Self-Test (boot with Option held) kills Atari 800xl?

[Arcanis-Will]

Just purchased Atari 800XL and power supply separately, both appear in good shape, and initially it booted to READY prompt and all keys worked. When I held Option and rebooted it, the screen came up black (aka not snow, it had a signal) and stayed black through successive power cycles and resets. How do I get out of self-test mode? I don't find any mention of needing to exit self-test mode but it is acting like the option-boot persists across power cycles. Does self-test take a long time to display anything? Maybe I am not waiting long enough. Very confusing that it worked until I booted this way and now appears permanently "stuck". As an aside, I have a 400 that shows the same symptom - power-up to a black screen - so that makes me wonder if it is in some weird boot mode but I find no mention of it if so.

 

 

[+bcombee]

I wonder if it's actually the Option key that's broken and now always indicating that it's pressed. You could try booting the 800XL with the keyboard unplugged internally and see if it comes up in BASIC.

[+slx]

The 400 should 'boot' to Memo Pad with no cart insrrted. Do you have a cart to try?

[Kyle22]

The 400 should 'boot' to Memo Pad with no cart insrrted. Do you have a cart to try?

It's not a 400, it's an 800XL...

 

Check the voltage output of the power supply. A stuck OPTION key shouldn't make the screen black. Self-Test should come up immediately when booting with OPTION down and no drives connected. RESET or power cycle will get out of Self-Test mode.

 

Hope this helps.

[re-atari]

Just purchased Atari 800XL and power supply separately, both appear in good shape, and initially it booted to READY prompt and all keys worked. When I held Option and rebooted it, the screen came up black (aka not snow, it had a signal) and stayed black through successive power cycles and resets. How do I get out of self-test mode? I don't find any mention of needing to exit self-test mode but it is acting like the option-boot persists across power cycles. Does self-test take a long time to display anything? Maybe I am not waiting long enough. Very confusing that it worked until I booted this way and now appears permanently "stuck". As an aside, I have a 400 that shows the same symptom - power-up to a black screen - so that makes me wonder if it is in some weird boot mode but I find no mention of it if so.

Hopefully you don't have an infamous Ingot power supply, there was a thread about this one recently. These are known to output a way too high voltage when failing, destroying several IC's inside the XL. I second the advice to check the voltage the power supply puts out. If that's within spec, try reseating the IC's inside the XL.

 

re-atari

[Bryan]

Powering the system off for 15 seconds or so will completely reset the computer. If they're not coming up properly, then something else is wrong.

[+slx]

It's not a 400, it's an 800XL...

.

OP said that he had a 400 with the same symptoms.

[russg]

Just purchased Atari 800XL and power supply separately, both appear in good shape, and initially it booted to READY prompt and all keys worked. When I held Option and rebooted it, the screen came up black (aka not snow, it had a signal) and stayed black through successive power cycles and resets. How do I get out of self-test mode? I don't find any mention of needing to exit self-test mode but it is acting like the option-boot persists across power cycles. Does self-test take a long time to display anything? Maybe I am not waiting long enough. Very confusing that it worked until I booted this way and now appears permanently "stuck". As an aside, I have a 400 that shows the same symptom - power-up to a black screen - so that makes me wonder if it is in some weird boot mode but I find no mention of it if so.

 

 

since the 800xl and 400 display the same thing, it may be your monitor or cable. Maybe try a RF connection to a TV. Be sure to set the channel.

[Rybags]

Regardless of if a drive is connected you should get the blue text screen periodically whether Option is held or not (assuming no cartridge present).

 

With no drives, the system does the SIO polling which takes a few seconds before going to Self-Test.

 

I would think something is wrong with the computer or PS. As was said before, wait several seconds between off then on again.

 

If you've got some cartridges, try them out. But it sounds like the computer itself does have a problem. Opening up then reseating all the big ICs would be the next thing to do.

[russg]

since the 800xl and 400 display the same thing, it may be your monitor or cable. Maybe try a RF connection to a TV. Be sure to set the channel.

I think that may be plugging the audio plug into the video jack. No, that isn't it, just tried it. Could be a RF with the wrong channel.

Nope, that not it either. And plugging the lumina and chroma plug in the video in isn't it either. So, good luck. Not the power supplies.

They are different for 400 and 800XL, I'm pretty sure 400 is a 9V AC.

Maybe chip swapping. Try GTIA and Antic. It is just strange that the two different computers give the same results. It certainly

isn't anything to do with the OPTION key, I think you can't bust a A8 with the OPTION key. Maybe check the jacks and switches

on your composite monitor. Or maybe you have the wrong channel selected if a TV, has to be channel 2 or 3..... Nope that not it either,

get 'snow' screen with wrong channel, not a grey screen.

