XEGS red/brown screen of death.

[deffroe]

While I've been sat tapping out posts I've left the GS running a good 20-30 mins and got the effect shown in the video.  Bit of white noise/interference coming through occasionally.

[deffroe]

8 minutes ago, flashjazzcat said:

Grab a pinout via Google and check the address and data pins. You should see some activity there. We know the system crashed early on since it didn't get far enough through the OS initialisation to even reset the screen background colour to $00.

The address pins show activity except pin 13 and moving up from pin 21 through to 40 everything shows activity except pin34.

 

Phil

[TZJB]

 

1 hour ago, deffroe said:

Evening all,

I get 3.54Mhz at both pin locations.

 

Is this pin7 on Sally?  If so, it just shows a steady 5v.

 

One thing I forgot to mention, and I wanted to wait til I got home to double check, inserting Star Raiders shows no difference.

 

Something else I've tried this evening, not sure if it help with diagnosing, I connected a Sophia 2 DVI.  The TV picks up a signal but only shows a black screen.

 

OSC looks good so check PH0 at Sally pin 37. This 1.79 MHz clock from Sally syncs everything else.

 

53 minutes ago, deffroe said:

The address pins show activity except pin 13 and moving up from pin 21 through to 40 everything shows activity except pin34.

 

Phil

 

Sally Pin 34 in not connected. And I am not sure that I understand the relevance of Sally pin 7 Sync as that isn't connected to anything either. Pin 13 is Bus Address 4 and if it's frozen something on the bus is causing it to be static. Sally Pin 21 and pin 38 are ground so I hope that there is no activity on those pins.

 

Your Sally shows Mexico markings, do you have one of an alternative manufacturer? Also did you change the OS ROM as this has an Add 4 connection too?

[DrVenkman]

51 minutes ago, TZJB said:

OSC looks good so check PH0 at Sally pin 37. This 1.79 MHz clock from Sally syncs everything else

Unless I'm misremembering, Phi0 is generated by ANTIC. 

OSC goes to GTIA, which then produces Fast Phi0, which feeds ANTIC. ANTIC then generates Phi0 and feeds it SALLY as an input. SALLY then generates Phi1 (which is part of the video color phase delay) and Phi2. Phi2 is what syncs up pretty much everything else in the system together after being buffered by the 74LS08 chip.

 

 

[TZJB]

22 hours ago, DrVenkman said:

Unless I'm misremembering, Phi0 is generated by ANTIC. 

OSC goes to GTIA, which then produces Fast Phi0, which feeds ANTIC. ANTIC then generates Phi0 and feeds it SALLY as an input. SALLY then generates Phi1 (which is part of the video color phase delay) and Phi2. Phi2 is what syncs up pretty much everything else in the system together after being buffered by the 74LS08 chip.

 

 

 

Aye, of course you are correct. This is what happens when I do this late at night.

 

Ph0 is an input to Sally pin 37 from Antic. Ph2 is then produced from it on Sally pin 39 and does as you say. Apologies and thanks for the correction.

 

Do you know what Sally pin 7 Sync would indicate?

[DrVenkman]

6 minutes ago, TZJB said:

Do you know what Sally pin 7 Sync would indicate?

According to SAM'S ComputerFacts for the 130XE, it should pulse but it's not connected.

 

[Rybags]

The Sync pin is used to indicate an instruction fetch.

Normally this goes unused though the arcade Missile Command puts it to good use.

The bitmap graphics are packed similarly to how a normal linear mode would do so (I think it's a mix of 2 and 4 bpp depending on vertical position) - as such you'd normally have that annoyance of needing to do bitshifts and masking.  But the machine has extra hardware that detects certain instructions [mostly (IND,X) ] then modifies the subsequent memory access to provide a full byte, lsb justified to the CPU.

 

I've read of other uses but that's the only one that comes to mind at the moment.

Another obscure pin is SO - Set Overflow.  This can be used with external hardware to provide a quick polling facility.  That usage case is also why we have a CLV instruction but not SEV.

Edited June 14, 2023 by Rybags

[_The Doctor__]

6 hours ago, Rybags said:

The Sync pin is used to indicate an instruction fetch.

Normally this goes unused though the arcade Missile Command puts it to good use.

The bitmap graphics are packed similarly to how a normal linear mode would do so (I think it's a mix of 2 and 4 bpp depending on vertical position) - as such you'd normally have that annoyance of needing to do bitshifts and masking.  But the machine has extra hardware that detects certain instructions [mostly (IND,X) ] then modifies the subsequent memory access to provide a full byte, lsb justified to the CPU.

 

I've read of other uses but that's the only one that comes to mind at the moment.

Another obscure pin is SO - Set Overflow.  This can be used with external hardware to provide a quick polling facility.  That usage case is also why we have a CLV instruction but not SEV.

Those are interesting and as mods that could be fun. Would be a nice to touch to add them into the 8 bit line.

Edited June 14, 2023 by _The Doctor__

[Rybags]

Fairly sure SO is tied to ground on all models.  Though nothing stopping someone from modding it to accept an external input.

The memory access trick in MC - that would involve a reasonable amount of external logic.

As I understand it from reading some MAME driver docs a few years ago, it allows pixel addressing by full byte and the hardware takes care of the shifting and masking that the program would otherwise need to waste lots of cycles doing.

[deffroe]

Was going to leave doing anymore until the weekend but sat down at my likkle desk and it was screaming at me for some attention..

Checked Sally pin 39 and get 1.77Mhz, same with pin 37.  Pin 36 has a VP of 3.6-4v with a bit of activity. A chunky square wave on pin 35, nothing on pin 34.

Still red screen.

