40MHz 68000 /Overclocking 68000 machines (Amiga especially)
Frostycat
In article <Pine.HPP.3.96.970808122803.23512I-
100...@demeter.sunyit.edu>, Pat Fish <fi...@sunyit.edu> writes
>If anybody has experience with homebrew 68000 speed-ups on Amigas, let me
>know. I've heard conflicting reports (including ones from Dave Haynie) that
>say these hacks aren't overly reliable. Yet I keep hearing people say they
>work. I wonder how they get the 68000 to slow down to R/W to the
>motherboard.
Actually, staying on this subject, it is possible to accelerate your
Amiga slightly to 12mh/z with a Sega Genesis/Megadrive processor.
I read this somewhere in Amiga Format ages ago.
This kid wrote in saying how he'd done it issuing the following
instructions on how to do it:
1> Obtain an old Sega Genesis/Saturn
2> Open up both the Amiga and Sega so the CPU is showing.
3> The CPU on the Sega is 12mhz compared to the A500's 7.1mhz
4> Note the notches on the chip :
The 68000 Chip
--------------
|||||||||||||||
---------------
notch ----> O
---------------
|||||||||||||||
5> Take out the CPUs and swap 'em, being careful to insert them
in the correct notch position. If they're anything to go by the
KickStart chips, be careful not to bend the damn pins!
6> Reconnect 'em
7> Boot up and check out the A500 cruising at 12mhz ;)
Try and play Sonic on the Sega ;)))
Actually, this leads to more interesting questions like "What if
I put a Blizzard 68000/14mhz in my Sega?" and "How high can I clock a
68000 at all?" (Best speed so far : 100 mhz)
Don't blame me if it don't work, it's just what I read!
---------------------------------------------------------------
******** ******** ******* ******* ******** ** **
** ** ** ** ** ** ** **
******** ** ** ** ** ****** ** ****
** ** ** ** ** ** **
** ** ******* ****** A1200 ** **
--------------------------------------------------------------
Amiga Music Massive => HTTP://www.jaygee.demon.co.uk
--------------------------------------------------------------
Kenneth Jennings
Nope. Not unless the directions also included a swap out of the clock
(oscillator) for the CPU. Otherwise, it's still running at 7.159..Mhz
...Unless maybe the Sega used a custom version of the 68000 that
supplies its own clock...
But, this upgrade sounds more like urban computing myth.
Kenneth Jennings -- SyntheToonz -- Miami, FL
kenneth at daffy dot aatech dot com
ken at ntalr dot aatech dot com
Pat Fish
If anybody has experience with homebrew 68000 speed-ups on Amigas, let me
know. I've heard conflicting reports (including ones from Dave Haynie) that
say these hacks aren't overly reliable. Yet I keep hearing people say they
work. I wonder how they get the 68000 to slow down to R/W to the
motherboard.
As for the debate over the 68HC000. Indeed, this 16MHz part can run at
40MHz. Motorola claims there is no 68HC000 that runs faster than 16MHz.
Yet Supra has 28MHz 68000 accelerators, and Motorola doesn't seem to mind.
Nobody in the Amiga community noticed this paradox. So I started making
phone calls to engineers familiar with the issue. . According to an
engineer who has little use for non-disclosure agreements, it's all about
suits and republicans.
Originally the 68HC000 was destined to be a 32 or 33MHz part. Around the
same time, Moto was about to release it's CPU32 cores for embedded use,
etc.. Trouble is, the pricing they set meant, that if you ran the numbers, a
32MHz 68HC000 gave you more bang for the buck than their new CPU32 cores.
Rather than accepting this, or even changing prices, the beancounters at
Moto decided that the 68HC000 series would be arbitrarily advertised as a
16MHz part (maximum). Another victory of brainless greed over good
engineering.
Anyway, he said they typically "burned in" (tested) the 68HC000 at 36 or
40MHz, and they used them around the company at those speeds.
Supra uses a cache apparently to get around the issue of how to get a
33/36/40MHz 68000 to talk to the motherboard (7 or 14MHz). The question is,
how does one get a 14MHz (or faster) 68000 to work correctly in a
do-it-yourself situation?
On 4 Aug 1997, Howard Chu wrote:
>In article <5rqkh4$g...@a-tuin.diginet.de>,
>Michael Schwingen <rinc...@a-tuin.diginet.de> wrote:
>>Thomas A. Sherman <to...@microware.com> wrote:
>> but I'm afraid the Motorola chips are not as
>>>overclocking friendly as x86 chips are.
Just the opposite is true. Intel overestimates their ratings, and tests
them as lots fail. If a 133MHz chip fails, they sell it as a 100MHz part,
etc. Moto is very stringent and does just the opposite. They underestimate
their envelope.
>My experience agrees with yours. But just stop for a second and think
>about this - if the chips really could run reliably at the higher clock
>rate, don't you think Motorola would have listed them at the higher
>clock rate when selling the part?
No. Engineers don't run these companies. Moron moneygrubbers do. See
above explanation.
>But as another footnote, I've seen 16 MHz 68000s running at 36 MHz without
>any problems. I guess these HCMOS parts are really even better than Motorola
>lets on, despite their own hype. Perhaps they're conservative with the 68000
>because they don't perceive a market demand for 32 MHz 68000s.
Very good guess. BTW where have you seen 68HC000s running at 36MHz?
-Pat Fish
PS I have Jaz disks for sale. Not coincidentally, I'll looking to buy a Jaz
drive, and am looking for comments on the Jaz (especially from Amiga users).
Howard Chu
In article <Pine.HPP.3.96.970808...@demeter.sunyit.edu>,
Pat Fish <fi...@sunyit.edu> wrote:
[Thanks for the story, very enlightening...]
>On 4 Aug 1997, Howard Chu wrote:
>>But as another footnote, I've seen 16 MHz 68000s running at 36 MHz without
>>any problems. I guess these HCMOS parts are really even better than Motorola
>>lets on, despite their own hype. Perhaps they're conservative with the 68000
>>because they don't perceive a market demand for 32 MHz 68000s.
>Very good guess. BTW where have you seen 68HC000s running at 36MHz?
Jim Allen used to run a company called FaST Technologies making accelerators
for Atari STs. You could order them in a variety of speeds, from 16 MHz up to
48 MHz, compared to the 8 MHz in the original ST. I think they're out of
business now, but not because their products were bad. I had one of their
original 16 MHz accelerators... They all used a 32K SRAM cache, because
they still had to interface to the bus at 8 MHz.
It still boggles my mind that the 68000 was pushed so far ahead and the
68010 was canned. I hope never again to see user code doing move sr,d0 ...
A 48 MHz 68010 - I would have bought that.
--
Howard Chu Principal Member of Technical Staff
h...@locus.com PLATINUM technology, Los Angeles Lab
Advertisements proof-read for US$100 per word. Submission of your ad to my
email address constitutes your acceptance of these terms.
Matthew Garrett
> Actually, staying on this subject, it is possible to accelerate your
> Amiga slightly to 12mh/z with a Sega Genesis/Megadrive processor.
>
> I read this somewhere in Amiga Format ages ago.
> This kid wrote in saying how he'd done it issuing the following
> instructions on how to do it:
> 7> Boot up and check out the A500 cruising at 12mhz ;)
>
speed that the computer drives it at that determines how fast it goes.
Just putting in a 12MHz chip won't make it go at 12MHz, so there'll be
no difference at all. If you put a 7MHz 68000 in the Megadrive, it'd
work, but would probably start crashing due to lower tolerances.
--
Matthew Garrett | ca...@enterprise.net | A3500/18/3.1/Retina |