I suppose a stuck OPTION key could cause something, but why both computers? The OPTION in a 400 is different than

the 800XL, the 400 should boot to memo pad with or without the OPTION key. The XL should boot to self test with OPTION key,

BASIC READY without.

Edited February 20, 2014 by russg

[Arcanis-Will]

The 400 should 'boot' to Memo Pad with no cart insrrted. Do you have a cart to try?

 

Yes, for the 400, I tried with and without a BASIC cart. Either way, same result - blank screen. I also removed both of the CPU and RAM card, with the same result. So for the 400 I assume something more serious is wrong internally if the output is the same with and without the CPU. Hmm, not sure what next step is for that one.

 

But I digress... back to the topic of the 800XL....

 

The 800XL displayed video just fine at first - therefore I will rule out the TV tuning and the connecting cable as a source of error.

 

Today, following the suggestion that it could be a stuck key, I unplugged the connector for the keyboard and upon the next power-up, I got the READY message - so at that point I figured great - I just have another keyboard to clean. However, upon subsequent power cycles, it is back to the blank screen again. So it was not the keyboard after all.

 

It has been sitting off for several hours while I work on my 800 keyboard, and now I am going to turn it back on again. My theory is that it will work for a short time and then die on me again. If so I am going to crack open the shield and reseat some ICs.

 

Oh and I don't think I have the ingot supply - based on a previous post from russg it appears I have The White Brick supply. It is the brown & cream one that matches the 800XL's case.

Edited February 20, 2014 by Arcanis-Will

[Arcanis-Will]

Reseated all ICs. Noticed that I had 3 or 4 different brands of RAM chips - 2 from West Germany, 1 from Japan, 1 from somewhere else, 4 from Malaysia. Must have been replaced at various times. Also found one of the larger ICs that had a bent pin crushed under it. Straightened it and reinserted it fine but that did not cure the problem.

 

I did run self-test. All keys work, sound works, all memory tests fine. However, I'm not sure I believe the memory results - the behavior fits what I expect from a memory problem.

 

On power-cycling, if you wait at least 5 seconds, first you get the blank screen then it boots to a blue screen then the READY prompt which overall takes 3-4 seconds. Seems ok.

However if you don't wait long enough between cycles, you get only the blank screen and it never boots.

On pushing reset button, 90% of the time you get only the blank screen. Very rarely you get back to READY prompt which takes less than 1 second if so.

While in BASIC, entering PRINT FRE(0) will usually cause the machine to invert colors (light blue background) and lock up. Sometimes I had success and saw the expected value printed.

Likewise doing something simple like 10 PRINT "HI" / 20 GOTO 10 will cause the inverted colors and lockup.

 

Booting with a BASIC cartridge installed seemed to improve things a little bit, where at least 50% of the time the simple programs would run. But then they also failed so it was not terribly conclusive.

[thorfdbg]

On power-cycling, if you wait at least 5 seconds, first you get the blank screen then it boots to a blue screen then the READY prompt which overall takes 3-4 seconds. Seems ok.

However if you don't wait long enough between cycles, you get only the blank screen and it never boots.

On pushing reset button, 90% of the time you get only the blank screen. Very rarely you get back to READY prompt which takes less than 1 second if so.

While in BASIC, entering PRINT FRE(0) will usually cause the machine to invert colors (light blue background) and lock up. Sometimes I had success and saw the expected value printed.

Likewise doing something simple like 10 PRINT "HI" / 20 GOTO 10 will cause the inverted colors and lockup.

 

Booting with a BASIC cartridge installed seemed to improve things a little bit, where at least 50% of the time the simple programs would run. But then they also failed so it was not terribly conclusive.

First, that you need to wait longer than 5 secs after poweroff to reboot the machine is normal. The machine runs into a cold reboot if three "magic values" in RAM are no longer correct, but even though it is dynamic RAM, it keeps its contents pretty long, even without refresh, depending on the type of memory.

 

That you get a lockup in Basic is an indicator of faulty RAM, though. I've seen similar. Do not trust the RAM test of the self test, it is pretty useless. First of all, it does not test all RAM, even though it inidicates it does (it's just lying). It does not test the zero page, nor page one, two or three. It neither tests the RAM under the self test. And it is not very reliable because it cannot test for common RAM defects like shorted address lines because it simply writes a sequential pattern into the RAM. If that pattern destroys something else in the RAM it does not test for. Thus, pointless test.

