XEGS red/brown screen of death.

[TZJB]

18 hours ago, deffroe said:

So...only tested A4 Sally to Freddie and Freddie to antic.  Sally and Fred are fine but antic only has continuity to the socket pin on the underside and NOT to the top side of socket or chip leg.

I'll have a quick play tonight and post an update either tomorrow or the weekend.

Think I've just seen a glimmer of light.🤞👍 

 

Excellent. You are getting closer to solving this one.

[+DrVenkman]

20 hours ago, deffroe said:

Sally and Fred are fine but antic only has continuity to the socket pin on the underside and NOT to the top side of socket or chip leg.

Remove ANTIC and take a very close look at the socket wipes - look for any that are obviously oxidized or corroded; bent or potentially even broken. Focus on the contact for the relevant signal of course, but examine them all. If one is damaged or oxidized, it may not be the only one. 

[deffroe]

I removed the antic socket in case it was bad and took the opportunity to inspect the PCB clearly in that area.

Took a couple of photos using an illuminated magnifying glass(definitely illuminated something🥴).

There is some crud visible in some of the holes but it was only a bit of flux.

I fitted another dual-wipe socket and the result was the same, got some machined sockets on order to replace that.

 

 

[TZJB]

5 minutes ago, deffroe said:

I removed the antic socket in case it was bad and took the opportunity to inspect the PCB clearly in that area.

Took a couple of photos using an illuminated magnifying glass(definitely illuminated something🥴).

There is some crud visible in some of the holes but it was only a bit of flux.

I fitted another dual-wipe socket and the result was the same, got some machined sockets on order to replace that.

 

 

For the result to be the same, the through hole may not be working. Looking at your images, if I have counted correctly, there is no trace to pin 28 on the underside of the board, so you are relying on the solder flowing into the hole to join up the top trace. Is there definitely no continuity on the socket?

[deffroe]

4 hours ago, TZJB said:

 

For the result to be the same, the through hole may not be working. Looking at your images, if I have counted correctly, there is no trace to pin 28 on the underside of the board, so you are relying on the solder flowing into the hole to join up the top trace. Is there definitely no continuity on the socket?

Think I've found the problem.  There was no continuity between top and bottom of hole, added a couple of strands of copper to it and soldered in a socket just at that pin, checked continuity again and still nothing.  Began tracing pin 28, it runs to a point just inside the chip outline and splits up to cartridge slot and down to Sally.  Had to remove her socket to see exactly where and came across a "Scrape" which I saw when I first started playing with this board but didn't think much of.  Upon closer inspection, there is a break in the trace to pin 13 of Sally.

The glimmer of light just got brighter. A job for tomorrow me thinks!

👍

[deffroe]

[_The Doctor__]

Classic prying damage.

[deffroe]

11 hours ago, _The Doctor__ said:

Classic prying damage.

I'm not sure it is.  When I first inspected the board I checked the underside for any signs of rework and it looked clean.  All IC's, as would be expected, were not socketed. So, if Sally had already been removed, a lot of care and skill had been used to solder her back in, but, not enough care to add a socket.

I certainly don't try to wedge anything under a chip when desoldering that would cause damage like that.  My initial thought when I saw it was, it must be factory.  I only wish I'd paid it a bit more attention as I'm sure I'll find out later today that that is the only issue with the board.

 

Phil

[deffroe]

Success!!!! Briefly... Managed to bridge the break to pin 13 of Sally, checked continuity to the other pins potentially affected, all good.

Installed new sockets to Sally and antic, checked continuity to pin 28 of antic, yey!

Popped in the sockets, powered on with RF and there was missile command ...for a second then screen turned blank.  Tried turning off and on again a few times with same result.  A few tries later and it now boots to self test and again after a second or 2 the screen goes blank but retains the colour of self test.

[Rybags]

So, a fair bit of progress then.

If it's getting that far then the buses should be OK.  Strange that you'd get a few seconds of what looks to be full functionality then it just dies.

Can you try Basic?  IIRC if you have no keyboard, holding Select should bring it up on coldstart.

[Beeblebrox]

I wonder if there is a solder point on a socket that isn't quite right. Have you tried leaving the machine off for a while and powering up? See if missile command stays on screen longer on first boot after the machine has been offer for a while. 

