800XL constant beep at power on

[c1ph3rpunk]

New to the Atari world, picked up an 800XL that had been reported working, but alas it’s not. When powered on I see a brief flicker on the screen, and what appears to be some green color, then it goes black and produces a constant beep. New power supply from 8-bit Classics, new video cable from them as well, same results on S-Video or composite.

 

Opened it up and visually inspected everything, nothing appeared damaged or blown, there are however signs of prior work on the OS ROM, a RAM chip and the parallel port chip (can’t remember the name), there’s obvious flux remaining on the bottom of the board on those chip pins. Most of the chips are socketed, including the RAM chips, all except the ROM ones and the parallel port one. Oddly, if someone removed/replaced the OS ROM chip, no idea why they didn’t socket it. Checked voltages coming in, it’s in the 5.1V range, appears nominal. Made sure all the chips are in the socket, a couple felt loose, tested with same results. Left the system on, with the volume down, for a while, 30 minutes or so, none of the chips feel abnormally hot, warmest one is the CPU.

 

Put the CPU on the scope and all 3 clock signals appear good, including the inverted one. Checked address and data bus there, and on the RAM chips, and activity is present. Went around and checked Vcc on all chips and all tested fine. Reset button on the keyboard appears to reset, at least when the scope is on it looks like it.

 

Read the service manual and nothing seems to match the “constant beep” condition I’m currently at, is that a sign of anything in particular?

 

Wondering if these next steps make sense:

 

1) Pull all socketed chips, de-oxit, check socket pins for continuity and re-seat all of them. Test.

2) If it persists, re-flow/clean all the socket pins that appeared to have work done. Test. I’m debating about just doing that anyways, the old yellow-brown flux mess is annoying me, prudence might force me there anyways.

 

From there I’m not sure on the 800XL to go next, help appreciated. TIA.

 

 

[TZJB]

23 hours ago, c1ph3rpunk said:

New to the Atari world, picked up an 800XL that had been reported working, but alas it’s not. When powered on I see a brief flicker on the screen, and what appears to be some green color, then it goes black and produces a constant beep. New power supply from 8-bit Classics, new video cable from them as well, same results on S-Video or composite.

 

Opened it up and visually inspected everything, nothing appeared damaged or blown, there are however signs of prior work on the OS ROM, a RAM chip and the parallel port chip (can’t remember the name), there’s obvious flux remaining on the bottom of the board on those chip pins. Most of the chips are socketed, including the RAM chips, all except the ROM ones and the parallel port one. Oddly, if someone removed/replaced the OS ROM chip, no idea why they didn’t socket it. Checked voltages coming in, it’s in the 5.1V range, appears nominal. Made sure all the chips are in the socket, a couple felt loose, tested with same results. Left the system on, with the volume down, for a while, 30 minutes or so, none of the chips feel abnormally hot, warmest one is the CPU.

 

Put the CPU on the scope and all 3 clock signals appear good, including the inverted one. Checked address and data bus there, and on the RAM chips, and activity is present. Went around and checked Vcc on all chips and all tested fine. Reset button on the keyboard appears to reset, at least when the scope is on it looks like it.

 

Read the service manual and nothing seems to match the “constant beep” condition I’m currently at, is that a sign of anything in particular?

 

Wondering if these next steps make sense:

 

1) Pull all socketed chips, de-oxit, check socket pins for continuity and re-seat all of them. Test.

2) If it persists, re-flow/clean all the socket pins that appeared to have work done. Test. I’m debating about just doing that anyways, the old yellow-brown flux mess is annoying me, prudence might force me there anyways.

 

From there I’m not sure on the 800XL to go next, help appreciated. TIA.

 

 

 

Welcome to Atariage.

 

A fully socketed 800XL normally means that extensive work has been done on the board as sockets are unlikely to be installed at manufacture, unless there was a chip shortage, in which case they may well be of the single wipe type.

 

From what you have written, you seem to be an experienced diagnostician and you have correctly checked the basics. However did you check for any chips getting hot?

 

With this type of fault it is always helpful to have another working board to swap chips into, but failing that it would be helpful if you could upload some high resolution images of your 800XL board top and bottom. Sometimes we can suggest or see something helpful.

 

If the CPU or other large chip were faulty there would be, more than likely, Data or Address bus signals frozen. Please re-check these at SALLY, ANTIC, GTIA, and POKEY, noting each one, checking for static or missing lines. The OS and BASIC sit on these bus lines too and either could also be a culprit.

 

The POKEY chip provides the sound and it may be processing false data which hopefully you can see on your 'scope.

 

Please find attached some schematics, which may help if you only have the Service Manual, and let us know the outcomes.

