Atari 400 Failure

[Panther]

12 hours ago, ClausB said:

Piper Arrow. No horn.

Boy you pros have all the gadgets!

You people and your fancy retractable gear.  Piper Archer, keeping it simple.

[Panther]

Does anyone know what the capacitor is that's added to the bottom of the 400 mainboards between C134 and POKEY pin 17?  I gather it's some post-production fix.

Actually, it's probably a resistor.

Edited April 1, 2021 by Panther

[Panther]

Now I'm getting somewhere...C183 is toast.

[Panther]

I'm seeing 5V at C183 on the bad board, only about 1V on a good board, so now I have to find where that's coming from.

[ClausB]

C183 is in the color circuit, so I would expect it to differ between a black screen and a blue screen.

[Panther]

I don't have the color connected to anything right now, but either way, I had to remove the system from my workbench for now so I can get some work done, but I'll resume this weekend.

[Panther]

I received a Rigol DS1202Z-E today and spent a couple hours figuring out how to operate it.  Reset goes low 44 to 70 ms after 5V goes high, usually in the 60's.

[ClausB]

Goes low and stays low? I thought it would go high after a bit and stay high, because it's active low.

 

Now you can use the score to compare waveforms between the bad and good boards. Maybe look for the Sams Computerfax trouble shooting guide.

Edited April 3, 2021 by ClausB

[Panther]

I've been working on it, just keep running into dead ends.  At least I now have a much more usable oscilloscope compared to the DS203 I was using before. 

[Panther]

Sorry, I meant it goes high after up to 70ms after going low right at power on.

 

My brain is a jumble of waveforms after my crash course study of the new toy.

Edited April 3, 2021 by Panther

[Panther]

The chip select (CS0) at the POKEY is a mess - it's staying high and extremely noisy.  Traced that back to the 74LS138 and found a mess of signals at it, not at all what I find on the good board.  I will have to track it down from there tomorrow.

 

[+DrVenkman]

9 hours ago, Panther said:

Sorry, I meant it goes high after up to 70ms after going low right at power on.

 

My brain is a jumble of waveforms after my crash course study of the new toy.

That’s what it’s supposed to do: go low on power-up, then jump high and stay there. If you can figure out your trigger settings and timebase to catch it, see if you can grab a full wave form. It should be darn close to a perfect negative square wave. 

[Panther]

While I'm learning a lot about the signaling on a 400, including how the decoders are utilized, I'm afraid I've still not found the source of the problem.  I find symptoms of the problem, such as CS1 (chip select at the GTIA which also affects the POKEY) being screwy, but keep running into dead ends trying to track down the cause.  I've not given up on this board yet though!

[Panther]

Does anyone know what this is from?  It's attached to the bottom of my spare 400 mainboard.

[+Stephen]

58 minutes ago, Panther said:

Does anyone know what this is from?  It's attached to the bottom of my spare 400 mainboard.

Yeah - the 48kB mod in a 400 requires some wires to be soldered.  I guess this PCB was done to make it hard to get the connections wrong.  Also, I could have sworn either this (or maybe a similar device) was offered as a solder free solution, I guess meant to be held in place with double sided tape or some such.

 

Oh - to test if your machine has 48kB, from BASIC type ?FRE(0).  A 48kB machine will report 37902.  if it's a 16kB machine, it will be 13,### (I don't remember the exact number).

[Panther]

This one was definitely meant to be soldered, but all this does is attach the $8000-$BFFF select/deselect lines to the cartridge port.  This particular board came to me as-is, no cards.

 

[Panther]

Oh, that's right, those are all the signals required if it's just for a 48K card.  I was thinking of the difference with the 64K mod.

[Panther]

I ended up removing the 64K mod on this mainboard and it works now.  I'm not sure why it is causing a problem unless the 16K cards simply don't work with the 64K modified mainboard.  I wasn't trying to use the 64K card since it's likely it was damaged when it was installed backwards.  I had traced the problems back to signals coming off of the RAM board, but the card I was testing with worked fine in another (less modified) mainboard.  Claus?  Are the mainboards not supposed to work with 16K cards after doing the 64K mod with the 74LS02 added?

[ClausB]

Wait, what?

 

Did I miss your original mention of a 48/64K RAM board mod? I was assuming 16K stock hardware all along!

 

Which 64K mod are you asking about? Mine? Let me check...

Edited April 6, 2021 by ClausB

[ClausB]

In my 48K mod, with the LS02 wired to the Mainboard, and a 16K board in the RAM slot, I think the RAM would be mirrored 3 times in the 48K space. I'm not sure how the OS would handle that.

 

But your photo shows a different Mainboard mod. With that mod a 16K board would short pins M and N of the RAM slot, which would short pins 1 and 14 of the cart slot. If a BASIC cart were plugged in, it would short 5V to /S4, which might make the LS42 unhappy.

Edited April 6, 2021 by ClausB

[Panther]

The last photo I sent with that other mod was some 400 mainboard I found in my collection.  The one that I've been working on had your 64K mod, but I removed that and tried installing the factory 48K mod, but I don't know how well that works with your version of the 48K board. I believe it's the same as the one I had removed from the other board, just with wires instead of a flexible board.  That would certainly explain why I can't use cartridges.  Other issues are popping up too, so maybe I'll install your mod back in now that I've replaced all the chips in the 64K RAM card.

[Panther]

You would think that Atari would have mentioned in the 48K upgrade kit manual that installing a 16K RAM card would be harmful.  By the way, I have photos I took of the factory upgrade manual and RAM card, in case you're interested.  I'd installed it for someone else awhile back.

[ClausB]

No, my mod is not compatible with the Atari 48K board, which has the LS02 equivalent logic on board.

 

Sure, post the photos. Someone else has already, but there might be some interesting differences.

[Panther]

6 hours ago, ClausB said:

No, my mod is not compatible with the Atari 48K board, which has the LS02 equivalent logic on board.

 

Sure, post the photos. Someone else has already, but there might be some interesting differences.

Attached.  Sorry, these were just taken with my phone just for future reference.  A 2015 phone at that.

Meanwhile, I've reinstalled your 48K mod on the mainboard and switched back over to the modified RAM card.  So far the system is mostly working, but I did have trouble with a couple of the games I launched, so still troubleshooting.

 

Atari 400 48K Memory Upgrade (Factory).pdf

[Panther]

Mostly works...just some issues with some cartridges.  I can load up an AtariMax cartridge for example, but some games will run from it while some won't.  No problems with anything over SIO.