 

Also what type of sockets have you installed? 

[flashjazzcat]

17 hours ago, deffroe said:

What happened to those gouged traces at the right hand edge of the socket (bottom of pic)?

 

EDIT: Doesn't matter - I assume it's what you already referred to as a scrape. Obviously someone's mangled it taking the CPU out at some time or another.

Edited June 18, 2023 by flashjazzcat

[+DrVenkman]

23 hours ago, deffroe said:

I fitted another dual-wipe socket and the result was the same, got some machined sockets on order to replace that.

Be EXTREMELY careful with repeated desoldering and resoldering of the same pads - these are consumer boards, and not particularly durable consumer boards at that. Repeatedly removing and replacing soldered components from the same pads runs risk of delaminating the board, pulling out vias and lifting traces just from repeat thermal cycles. Dual-wipe sockets are already 100% better than stock and handle removing and reinstalling socketed chips very well due to all that surface area to touch both sides of the chip legs.

[deffroe]

Hi all,  work commitments got in the way of continuing with this for a while, so, quick update...after my brief success the machine has reverted back to booting to red/brown screen  .

On 6/18/2023 at 2:59 PM, deffroe said:

Popped in the sockets, powered on with RF and there was missile command ...for a second then screen turned blank.  Tried turning off and on again a few times with same result.  A few tries later and it now boots to self test and again after a second or 2 the screen goes blank but retains the colour of self test.

One evening last week, when I was able to have a tinker for half an hour, I was probing the chips with my handheld scope.  Touching pin 1 of Freddie killed the video, going from brown screen to white noise.  Removing probe and video would return.

 

Moving on to today, I connect the cables and plan to make notes of every chips pin activity/frequency/anything worthwhile.  Hit power and cue the Atari drum roll and lovely blue screen as if booting to basic but then goes to Self Test.  Then there was a thud as my chin hit the desk 😮.  So, again, for a few brief seconds I had sound and graphics all working but then the screen garbled and went blank BUT retained Self Test green.

 

Any suggestions where to look?

 

Booted it again before submitting post and back to brown screen.  aggghh!!

[_The Doctor__]

Yes, be careful not to short things when probing about, and you also may have an intermittent connection that happens as you apply pressure to probe, so bad sockets and dirty connections, clean make sure no shorts/cracks(slightly open) or bad sockets. Wiggle things and see what happens

Edited July 12, 2023 by _The Doctor__

[deffroe]

19 minutes ago, _The Doctor__ said:

Yes, be careful not to short things when probing about, and you also may have an intermittent connection that happens as you apply pressure to probe, so bad sockets and dirty connections, clean make sure no shorts/cracks(slightly open) or bad sockets. Wiggle things and see what happens

Here's a little clip, you can see the breaking and killing of video when I touch the probe to pin 1 and returning once I remove probe, incidentally, same thing happens on pin 2 as well.  Pin1 gives a frequency of 10Mhz.

 

All looks good with socket and solder points, giving it a wiggle shows no difference.

[TZJB]

3 hours ago, deffroe said:

Here's a little clip, you can see the breaking and killing of video when I touch the probe to pin 1 and returning once I remove probe, incidentally, same thing happens on pin 2 as well.  Pin1 gives a frequency of 10Mhz.

 

All looks good with socket and solder points, giving it a wiggle shows no difference.

 

Freddie pin 1 is an unused clock output for the Atari XE computer line so no point in checking that.

 

Freddie Pin 2 on the other hand should show as the main 14.187576 MHz clock input. Please check this is the frequency seen at pin 2.

 

This should then produce the 3.546894 MHz output on Freddie pin 37 OSC.

Edited July 12, 2023 by TZJB

[deffroe]

10 hours ago, TZJB said:

Freddie Pin 2 on the other hand should show as the main 14.187576 MHz clock input. Please check this is the frequency seen at pin 2.

 

This should then produce the 3.546894 MHz output on Freddie pin 37 OSC

Checked pin 37 first, 3.54Mhz.  pin 2 showed nothing, no frequency or activity.  Realised I'd not turned my screen on and the following screen was present...

 

Pressing reset blanked the screen, powering off and on, returned the brown screen.

Pin 2 gives the same video result as pin 1.

