14 mhz

[RUDOLPh ERIC]

Here goes again:
I installed a 16mhz 68000p12F (F!) in the rev 5 amiga 500. For the 14mhz signal,
I tried CDAC* XOR 7mhz and (NOT) CDAC* XOR 7mhz with a 74Fast86.
The leads fron agnus to the 86 are about two inches long, solid wire.
The leads from the 86 to the 68k are about 2 inches long also, with ferrite bead.
Now when I turn on the computer, it ALWAYS comes up dark grey first, then
bright green replaces it. I get 10 short flashes, and one long flash, and it
tries to restart.

I am under the impression that DTACK will hold off weird bus cycles. What is the
problem here, will somebody explain why this is going bad and how I might get
it to work with some extra logic? In my opinion, this is a very very poor Hack.

Also, please include the number of your server. My server has a HARD time
identifying names. I.E. umbra.cs.orst.edu=128.???.???.??. Thanks.

[Dave Haynie]

In article <21...@orstcs.CS.ORST.EDU> rud...@upas.CS.ORST.EDU (RUDOLPh ERIC) writes:
>Here goes again:
>I installed a 16mhz 68000p12F (F!) in the rev 5 amiga 500. For the 14mhz signal,
>I tried CDAC* XOR 7mhz and (NOT) CDAC* XOR 7mhz with a 74Fast86.

If that's the only logic, you're doomed to failure. Even ignoring the problems
with the E/VPA*/VMA* logic for now, any accelerated 680x0 must still respond
properly to the standard 68000 bus signals at 7MHz. So, for example, that
means that you don't drive AS* on the bus until S2, in 7MHz time. Same for
the DS* lines on read, and on write you wait until S4, again in 7MHz timeframe.
You don't look at DTACK* until the S4->S5 transisiton, 7MHz timeframe, and you
don't count on any valid data until the S6->S7 transition.

That's the simple basics. From there, you have to worry about the E/VPA*/VMA*
logic; you can't feed an 8520 the E from a 68000 clocked at 14MHz and expect
it to work right, if at all. If you're trying this with other hardware, you
really need BR* and BG* to be clocked in and out of the 68000 on the right
clocks, or DMA devices may lock up.

In short, there's a reason accelerator boards cost more than 74F86s and
74F74s, and there's also a reason even some commercial accelerator boards
have trouble with expansion hardware.

>Now when I turn on the computer, it ALWAYS comes up dark grey first, then
>bright green replaces it. I get 10 short flashes, and one long flash, and it
>tries to restart.

Your 68000 is having trouble communicating either with the 8520, Chip RAM,
or very likely both.

>I am under the impression that DTACK will hold off weird bus cycles.

DTACK* gates 68000 access to Chip RAM, true. But you have to play by its
rules, as described above.

>In my opinion, this is a very very poor Hack.

In that, we are in complete agreement.

--
Dave Haynie Commodore-Amiga (Amiga 3000) "The Crew That Never Rests"
{uunet|pyramid|rutgers}!cbmvax!daveh PLINK: hazy BIX: hazy
Standing on the shoulders of giants leaves me cold -REM

[totoon...@gmail.com]

Thank you very much Dave for this answer.
I hope that will help me too for building a simple 14Mhz accelerator board for my CDTV.

Kind regards,

Richard