI held. I unplugged everything that had been plugged into the power strip and checked everything out closely. I couldn't find any reason for it. Then I plugged the strip back in and slowly started plugging things back in and waited about five minutes between devices, thinking maybe I'd find the problem device. I plugged my 800XL back in last. I've not had another trip since, but an hour or two later was when I tried to fire my 800XL back up after my ANTIC/GTIA swap with my 1088XLE and I noticed it wasn't working. I don't have one of those massive brick style PSUs that are so prone to blowing and taking the computer with it when it goes, and my PSU seems to be putting out good, steady voltage at the moment. Still, I'm wondering now if something on my 800XL could've caused the ground fault when it blew. If it did, there've been no further ground faults, but my 800XL is toast. I guess my concern is, if/when I find the problem component, I'm almost afraid to just swap it, thinking something further back may have taken it out and will do the same to my replacement.

 

Anyway, I'm only working onsite one day this week, and from home the rest, so hopefully I'll have time for some more troubleshooting. I'll move my focus to CPU and RAM for now.