[Beeblebrox]

@deffroeWhen you get your blue screen above do you have audio connected? If not, make sure you do and have the volume turned up. What do you hear, if anything?

 

 

Edited July 13, 2023 by Beeblebrox

[TZJB]

1 hour ago, deffroe said:

Checked pin 37 first, 3.54Mhz.  pin 2 showed nothing, no frequency or activity.  Realised I'd not turned my screen on and the following screen was present...

 

Pressing reset blanked the screen, powering off and on, returned the brown screen.

Pin 2 gives the same video result as pin 1.

 

I don't think that your 'scope will detect a 14 MHz frequency so we will asume it to be good as your OSC is 3.54 MHz.

 

I would suspect that there is still a bad connection/continuity on at least one of the pins of the CPU or other large IC that has previously been socketed, which is most of them.

 

You may need to painstakingly continuity check each trace between each chip to find the problem.

 

Just don't play with Freddie pin 1 or 2. 😁

[deffroe]

55 minutes ago, Beeblebrox said:

@deffroeWhen you get your blue screen above do you have audio connected? If not, make sure you do and have the volume turned up. What do you hear, if anything?

 

 

That was the first time seeing that screen, connected by RF.  There was no audio heard and machine was already booted when I turned in display.

[deffroe]

3 minutes ago, TZJB said:

You may need to painstakingly continuity check each trace between each chip to find the problem.

That was my plan when I got the unexpected Self Test.  I'll start the process this afternoon 👍.

[Beeblebrox]

1 hour ago, deffroe said:

That was the first time seeing that screen, connected by RF.  There was no audio heard and machine was already booted when I turned in display.

sure. If you have a CRT display you'll see and hear things that might have occured by the time a modern LCD kicks in and you'd therefore miss.

 

I'd advise either:

 

Hooking up via RF to an old CRT TV

Or

hooking up to a composite cable to your modern LCD but hook up the audio to external speakers.

 

Basically what I am thinking is to see if  - where you can get the blue screen up - either you hear the Atari fart sound at all, (doubt it), or any audio blips.

 

Trying to see if the OS chip is coming online.

 

I've not 100% caught up with the replies in this thread but it certainly looks like you have several problems here all contributing, so the issues are layered.

 

In my experience the audio cues are sometimes as important as the visual ones.

 

Can you pull the Basic chip as the machine can still boot with it removed?

 

My pennies worth. I am thinking aside the previously mentioned excellent advice from others, you could have issues with one or more of the following (chips themselves or their sockets/solder connections):

 

OS Rom

Basic rom

one/both of the graphics chips

Pokey

 

Finally make sure you have a mains powered, (ideally original Atari) PSU and not a modern switching PSU or USB power. If you are using USB make sure you have a decent adapter and one that has 5v 2amp. TBH I don't think it is a contributing factor but worth ensuring your power is clean/reliable.

 

 

Edited July 13, 2023 by Beeblebrox

[deffroe]

9 hours ago, Beeblebrox said:

Hooking up via RF to an old CRT TV

Or

hooking up to a composite cable to your modern LCD but hook up the audio to external speakers.

This is not possible I'm afraid, don't have a CRT.

9 hours ago, Beeblebrox said:

where you can get the blue screen up - either you hear the Atari fart sound at all,

 

On 7/12/2023 at 12:34 PM, deffroe said:

Hit power and cue the Atari drum roll and lovely blue screen as if booting to basic but then goes to Self Test. 

When I posted this the other day, the video was perfect(except it should have booted to missile command) the Atari drum roll/fart sound sounded correct.  This was a totally unexpected outcome and lasted all of 10 seconds before glitching and blanking to Self Test green.  Rebooting sent it back to brown screen.

 

That light blue screen from earlier today, first and only time that has happened/been seen.

Spent a couple of hours this afternoon checking continuity of Sally, Fred, Antic and GTIA following the diagrams posted by TZJB and the 130XE Hardware manual and everything seems OK...but it isn't  

 

Tomorrow I will try swapping out chips from another machine from my collection to confirm again that they are not the problem.

 

I'm done for tonight!

[+DrVenkman]

6 minutes ago, deffroe said:

When I posted this the other day, the video was perfect(except it should have booted to missile command)

Something's up with your GTIA and/or PIA signals for sure.