Troublshooting an Atari 400 No Boot scenario

[Trellot]

Greetings all,

 

Purchased an Atari 400 and all I get is a black/brown screen on the TV; with or without a cartridge. Channel select on 3 as is the TV.  Heads up: I am using the following tech manual "Sams_Computer_facts_Atari_400" to troubleshoot from: http://www.atarimania.com/documents-atari-400-800-xl-xe-technical-documents_3_8.html

 

Power supply voltages: (Note: this is the second power supply; first one was bad and I'm currenty recapping it.)

CR206 BR Diode -> 11.58v

CR208 BR Diode -> 11.79v
VR A202 (5v) pin3 -> 5.06v
VR A201 (12v) pin1 ->23.86v 
VR A201 (12v) pin3 -> 12.20v
CR201 Zener Diode ->  -5.12v

 

Verify Oscillating Crystal on mainboard: (should be 3.579575mhz)

X101 -> (using a multi-meter on Hz setting: I'm getting varied results across the board nothing particularly solid and the reading doesn't stay put, it zeros out after holding the probes for a while.  Any tech tips on getting a good read on this part? All I have at the moment is this multi-meter.  

 

Thanks,

 

~ Trellot

 

 

Edited October 5, 2019 by Trellot

[Rybags]

It's likely with brown screen the OS isn't performing the early initialization (zeroing registers which would make the screen black is part of that).

If you have a diagnostic mode cart such as Star Raiders, Asteroids and some other early Atari releases, give them a try.

 

Since you have test equipment maybe try the Sync pin on the CPU.

It's an old machine and oxidization of chip pins can be an issue so it's worth lifting/reseating the ICs on the CPU board and do the same for the OS though with the 400 this does mean a near full pull apart process.

Note that the CPU won't necessarily be interchangable with the Sally one as used on XL/later - some late 400/800 used it but check the part # before ordering a replacement.

[Trellot]

1 hour ago, Rybags said:

Since you have test equipment maybe try the Sync pin on the CPU.

Thanks for the response!  I'll try the above and post results.  I have reseated all the ICs on this unit previously, hoping that'd get be booting but alas...

 

~ Trellot

[Trellot]

So I'm getting 5volts on pin8 of the 6502 chip and .016v on the SYNC pin7.  Note: I was only testing with the mainboard and power supply powered up, no other connections.

 

~ Trellot   

[Rybags]

SYNC should go high (?) each time an instruction is read - the usefulness in this case might be limited since the CPU might be in a hang in which case I don't think it changes state.

[eegad]

If you do manage to figure it out,  I'd very much like to hear what you did.  My first computer was a 400. Didn't touch it for over 30 years...  When I got back into Atari stuff a dozen years ago or so,  I only hooked up my 800xl,  which replaced my 400 in 1984. Finally I decided to dig out and play a few cart games on the 400 one day a couple years ago.  I got the same brown screen as you.  Tried re-seating socketed chips on the board and such.  Only thing that ultimately accomplished was getting different solid colors on screen... Very dark green,  maybe maroon,  instead of brown.  One thing I will ask - does your 400 have a stock (Atari installed) RAM module?  The one thing I did to my 400 back in the day was upgrade to a third party 32k module.  When I was tinkering with my 400,  I was wondering if the RAM was shoddy and was the culprit,  but I don't have the original 16k module any more to test-swap. 

 

[Trellot]

On 10/6/2019 at 6:36 AM, eegad said:

If you do manage to figure it out,  I'd very much like to hear what you did.  My first computer was a 400. Didn't touch it for over 30 years...  When I got back into Atari stuff a dozen years ago or so,  I only hooked up my 800xl,  which replaced my 400 in 1984. Finally I decided to dig out and play a few cart games on the 400 one day a couple years ago.  I got the same brown screen as you.  Tried re-seating socketed chips on the board and such.  Only thing that ultimately accomplished was getting different solid colors on screen... Very dark green,  maybe maroon,  instead of brown.  One thing I will ask - does your 400 have a stock (Atari installed) RAM module?  The one thing I did to my 400 back in the day was upgrade to a third party 32k module.  When I was tinkering with my 400,  I was wondering if the RAM was shoddy and was the culprit,  but I don't have the original 16k module any more to test-swap. 