 

You could try to re-seat all RAM chips or, if that does not help, swap them one by another from a known good machine and check whether this helps. If not, you could also swap the CPU and ANTIC, both of which communicate with the RAM and could create similar defects.

[russg]

since the 800xl and 400 display the same thing, it may be your monitor or cable. Maybe try a RF connection to a TV. Be sure to set the channel.

Well, you're using a TV, the 400 doesn't have a composite jack. Maybe try a different TV. You can get a picture with an antenna?

[Arcanis-Will]

Well, you're using a TV, the 400 doesn't have a composite jack. Maybe try a different TV. You can get a picture with an antenna?

 

Hi russg, yes, the TV input is working ok with cable box, Atari 800, and this 800XL (sometimes) -- I tore down the 800XL, reseated all ICs, and spent lots more time power-cycling with longer delays -- I described the behavior in more detail up in post #12.

 

The problem I have now is that if it is a bad IC, I don't have a donor machine to do chip swaps. I found a good price for 4x RAM chips on ebay so I ordered them. Now I can swap out 1/2 at a time to rule out memory issues. But I don't have a second 800XL for the bigger chips.

[AtariLeaf]

First, that you need to wait longer than 5 secs after poweroff to reboot the machine is normal. The machine runs into a cold reboot if three "magic values" in RAM are no longer correct, but even though it is dynamic RAM, it keeps its contents pretty long, even without refresh, depending on the type of memory.

 

 

Tandy color computers are the same. Manuals even state that you should wait 10-15 seconds after power off to turn it back on or else you could get a warm boot instead of a cold boot lest you use a special poke command, which I can't remember at the moment. These machines should never be turned off and on after only 1 second. I would imagine that would cause problems down the road if done often enough.

[The Usotsuki]

I got in the habit of waiting 5 seconds before rebooting anything from using the Apple //e.

[atari8warez]

First, that you need to wait longer than 5 secs after poweroff to reboot the machine is normal. The machine runs into a cold reboot if three "magic values" in RAM are no longer correct, but even though it is dynamic RAM, it keeps its contents pretty long, even without refresh, depending on the type of memory.

 

This is a bit off topic but, I started to have that "memory still alive" problem with my 600XL after the 64K upgrade, I am wondering if there is a way to build a simple circuit which will erase/drain/corrupt etc.. the RAM contents so that I don't have to wait for several seconds before re-powering the computer.

[Arcanis-Will]

Now that it has been explained, the warm-boot behavior makes sense. I did notice that Reset takes less time than a cold boot.

 

The underlying problem affecting my 800XL prevents a successful warm boot. Instead, it hangs. For example, pushing Reset does not return to the Ready prompt as it should. Instead I will get a blank screen.

[russg]

Well, you're using a TV, the 400 doesn't have a composite jack. Maybe try a different TV. You can get a picture with an antenna?

I've had a senior moment, or rather an afternoon. You're not going to get a picture with an antenna. Unless you have a digital to analog adapter.

You already answered my question, your cable box works with it, puts an RF into it.

I should read before I post.

Well, you have to have an Atari RCA to TV coaxial adapter, which you have for the 800 that works, sometimes works with 800XL.

It isn't a problem with the video, it is something internal.

Cold starts should be at least ten seconds delay. Especially with higher RAM. If possible you should RUN E477 to do a re-boot, from

MyD/2.x 'M' or Sparta command line.

Are you sure your 800XL doesn't have a memory upgrade? Test with FRE(0) or memory tester like 'SHORTEST.COM'..

No, FRE(0) doesn't work with my Newell 256K. You gotta use a memory tester.

 

Well, SHORTEST.COM doesn't measure my 256K Newell correctly, says I have 131k. Here's RAMTEST.COM that does.

testsngl.atr

SHORTEST.zip

RAMTEST.zip

Edited February 20, 2014 by russg

[Arcanis-Will]

Especially with higher RAM. If possible you should RUN E477 to do a re-boot, from

MyD/2.x 'M' or Sparta command line.

Are you sure your 800XL doesn't have a memory upgrade? Test with FRE(0) or memory tester like 'SHORTEST.COM'..

No, FRE(0) doesn't work with my Newell 256K. You gotta use a memory tester.

 

Well, SHORTEST.COM doesn't measure my 256K Newell correctly, says I have 131k. Here's RAMTEST.COM that does.

 

Nope, it has 8x of the HM4864-P2 chips which is 64k x 1bit so just a stock XL. The fact that I have 2 from W. Germany, 2 from Japan, and 2 from Malaysia indicates there has already been some swapping. Since I can't run anything on it including cartridges, I'm not sure I will have any success running your attachments. My new RAM chips will be here Monday or Tuesday and that should be informative.

