exxos's Atari, Amiga & retro forum

Okay, I'm still checking the problem. when I remove the ram card, the values obtained at cpu init are the same: ssp =$0E2040 and PC=$217900000. So I've removed the rom chip and the processor does not gotten anything ... The chip select is correctly made by Combel... So my problem is "maybe" the rom chip ... I need a new :)

Could the problem be with one of the GAL's?

I've maybe some problems with GAL but, now, values received from the rom are never good even in transitory states. So I think my first problem is the rom chip, but it might not be the last...

My order is placed on Exxos Store :) Now, i have to wait...

Hi folks,
I changed the Rom: still nothing.
I reprogram all the GALs: still nothing
I check the data and address continuity between CPU and ROM : All ok.
Have you an idea ? I'm lost :(

Maybe this will help ? https://www.exxosforum.co.uk/forum/viewt ... &t=80&p206

Just a couple of things I'm wondering... (sorry if you've already answered these)

* have you owned this Falcon for a while & it used to work? In what circumstance did it stop working
* Is there any board modifications at all? ie; clock patches, nvram, boosters

silversword31 wrote: 10 Jul 2020 07:09 Hi,
i've another capture on cold start (tigger on vcc). I don't understand how i can have cpu activity without clock... Is the expansion bus directly connected on 68030 ?

No, no it's not. The expansion bus is actually a very bad place to test the CPU, unfortunately. It's translated and not fully.

silversword31 wrote: 18 Jul 2020 19:01 I've done another capture. I don't understand why the first action of cpu is to write $FFFF. Another curious thing is that /DTACK is totally inactive...
capture1.png

Which explains this. You'll only see DTACKs during DMA.

Looking at your snapshot of the start-up sequence:
You can see the CPU looks good -- it asks for the correct two longwords (four 16 bit accesses). The reply to the first two accesses should be 00E0, 0030. This will result in E00030 being the next access. It's not, it's 780000. That's wrong.

So it looks like you're not getting the results from ROM.

That could be caused by COMBEL not decoding the ROM address, but it is issuing a DSACK1 and in another post you see a bus error, so I'd be inclined to think COMBEL is decoding ok.

It could be COMBEL isn't asserting /XROM2, but you've tested that.

It could be the ROM? You've replaced that, but what about its socket? Is the chip getting power, perhaps?

It could be the data lines? Continuity's not enough. We can check it better: remove the ROM chip again. You should be getting FFFF on the data line from the pull-ups. That startup sequence above should therefore try to access FFFFFFFF on the fifth cycle.

This would seem the next step to me. If something's pulling the lines low, or talking on the data bus, you might be seeing this kind of thing.

Good luck,

BW.

Thanks a lot for your help :) I’ll try to remove rom and watch the result. I’ll give you the result. Thanks :)

When i remove the rom, i get $FFFF as Data. It's good. But, now, the cpu doesn't continue after reading $000004.... It stops after reading :(