 

Yes, I will post my findings as I go.  Regarding my Atari unit, it does have the stock RAM module inside.  I think my next step is to order new populated CPU/RAM boards and play the substitution game.  Even if they don't solve my problem I'll at least have a good set for future repair projects.

[+Nezgar]

8 hours ago, Trellot said:

I think my next step is to order new populated CPU/RAM boards and play the substitution game.  Even if they don't solve my problem I'll at least have a good set for future repair projects.

In not as familiar with the behaviours of the 400/800, but the XL's and XE's can exhibit the solid color screens when there is a bad OS ROM. I think replacing the ROM's on the personality board might be a good start... (Or a new personality board).  If you can find them. Replacing the ROM chips with EPROM's that are common today would be most easily achieved with 23xx adapter PCB's from Retro Innovations to convert 27xx type eproms to 23xx pinout..

 

If you have another working Atari, you could swap the 24 pin OS ROM chips into a socketed brown-shell Atari cartidge (also 24 pins), and use DOS to save memory to disk, and analyze the file to see if the ROM contents show up, for each chip. (one 2K fp math chip, 8KB OS split into two 4K chips)

[_The Doctor__]

what were the values on the bad power supply? if it fed the wrong voltage on the 12 volt side... check the ram and support...

if it fed the wrong voltage on the 5 side check the os and other lsi's first...

 

The whole of clean reseat, make sure no bent pins/bad sockets joints thing still holds true... and change the order of the ram chips... sometimes a bad chip in a different slot is enough to get a machine partially working...

 

RAM almost always seems involved lately on the out of storage fix train though...

[Trellot]

 

17 hours ago, Nezgar said:

If you have another working Atari, you could swap the 24 pin OS ROM chips into a socketed brown-shell Atari cartidge (also 24 pins), and use DOS to save memory to disk, and analyze the file to see if the ROM contents show up, for each chip. (one 2K fp math chip, 8KB OS split into two 4K chips)

I'm not familiar with the process you're describing above although I understand the value of this test would be to show that the ROMs still dump some kind of data, is that right?  So if I were to conduct this test then any data dumped would show the chip to be good? or would I need to compare my dump to some other known good dump out there?  Thanks for your reply, btw!

Edited October 8, 2019 by Trellot

[Trellot]

2 hours ago, _The Doctor__ said:

what were the values on the bad power supply? if it fed the wrong voltage on the 12 volt side... check the ram and support...

if it fed the wrong voltage on the 5 side check the os and other lsi's first...

 

The whole of clean reseat, make sure no bent pins/bad sockets joints thing still holds true... and change the order of the ram chips... sometimes a bad chip in a different slot is enough to get a machine partially working...

 

RAM almost always seems involved lately on the out of storage fix train though...

The bad power side board never even powered the LED, it was just dead, I couldn't get any voltage readings anywhere on the 400... it ended up causing my power supply to blow a fuse.  I'll try the RAM swapping too.  Thanks!

Edited October 8, 2019 by Trellot

[+Nezgar]

14 hours ago, Trellot said:

 

I'm not familiar with the process you're describing above although I understand the value of this test would be to show that the ROMs still dump some kind of data, is that right?  So if I were to conduct this test then any data dumped would show the chip to be good? or would I need to compare my dump to some other known good dump out there?  Thanks for your reply, btw! 

Yes, the idea is exactly that - A) verify data can be read from each of the 3 chips, B) verify the data that was read was GOOD. If you save the rom dumps using DOS to disk or an ATR, I can check them / piece them together again if you upload them here to verify.

 

I successfully used this procedure to dump the 4K mask ROM from the Atari 850 in this topic, the process will be similar:

 

In summary:

1. Put one of the ROM chips into one of the slots in an old socketed Atari brown shell cartridge
2. Boot DOS 2.5, Save Binary file: ROM1.ROM,A000,BFFF
3. repeat 1-2 for each rom chip, making computer is off before inserting/removing cartridge.
4. Upload resulting ATR file here for anaylsis
5. OR if you feel technically incined, use Altirra or AspeQT to extract the files from the ATR images to the PC, and examine them in a hex editor.