[russg]

 

Nope, it has 8x of the HM4864-P2 chips which is 64k x 1bit so just a stock XL. The fact that I have 2 from W. Germany, 2 from Japan, and 2 from Malaysia indicates there has already been some swapping. Since I can't run anything on it including cartridges, I'm not sure I will have any success running your attachments. My new RAM chips will be here Monday or Tuesday and that should be informative.

I'm sorry. I didn't understand your problem. I thought you indicated the 800XL would start sometimes, and that you were having problems re-starting it with or without the OPTION pressed.

It doesn't start with or without a cart installed, you just get a dark grey screen, and maybe a brief white flash at powerup. It never goes to self test? I'm leaning toward sending you a set

of 40 pin LSI. Antic, GTIA, 6502C, POKEY, PIA, for you to try and return. Are the LSI socketed, which ones are socketed?

Edited February 22, 2014 by russg

[Arcanis-Will]

I'm sorry. I didn't understand your problem. I thought you indicated the 800XL would start sometimes, and that you were having problems re-starting it with or without the OPTION pressed.

It doesn't start with or without a cart installed, you just get a dark grey screen, and maybe a brief white flash at powerup. It never goes to self test? I'm leaning toward sending you a set

of 40 pin LSI. Antic, GTIA, 6502C, POKEY, PIA, for you to try and return. Are the LSI socketed, which ones are socketed?

 

Hi Russ, let me try to describe the problem behavior:

 

There are 2 failure modes: 1) A blank screen. 2) An inverted colors screen where everything is powder blue / light blue background.

 

It does give me a READY prompt sometimes, but even then, it is unstable. If I get the READY prompt, it will not run well -- even a simple 2-line program will crash (failure mode 2).

 

Entering PRINT FRE(0) will crash (failure mode 2). Pushing RESET will lock up with blank screen (failure mode 1). Holding OPTION during boot will mostly lock up (failure mode 1). Turning power off and on, without waiting 10 seconds, will also give a blank screen.

 

For self-test, yes, I can enter self test sometimes also, but that does not show any error.

 

For sockets -- Every chip is socketed and I have reseated them all but that part did not help.

 

I appreciate your offer of loaning some chips to me. I might take you up on it. I'm hoping that the new memory chips that I ordered will help. I should receive them in the next few days.

 

Thanks!

Edited February 24, 2014 by Arcanis-Will

[thorfdbg]

There are 2 failure modes: 1) A blank screen. 2) An inverted colors screen where everything is powder blue / light blue background.

 

It does give me a READY prompt sometimes, but even then, it is unstable. If I get the READY prompt, it will not run well -- even a simple 2-line program will crash (failure mode 2).

Well, it *need not* to be the RAM chips. I had a similar case with one of my 800XL which caught a static discharge on the parallel port. RAM was actually not the issue, but the MMU chip burned. I believe it also contains a delay line to generate the RAS/CAS addressing of the RAM chips, and part of this logic did not work. Also, the BASIC ROM burned (interestingly, I thought that this part would be pretty robust). I had *some* red blocks in the "self test" that indicated a faulty RAM, but in the end, swapping chips from my spare part collection brought it back to life. As said, the RAM was actually fine. I still use the machine today, occasionally.

[Arcanis-Will]

The memory came today but as was foreshadowed, it did not fix the problem. Still same problem as before. Oh well.

 

I did a "push check" on each IC - making sure they were all solid.

 

Interesting result - pushing on U17 made the system work better for a short time. Immediately after pushing on U17 I had 2 successful boots before it failed again. U17 is labeled 1100008-01 which is the GTIA.

 

Separate Note - the chip that originally had a bent pin was U22 - C012294-19 which is Pokey. I had found either pin 20 or 21 was bent under the IC and not making contact. I have since straightened it and made sure it is seated OK. That would be /K2 or /K3 as per schematic. That would cause issues with Keyboard scan. This could not have passed initial burn in that way. Must have been replaced by an owner, and done poorly.

 

In summary, I should start next with trying to swap these 2:

U17 - since physical motion affects its behavior.

U22 - since the prior owner was running it with one pin open-circuit for who knows how long.

 

Is there any commonality with the 400? I have a dead 400 I could rob. Looks like I could steal both of them (depending on whether I have a GTIA or CTIA) - A101 / Pokey is on the motherboard, and A301 is the CTIA/GTIA on the CPU board.