exxos's Atari, Amiga & retro forum

stephen_usher wrote: 14 Jun 2024 11:50 It shouldn't matter which ROM you have in the machine for the /OE signal as that's an input to the ROM to tell it to output data onto the bus.

The Amiga 2000's ROM socket is wired up to accept the type of ROM chip that DiagROM is written into isn't it? Something to check as at least some of the A500s have different pin outs and the 2000 is based upon the same design.

What revision board is in the 2000?

It is a 6.2.

stephen_usher wrote: 14 Jun 2024 11:50 It shouldn't matter which ROM you have in the machine for the /OE signal as that's an input to the ROM to tell it to output data onto the bus.

The Amiga 2000's ROM socket is wired up to accept the type of ROM chip that DiagROM is written into isn't it? Something to check as at least some of the A500s have different pin outs and the 2000 is based upon the same design.

The only reason why I mentioned "with Diagrom" is because the motherboard seems to progress further in the boot process with the Kickstart ROM than with Diagrom...well, it doesn't at least halt the system. That is why I try everything with both ROMs.

Have you toned out the buses as suggested by TF? Checking for dead shorts to ground, 5v, and/or continuity (obviously bear in mind that there is some buffering so trace accordingly).

Kepler wrote: 14 Jun 2024 11:16 Thanks.
ROM OE is high at power on with Diagrom.

Then the ROM isnt being selected and code isnt being run. Check pin 21 on Gary at boot now + measure the resistance between Gary 21 and /OE on the ROM.

If the connection to Gary(21) and ROM (13) is fine then check Gary's address pins A23-A19 for a short to ground (Pins 39 to Pin 35).

Also just try pulling /OE low and running diagrom (i.e. short it to ground).

EDIT: Finally check OVL on Gary (30) is briefly low at start up. It should be for around 500ns. If that doesnt happen then check OVL's connection to ODD CIA Pin 2.

EDIT2: What i suspect is happening if all the info i've been given is correct is that for some reason the ROM isnt being selected. There can only be 5 reasons for this, Dead ODD CIA, Dead Gary, short on the upper address bus, track break from ROM to Gary, Track break from Gary to ODD CIA.

Kepler wrote: 14 Jun 2024 12:33 The only reason why I mentioned "with Diagrom" is because the motherboard seems to progress further in the boot process with the Kickstart ROM than with Diagrom...well, it doesn't at least halt the system. That is why I try everything with both ROMs.

It probably just runs more instructions before it hits a double bus fault. Its not really an indication of anything.

terriblefire wrote: 14 Jun 2024 13:08

Kepler wrote: 14 Jun 2024 11:16 Thanks.
ROM OE is high at power on with Diagrom.

Then the ROM isnt being selected and code isnt being run. Check pin 21 on Gary at boot now + measure the resistance between Gary 21 and /OE on the ROM.

If the connection to Gary(21) and ROM (13) is fine then check Gary's address pins A23-A19 for a short to ground (Pins 39 to Pin 35).

Also just try pulling /OE low and running diagrom (i.e. short it to ground).

EDIT: Finally check OVL on Gary (30) is briefly low at start up. It should be for around 500ns. If that doesnt happen then check OVL's connection to ODD CIA Pin 2.

EDIT2: What i suspect is happening if all the info i've been given is correct is that for some reason the ROM isnt being selected. There can only be 5 reasons for this, Dead ODD CIA, Dead Gary, short on the upper address bus, track break from ROM to Gary, Track break from Gary to ODD CIA.

First of all I want to say that this is one of the best responses I have read on fault diagnosis with a level of detail that makes it useful to anybody else with a similar issue.

These are my findings:

Check pin 21 on Gary at boot now + measure the resistance between Gary 21 and /OE on the ROM.

Both Gary pin 21 and ROM /OE start high on boot.
I then checked my (working) test board and sure enough, that starts high too and then drops low, about 1.7Mhz according to my scope. So the difference between my non working and working board is /OE never goes low on my non working board.

If the connection to Gary(21) and ROM (13) is fine then check Gary's address pins A23-A19 for a short to ground (Pins 39 to Pin 35).

None of those pins are shorted to ground except pin 40 and 48 - which are supposed to be shorted to ground.

terriblefire wrote: 14 Jun 2024 13:08

Kepler wrote: 14 Jun 2024 11:16 Thanks.
ROM OE is high at power on with Diagrom.

