Date : Fri, 24 Nov 2006 10:44:40 +0100
From : s-whs@... (W.Scholten)
Subject: : BASIC variations
Pete Turnbull wrote:
>> Littlefield Aaron wrote:
>>
>>
>>> Please excuse my ignorance here, but purely out of interest, is it
>>> possible to put a Master BASIC IV ROM into a humble Model B, or B+?
>>>
>>>
>>
> If you replace the 6502 in the Beeb with the right type of 65C02, it
> will work. I've done that. However, there are two families of 65C02
> with slight differences in the extended ocodes, and only one works.
> Sorry, I can't remember which ;-)
>
>
Rockwell's 65C02 has more instructions. I've seen two soli 4MHz 256K
accelerator boards, one had a 65SC02, the other a R65C02 with basic 4
EPROM. The second board's CPU may have been replaced with the Rockwell
version by the owner to get Basic 4 running. You can then also use the
Basic time functions with e.g. the Solidisk RTC board. It's also
interesting that Basic 4 is much faster at math. See my webpages for
more information (pictures section). Btw, I had no problems with any of
the 4MHz boards being unreliable, although I did not use any for long
periods of time. The only problem I saw was that switching between 4MHz
and 2MHz usually works fine but there's about 1 in 10 chance of the BBC
hanging. I switched between 2MHz/4MHz a lot for example while testing
the speed with Elite (Solidisk mentioned the speed at which Elite runs
in their ads, so I checked if this speed difference was really as
impressive as they implied. I wasn't too impressed. In reality, the ca.
1.4x speedup is not all that noticeable. I think Solidisk could only
have made it look impressive by setting up 2 BBCs, one on which Elite is
running at 2MHz, and another at 4MHz)
As I mentioned a few years ago, the Micro user mentioned not relying on
the Rockwell instructions but for the 6502 2p I've only come across
units with the Rockwell CPU. I was also interested in knowing which
games or other programs depended on the 6502's illegal ops to know if it
would be much trouble to replace the 6502. It would be nice to replace
the 6502 and 6522s with CMOS versions, to reduce the heat a bit :)
Conclusion: 65C22s are almost impossible to find, and there may be too
much hassle using a 65C02.
Another thing: When the 65C02 came out, I was disappointed. I'd much
preferred a CPU that was just the 6502 but which eliminated as much as
possible the dead bus cycles (to get to near 100% bus usage). I made
some stats using optimized assembler examples at that time which showed
a large increase in speed in case that was implemented. Much more than
what you can get with the new instructions. The examples with PLX/PHY
etc. didn't impress me as all that great. And why no MUL?
Wouter
--
BBC micro | Calculators | Classic PC games: http://www.xs4all.nl/~swhs/whs/