exxos's Atari, Amiga & retro forum

stephen_usher wrote: 11 Sep 2024 09:16 Check U601, that 74LS245 is bidirectional and if faulty could pull those lines down potentially.

Thanks when you say check do you mean take it out and test it or can I probe a pin to check? I guess its hard to know what logic chip is pulling the line down, are the bidirectional logic chips more likely to pull the line down that say the 74F244. although looking at the schematic the dram chips only get A0-A8 which goes via the 74f244's

Take it out and test. You probably won't be able to tell anything in circuit.

Also check the resistance between those address lines and ground (preferably with the CPU and U601 removed) to check for shorts.

Of course, there is a very remote possibility that the system just wasn't trying to use those addresses, but it's unlikely.

stephen_usher wrote: 11 Sep 2024 12:25 Take it out and test. You probably won't be able to tell anything in circuit.

Also check the resistance between those address lines and ground (preferably with the CPU and U601 removed) to check for shorts.

Of course, there is a very remote possibility that the system just wasn't trying to use those addresses, but it's unlikely.

I have now removed the top 8 drams, all 8 were faulty, added sockets and now diagrom gets past the initial data tests, but there is still random address errors reported, sometimes none, sometimes 1 or 2, sometimes odd characters displayed over serial. With the KS ROM in it now gets to the light grey screen.

So slowly making progress.

:dualthumbup:

stephen_usher wrote: 12 Sep 2024 16:58 :dualthumbup:

Thanks, slowly slowly

After some piggybacking if I piggy back the chip marked with a cross the KS screen appears (corrupt), piggy backing the other chips makes no difference at this point until I put a chip on the chip marked with a white dot, when I do this I get a yellow screen or sometimes a grey screen. So I know the chip marked with a white cross is faulty and I suspect the white dot chip is as well. So will remove those tomorrow.

You're definitely getting close.

If you piggyback a chip and it doesn't change anything it does not mean the chip is good..
It all depends on how the chip is defective on the inside.

With some broken chips the piggyback works, others do not.
(make/model or function does not matter for this)

Indeed.

Given the number of bad chips found I would not trust any of them. Desolder them all and test every one.

Maximilian wrote: 13 Sep 2024 01:44 If you piggyback a chip and it doesn't change anything it does not mean the chip is good..
It all depends on how the chip is defective on the inside.

With some broken chips the piggyback works, others do not.
(make/model or function does not matter for this)

Yes thanks I know this, I'm just hoping these next two chips are the final step otherwise I think it comes down to randomly removing and testing DRAMs until I stumble upon any faulty ones.(After probing DRA lines to check for anomalies)

stephen_usher wrote: 13 Sep 2024 07:04Given the number of bad chips found I would not trust any of them. Desolder them all and test every one.

Nooooooo :o :D

I'm making progress, after replacing the chip marked with "x" I now get the KS screen and I can test RAM using ATK, all passes with over 20 rounds of memory testing.

Only 512k of chip RAM is detected though, no slow RAM. I have jumper J500 closed and J101 set to 1-2 with a 8371 Agnus.

When I replaced the chip with a white dot I get a yellow screen. The chip with a white dot when put in a working rev5 A500 produces a green screen so I know that chip is faulty. I think there must be another chip that is faulty somewhere that is stopping the 512k slow RAM/Ranger RAM being detected.

Does anyone know which one it might be?