Then the ROM isnt being selected and code isnt being run. Check pin 21 on Gary at boot now + measure the resistance between Gary 21 and /OE on the ROM.

If the connection to Gary(21) and ROM (13) is fine then check Gary's address pins A23-A19 for a short to ground (Pins 39 to Pin 35).

Also just try pulling /OE low and running diagrom (i.e. short it to ground).

EDIT: Finally check OVL on Gary (30) is briefly low at start up. It should be for around 500ns. If that doesnt happen then check OVL's connection to ODD CIA Pin 2.

EDIT2: What i suspect is happening if all the info i've been given is correct is that for some reason the ROM isnt being selected. There can only be 5 reasons for this, Dead ODD CIA, Dead Gary, short on the upper address bus, track break from ROM to Gary, Track break from Gary to ODD CIA.

First of all I want to say that this is one of the best responses I have read on fault diagnosis with a level of detail that makes it useful to anybody else with a similar issue.

These are my findings:

Check pin 21 on Gary at boot now + measure the resistance between Gary 21 and /OE on the ROM.

Both Gary pin 21 and ROM /OE start high on boot.
I then checked my (working) test board and sure enough, that starts high too and then drops low, about 1.7Mhz according to my scope. So the difference between my non working and working board is /OE never goes low on my non working board.

If the connection to Gary(21) and ROM (13) is fine then check Gary's address pins A23-A19 for a short to ground (Pins 39 to Pin 35).

None of those pins are shorted to ground except pin 40 and 48 - which are supposed to be shorted to ground.

try pulling /OE low and running diagrom (i.e. short it to ground).

I did. It made no difference.

EDIT: Finally check OVL on Gary (30) is briefly low at start up. It should be for around 500ns. If that doesnt happen then check OVL's connection to ODD CIA Pin 2.

On my non working board, OVL is high on start up and stays high.
On my working board, OVL is high on start up and then drops to low.
The connection from OVL to ODD CIA pin 2 is fine.

EDIT2: What i suspect is happening if all the info i've been given is correct is that for some reason the ROM isnt being selected. There can only be 5 reasons for this, Dead ODD CIA, Dead Gary, short on the upper address bus, track break from ROM to Gary, Track break from Gary to ODD CIA.

Dead ODD CIA - tested fine.
Dead Gary - tested fine.
Short on the upper address bus - tested each pin to ground, no short.
Track break from ROM to Gary - tested the two address lines and ROMEN
Track break from Gary to ODD CIA - tested OVL.

Finally, in case my Diagrom is faulty...I tested it in my working board...it works perfectly! 😊

Kepler wrote: 14 Jun 2024 14:51
On my non working board, OVL is high on start up and stays high.
On my working board, OVL is high on start up and then drops to low.
The connection from OVL to ODD CIA pin 2 is fine.

So this statement is at odds with a working OVL. OVL should always start low. Is something shorting it or pulling it high?

What this signal does is tell Gary that we've just had a RESET signal so please overlay ROM at address 0. Without this there is no initial code to get us running ROM and basically the CPU will run random crap in ram.

It would be worth pulling the ODD CIA... checking the voltage on PIN2 .. if there are no shorts to VCC try shorting it to GND and measure the voltage at Gary. The problem is clear now we just dont know why its happening.

terriblefire wrote: 14 Jun 2024 15:33

Kepler wrote: 14 Jun 2024 14:51
On my non working board, OVL is high on start up and stays high.
On my working board, OVL is high on start up and then drops to low.
The connection from OVL to ODD CIA pin 2 is fine.

So this statement is at odds with a working OVL. OVL should always start low. Is something shorting it or pulling it high?

What this signal does is tell Gary that we've just had a RESET signal so please overlay ROM at address 0. Without this there is no initial code to get us running ROM and basically the CPU will run random crap in ram.

It would be worth pulling the ODD CIA... checking the voltage on PIN2 .. if there are no shorts to VCC try shorting it to GND and measure the voltage at Gary. The problem is clear now we just dont know why its happening.

I've actually given you wrong information here.
So if you've measured it high then ROM is in the right place. But it does not explain why you are not seeing /OE go low on at the ROM.

Expanding a little here are the conditions that need to be met for Gary to assert ROMOE.