Broken reset circuit / 555?

[dml]

\o/

nice work

[redpixel]

After replacing the Videl ... both the normal and turbo mode work. So that puts the ability to boot without Videl really into question.

Not sure if I'm misunderstanding what you mean by boot, but isn't VIDEL sitting in-between the COMBEL and the DRAM data lines? I.e. without a VIDEL you wouldn't be able to access DRAM at all? Then I don't expect it to be able to boot very far. Sort of impressive that it got as far as it did with the diag cart (first stage of memory testing).

And P.S. great job fixing the Falcon!

[mikro]

By booting I mean booting with the diag cart over serial port. The service guide hints on several places that the Falcon should be able to boot without it and it did in the CT2 mode, so who knows for sure.

[redpixel]

By booting I mean booting with the diag cart over serial port. The service guide hints on several places that the Falcon should be able to boot without it and it did in the CT2 mode, so who knows for sure.

I think what you saw is how far it's able to boot without a VIDEL. It will of course die on the first RAM test and in the diag cart source it seems to indicate that it will repeat RAM tests endlessly if they fail:

Code: Select all

*********************************
*       Test system RAM         *
*       If error, branch to FATAL routine with d0=data written,
*       d1=data read, a0=location

*       ========
*       Test high and low
tstdat: moveq   #0,d6
tstd6:  move.w  d6,d0           ;d6=data for repeat
        movea.l #8,a0
tstd0:  move.w  d0,(a0)
        nop
        move.w  (a0),d1
        cmp.w   d0,d1
        bne     fatal0

...

*       Failed during data line check
*       d0=data written, d1=data read
fatal0:
        movea.l #datlin,a5
        bra.s   fatal

...

****************************************************************************
*       Handle a fatal error
*       Print out RS232; assume no RAM.
*       Clear screen and turn red. Then repeat the test(s).
*       d0=data written, d1=data read, a0=address, a5=message ptr
****************************************************************************
* MRW
*
****************************************************************************

fatal:
[... print the results over RS232 ...]
        bra     tstd6           ;repeat the RAM tests endlessly

[mikro]

Nice investigation! However it still doesn't explain why it worked only in the CT2 mode (well, except the mysterious two chip legs doing something nasty on the motherboard of course).

[redpixel]

Nice investigation! However it still doesn't explain why it worked only in the CT2 mode (well, except the mysterious two chip legs doing something nasty on the motherboard of course).

No, that's hard to say for sure I guess. Doesn't sound super tempting to desolder a VIDEL to corroborate your findings :) Maybe the CT2 helped in some way, or I guess it could just be behaving a bit unpredictably when you don't even have a VIDEL physically on the board. Some control lines might be floating etc.