 

Atari800XL_schematics_rev2.pdf

 

 

 

 

[+MrFish]

43 minutes ago, TZJB said:

Please find attached some schematics

Do you know who the authors of these schematics are?

 

[+kheller2]

On 9/21/2023 at 7:08 PM, TZJB said:

 

Welcome to Atariage.

 

A fully socketed 800XL normally means that extensive work has been done on the board as sockets are unlikely to be installed at manufacture, unless there was a chip shortage, in which case they may well be of the single wipe type.

 

 

 

That’s isn’t exactly true.  There are many 800XLs that are mostly, to fully socketed.  It depends when it was manufactured, late 84 into 85 will have a lot less.  XLF boards have hardly any. 

[Beeblebrox]

@c1ph3rpunk where are you based? Sounds like an NTSC machine given your ref to detoxit and 8bit classics. (Not that this has any bearing here).

 

Stock machine? (No upgrades?) 

 

Given ram is socketed I'd replace all ram with new chips anyway as its likely they could still be am issue. Is it the notorious mtram per chance? 

 

Also can you take a photo of your pcb and also was there an original power supply used and when you experienced issues you bought the modern one? 

 

As mentioned at factory some chips may have been socketed and some not. Atari did this with some many. Others (Hong Kong made), will be fully socketed. 

 

Pics of the board will help. 

Edited September 22, 2023 by Beeblebrox

[+x=usr(1536)]

On 9/20/2023 at 5:48 PM, c1ph3rpunk said:

Opened it up and visually inspected everything, nothing appeared damaged or blown, there are however signs of prior work on the OS ROM, a RAM chip and the parallel port chip (can’t remember the name), there’s obvious flux remaining on the bottom of the board on those chip pins.

What you're seeing may also have been someone's attempt at chasing down cold solder joins.  Instead of hitting everything that may have been an issue (and socketing things like the OS ROM in the process), they likely just went after individually-suspect pins.

On 9/20/2023 at 5:48 PM, c1ph3rpunk said:

Left the system on, with the volume down, for a while, 30 minutes or so, none of the chips feel abnormally hot, warmest one is the CPU.

The CPU being a bit warmer is pretty much normal.

On 9/20/2023 at 5:48 PM, c1ph3rpunk said:

Wondering if these next steps make sense:

 

1) Pull all socketed chips, de-oxit, check socket pins for continuity and re-seat all of them. Test.

2) If it persists, re-flow/clean all the socket pins that appeared to have work done. Test. I’m debating about just doing that anyways, the old yellow-brown flux mess is annoying me, prudence might force me there anyways.

Yes on both.

 

This probably won't work, but: if you turn the machine on while holding down OPTION, does it go into the self-test menu or do any other changes take place?

 

One other possibility is a shot MMU / delay line IC, but I don't want to jump to that just yet.

[woj]

I'd also give the board a good clean, you mention some prior work and flux residue, not everyone is as tidy as they should be. I chased problems with my 130XE for almost a month, the last remaining problems that took me over a week sitting on them were due only to gunk on the board.

[TGB1718]

As @Beeblebrox said, can you provide good quality photo's of the front and back of the board, you'd be surprised

how many faults have been diagnosed by members here just by inspecting photo's (and we like doing it  )

[TZJB]

18 hours ago, MrFish said:

Do you know who the authors of these schematics are?

 

 

Unfortunately I do not know the authors name or even where I got them. They are in my 800XL folder and thought that they may be useful.

[+MrFish]

11 minutes ago, TZJB said:

Unfortunately I do not know the authors name or even where I got them. They are in my 800XL folder and thought that they may be useful.

 

Thanks anyway; I appreciate the upload. I see the KiCad docs are dated 2020; the GIF file is dated 1998.

 

[TZJB]

18 hours ago, kheller2 said:

That’s isn’t exactly true.  There are many 800XLs that are mostly too fully socketed.  It depends when it was manufactured, late 84 into 85 will have a lot less.  XLF boards have hardly any. 

 

You may well be correct which is why I gave the caveat of chip shortage. In order to keep an assembly line running it is historically cheaper to fit IC sockets when chips are not available, and fully assemble later when they are, rather than shutting the assembly line down.

[TZJB]

@c1ph3rpunk an old trick you could try is to boot a Star Raiders cartridge which will bypass the onboard OS.

[c1ph3rpunk]

Wow, forgot to check this since my forum join was approved, have some reading and reply work to do, but thanks a ton for all the initial suggestions.

 

I’ll start with the board pics, attached are from when I first opened it up. Since then I did remove the socketed chips, de-ox it’s them, and put them all back in. At the same time I went ahead and cleaned up the previous work, re-soldered many pins and cleaned up the old flux.

 

When powered on now I do get some video, dark blue for a second, sometimes white lines and other times green bars. There’s unfortunately no apparent consistency to this.

