Re: TWGS

[supertimer@aol.com (Supertimer)]

Liam Busey <liam1@ix.netcom.com> wrote:

>CBS wrote:
>>
>> (FOLLOWING IS THEORY, based on the facts I've collected)
>> ***************************************************************
>> the REV A Gals were made when the FIRST 65816 chip came out.  This
>> would mean no bugs that you'd have to slow it down for, thus the A
>> gals. Then they sped up the CPU, but not flawlessly, so they had to 
>> slow down certain instructions in order for them not to crash the
>> CPU/TWGS card. That required the addition of the B (and other???) gals
>> ***************************************************************
>
>I don't believe the original 65816 was ever sped up. I think WDC just
>had certain batches that were more reliable at higher speeds than
>others. Kind of like CRT quality in monitors.

That was the case with the older 65816 chips.  I think CBS meant
that AE and Zip Technologies overclocked the original 65816
to make accelerators.  But they could not do this without breaking
the chip at certain instructions.  That's why they needed "slow
down" bug traps (the "B" GALs in TWGS and the slow down
logic in the ASIC of the ZipGS).  When the redesigned by
Sanyo WDC 65C816 units came out, the "buggy at high speed"
instructions became reliable at high speeds, so the "B" GALs
were no longer needed (except for 2B if you have a RamFAST,
it seems).

Incidentally, does anyone notice that the TWGS runs A LOT
cooler with the 14Mhz redesigned 65C816 units than the
original "buggy" 65C816?