[DrVenkman]

40 minutes ago, deffroe said:

Checked Sally pin 39 and get 1.77Mhz, same with pin 37.  Pin 36 has a VP of 3.6-4v with a bit of activity. A chunky square wave on pin 35, nothing on pin 34.

Pin 39 is Phi2 (output from SALLY), and pin 37 is Phi0 (input to SALLY, generated by ANTIC). Pin 36 is R/W (Read/Write), so you should expect various degrees of pulsing indicate the system trying to read and write to the data and address busses. Pin 35 is the HALT line, which is normally held High until ANTIC is trying to access to bus and draw the screen, when it is pulled low. So the pulse means ANTIC is basically working since it's accessing memory and (via GTIA) drawing a stable solid-color screen. Pin 34 is not connected so you won't see activity there.

 

[deffroe]

Discovered a bit of odd behaviour.  If I boot with just RF connected I get red/brown screen, if I boot with both RF and composite I can flick between the two with same red screen, if I boot with just composite I get NO signal.

[_The Doctor__]

If you are absolutely sure Ram, Capacitors, and MMU are good, you can always try the 4050 and related video transistors

[Beeblebrox]

@deffroeI just did a quick scan of this post and not sure but I don't think it's been mentioned. Have you powered up with audio connected at all and volume turned up? If so try it. Do you hear anything in terms of little pops, usually two?  Irrc it indicates the os rom is at least trying to kick in. (I am mobile so can't cross check this). 

 

Also (again not sure if mentioned) if you have a CRT then use that instead of an lcd. The lcd takes a few more seconds sometimes to display and you can miss other initial in screen events that might flash up beforehand before the red/brownish screens. Can flash bars or other colour screens briefly which itself can indicate things. 

 

I am movile

 

 

Edited June 15, 2023 by Beeblebrox

[DrVenkman]

OP’s problem does not sound like a video issue, but one way to find out is to power the Atari on while holding START - that will trigger a cassette boot and if the system is working but with bad video, he should hear the cassette loading sound/prompt. 

 

That said, I don’t believe that this is a video issue. It sounds very much to me like the system is “alive” (the OSC and clock signals are there, and the system is trying to boot up) but something is preventing it from completing. About 6 years ago, I resurrected a “red screen” 1200XL that turned out to have several bad sockets. But the first one I discovered was a bad connection on the Phi2 pin of POKEY. The socket it was installed in had a physically broken wipe (contact) and the chip wasn’t making contact on that pin. Once I fixed that, the system would boot but was unstable. I ended up finding bad/damaged wipes on the sockets for SALLY and PIA, too, so those got replaced as well. The system works great today. 

[Beeblebrox]

8 minutes ago, DrVenkman said:

OP’s problem does not sound like a video issue, but one way to find out is to power the Atari on while holding START - that will trigger a cassette boot and if the system is working but with bad video, he should hear the cassette loading sound/prompt. 

 

That said, I don’t believe that this is a video issue. It sounds very much to me like the system is “alive” (the OSC and clock signals are there, and the system is trying to boot up) but something is preventing it from completing. About 6 years ago, I resurrected a “red screen” 1200XL that turned out to have several bad sockets. But the first one I discovered was a bad connection on the Phi2 pin of POKEY. The socket it was installed in had a physically broken wipe (contact) and the chip wasn’t making contact on that pin. Once I fixed that, the system would boot but was unstable. I ended up finding bad/damaged wipes on the sockets for SALLY and PIA, too, so those got replaced as well. The system works great today. 

Don't think your reply was following/realted to my reply, but yeah I agree, I don't think it is a video issue either.   (Just pointing out with audio connected and a CRT rather than an LCD you are more likely to pick up on some indicators at power on to indicate possible single or multiple issues). I had a similar issue with PIA and Pokey on 65XE and 130XE machines IIRC.

 

 

[TZJB]

To point out the elephant in the room, if Bus Address 4 is not active at Sally pin 13, activity and continuity between all chips with Bus Address 4 needs to be checked and confirmed. There may be a dis or a short on the bus, or a chip may be causing this.

 

This involves Sally pin 13 already stated, Antic pin 28, GTIA pin 38, Freddie pin 12 and OS ROM pin 6.

[flashjazzcat]

On 6/12/2023 at 10:13 PM, deffroe said:

Here's a couple of better less blurry photos.

Can you take some of the underside of the board as well? That MMU socket and one or two others look a bit troubled. It's perfectly possible to introduce a secondary problem when socketing a board whose symptoms are identical to the primary issue one is attempting to fix.

[deffroe]

Here's an underside shot of the board.

[deffroe]

Just tried booting with composite and audio only and it boots with the erratic screen below...and as before, just composite= no video signal, add RF and boots with red screen.

 

[deffroe]

4 hours ago, TZJB said:

Bus Address 4 is not active at Sally pin 13, activity

Pin 13 is held high with a Vp of 3.6v

[flashjazzcat]

1 hour ago, deffroe said:

Here's an underside shot of the board.

Thanks. The underside looks good (assuming this isn't a bridge near the RAM, which I guess it isn't):

 

[deffroe]

52 minutes ago, deffroe said:

Pin 13 is held high with a Vp of 3.6v

I've been able to compare pin 13 on my working GS and can see activity on that that isn't see on this GS.

[deffroe]

4 minutes ago, flashjazzcat said:

The underside looks good (assuming this isn't a bridge near the RAM, which I guess it isn't

No not a bridge.  Gave it a clean though but same erratic screen on composite a/v.

[deffroe]

6 hours ago, TZJB said:

To point out the elephant in the room, if Bus Address 4 is not active at Sally pin 13, activity and continuity between all chips with Bus Address 4 needs to be checked and confirmed.

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.🤞👍