 

I then went back to check Vcc all over and when I got to the GTIA chip I get effectively 0V on Vcc. When I check across Vcc and Vss on that chip, it shows -5V, I’m guessing that’s not good.

 

I do have a friend local (Chicago area) that has a working 800XL that I think I’ll visit and try some chip swaps. It seems like the GTIA has … issues. 

[Beeblebrox]

Can you share pics of the various screens you are experiencing? 

 

Also have audio turned up and tell us what you hear on booting up. Is it always a tone? 

 

I'd definitely replace the ram chips even if you think the current mtram chips are OK. 

 

 

[c1ph3rpunk]

I wish the video issue was consistent, here’s one of that, plus the sound, just after I did the most recent work.

 

The video appears after it’s been sitting for a bit, it goes straight black if I turn it on right after powering down. It’s almost like capacitors have to drain before any video will appear again.

 

 

[+x=usr(1536)]

46 minutes ago, c1ph3rpunk said:

I then went back to check Vcc all over and when I got to the GTIA chip I get effectively 0V on Vcc. When I check across Vcc and Vss on that chip, it shows -5V, I’m guessing that’s not good.

Yeah, that's not the best.

 

Can you use a different power supply?  If you happen to have the original one, please post a photo of it here before using it - there's one 110V PSU (the 'Ingot') that's known to kill systems.

 

Any solder bridges and/or splatters anywhere on the board?

 

One thing worth noting: as @Beeblebrox mentioned, you have MT RAM in that machine.  It's known to be not the most reliable.  This may be a contributing factor to the behaviour you're seeing if not outright responsible for it.

 

[c1ph3rpunk]

6 minutes ago, x=usr(1536) said:

Yeah, that's not the best.

 

Can you use a different power supply?  If you happen to have the original one, please post a photo of it here before using it - there's one 110V PSU (the 'Ingot') that's known to kill systems.

 

Any solder bridges and/or splatters anywhere on the board?

 

One thing worth noting: as @Beeblebrox mentioned, you have MT RAM in that machine.  It's known to be not the most reliable.  This may be a contributing factor to the behaviour you're seeing if not outright responsible for it.

 

Sure, here’s the PSU that came with it. I had read about issues with some of them previously so I haven’t used it, waited until I got a new one at VCF from 8-bit Classics.

I had seen what looked like some bridged connections but after cleaning them up I found the traces where they were meant to be. I looked over the board as closely as possible, twice, and haven’t seen anything obvious. I’ll give it a third try, a little closer this time.

 

I was going to buy RAM and likely a GTIA, ANTIC and, if possible, a POKEY (they seem costly now). Is there a known reliable source for them in the US? Searching eBay has results but some of them immediately look suspect.

 

I suppose second question on the RAM, if the Mt ones are known to have issues, is there a more reliable replacement manufacturer?

 

[ivop]

2 hours ago, c1ph3rpunk said:

I then went back to check Vcc all over and when I got to the GTIA chip I get effectively 0V on Vcc. When I check across Vcc and Vss on that chip, it shows -5V, I’m guessing that’s not good.

You sure you didn't have Vcc and Vss reversed? That would explain the -5V reading you got.

[c1ph3rpunk]

6 hours ago, ivop said:

You sure you didn't have Vcc and Vss reversed? That would explain the -5V reading you got.

It must be, the pinout I printed says pin 3 is Vcc, another one shows it as Vss, explains that, thx.

[+x=usr(1536)]

14 hours ago, c1ph3rpunk said:

Sure, here’s the PSU that came with it. I had read about issues with some of them previously so I haven’t used it, waited until I got a new one at VCF from 8-bit Classics.

That one should be OK, but testing it before plugging it into the machine is very much recommended   Pinout is:

 

 

14 hours ago, c1ph3rpunk said:

I was going to buy RAM and likely a GTIA, ANTIC and, if possible, a POKEY (they seem costly now). Is there a known reliable source for them in the US? Searching eBay has results but some of them immediately look suspect.

POKEYs are not cheap these days.  Best Electronics has them, as does The Brewing Academy.  Other than that, outside of taking a chance on a non-working machine and hoping that it yields good ICs the options are fairly slim.

15 hours ago, c1ph3rpunk said:

I suppose second question on the RAM, if the Mt ones are known to have issues, is there a more reliable replacement manufacturer?

A couple of years ago, I picked up somewhere in the region of 384K of mid-1990s manufacture NEC RAM.  The ICs were pulled from unknown sources, but it's been working great.

 

Can't remember where I got it from, but it was probably an electronics surplus site.  Too many counterfeits on eBay / Amazon to make it worth buying it there unless I'm familiar with the seller and their reputation.