 

[Trellot]

15 hours ago, Nezgar said:

1. Boot DOS 2.5, Save Binary file: ROM1.ROM,A000,BFFF

Thank you for your comment.  I'd very much like to try this process.  I do have one cart that could be used, such as the one you pictured above. 

 

Questions: 

1. Can DOS 2.0 virtualization be used for this process? 

2. How do I connect the ROM cart to the computer so that the OS sees it?  I've dumped spis and such before but through a USB connection.

3. Lastly, as I'm thinking more about this, I feel like I might need a working Atari 400 for this?  If so, I guess this test won't work for me yet, lol.

 

Thanks,

 

~ Trellot

[+Nezgar]

Sorry yes, I was assuming the cart would be plugged into a second real computer ... if the problem 400 was working, you could easily dump the OS ROM without even opening it.

 

Only the disk drive would need to be virtualized to facilitate transfering the saved dump to a modern PC via a virtial disk image. (SIO2PC-USB, SIO2SD, etc).

 

I guess main focus then would be to acquire a new personality and/or CPU board to try swapping out, or another working 400 to swap parts with for troubleshooting.

[Trellot]

21 hours ago, Nezgar said:

Sorry yes, I was assuming the cart would be plugged into a second real computer ... if the problem 400 was working, you could easily dump the OS ROM without even opening it.

 

Only the disk drive would need to be virtualized to facilitate transfering the saved dump to a modern PC via a virtial disk image. (SIO2PC-USB, SIO2SD, etc).

 

I guess main focus then would be to acquire a new personality and/or CPU board to try swapping out, or another working 400 to swap parts with for troubleshooting.

Thanks Nezgar, yeah I will try to acquire more working pieces for this troubleshooting. I like your neat trick here however!  I will try it nonetheless once I have a working unit

[Trellot]

Evening All,

 

Finally found some time to continue debugging what I started on this posting regarding a non-booting Atari 400 I purchased last year.  Previously I was seeing a dark brownish screen at power up and we were able to narrow down the issue to potentially the OS ROM chips.  Since my last posting, I was able to acquire another working Atari 400.  Below is a table of results on some chip swapping I did to try to segment which chip is bad, or a combination of chips, etc.  In the table each chip configuration, per test, is exactly the opposite of the other on each Atari system for clarity.

 

Note_1: 400_A = original failing system which prompted this post; 400_B is the new working system I recently acquired. 

Note_2: In the table, when I declare a Pass what I mean is that the system boots to Memo Pad...i didn't have the keyboards installed to test for full functionality, FYI.  

 

Observation_1: Notice Test #4 where, based on the chip configuration, I found two simultaneous passing configurations causing two working systems...this confuses me since, based on the table, and with all three original ROM chips installed into 400_A the system won't boot; one of the original ROM chips from the Atari 400_A system appears to be bad.

 

Observation_2: on the 400_A side, tests 5,6,7,8 seem to suggest that A105 is the culprit, but then see test #4 on the 400_B side and that chip passes when installed alone.    

 

Observation_3: Also notice Test 2,3,4 from 400_B ... it shows a Pass when swapping one 400_A chip at a time into 400_B, effectively declaring all three chips as good...ugg?   

 

Let me know if any of you have an insight into this data.  Thank you!  

 

 

Edited March 9, 2020 by Trellot

[Trellot]

Afternoon everyone,

 

I've made a final determination of which chip is at fault after gathering the data for the table above.  The A105 OS ROM chip for the 400_A system is bad.  For Test #4 in the above table, I was able to get a passing set across two systems by swapping the A105 chips into each, however, repeating the swap test over and over resulted in the failure following chip A105 from Atari 400_A more often than not.  With this data, I'll be placing an order in for a new one very soon.  

 

Happy day to all,

 

~ Trellot 

[eegad]

I had similar problems with my 400. I had replaced the 16k ram card with 32k intec card way back when.  A couple years ago, when I tried to use it for the first time in like 30 years,  all I got was a black screen.... Or brown.... Or green.  A few weeks ago I decided to mess with it again finally.  I took a 16k board out of my 800 and put into the 400. Bingo,  it booted up fine. I then ordered a 16k board from ebay for $20 and my 400 is in business again after all these years. 

 

So if you haven't  tried it yet - swap the ram card with a known working one.