--
__ ____
/// Daniele Gratteri, Italian Commodore-Amiga user / ___|____
__ /// E-MAIL: daniele_...@inwind.it RITMO S75 | | |___/
\\\/// Nickname: FIAT1100D - ICQ: 53943994 FOREVER! | |___|___\
\/// URL: http://spazioinwind.libero.it/danielegratteri \____|
You could try dropping in a 68010/14 processor, no software needed.
There was also an old hack called AmiFast or something that had a
14mhz 68000 replacement chip and used software to select 14 or 7 mhz
for compatability. I sold mine years ago, when I got the wife an
030/25 accelerator for her then A2000.
--
Mike Leavitt ac...@lafn.org
> You could try dropping in a 68010/14 processor, no software needed.
Uhm... it seems the 68010 isn't really 100% compatible with the 68000 CPU...
If you are that worried about 100% compatability, even a 14 mhz 68000
is not totally compatible with a 7 mhz one, that is why the old
package I had, had software to run the machine at 7 mhz or 14. It
was called AdSpeed, BTW, and was a 14 mhz 68000 and a software
package. I never had any compatibility problems under OS 1.3 with it,
and never had to resort to running it at 7 mhz. Of course the speed
increase was negligible, even at twice the mhz. The 68010 is a little
faster.
--
Mike Leavitt ac...@lafn.org
>> You could try dropping in a 68010/14 processor, no software needed.
>Uhm... it seems the 68010 isn't really 100% compatible with the 68000 CPU...
Correct, but it's like 99% compatible with the 68000 in the Amiga.
--
| Apollo fastslot accelerators page - Http://www.canit.se/~glenn/apollo.html |
|----------------------------------------------------------------------------|
| ___ | Email 姣 Sha...@bay-watch.com |
| / __\ __ | Homepage 姣 http://www.canit.se/~glenn |
| __ / /__ / /__ ____ ____ __ | IRC 姣 XT600 @ IRC-net |
| (__/ /_ // / -_) _ ) _ )__) | Amiga - Silicon Graphics - 8bit comps. |
| \___//_/\__/_//_/_//_/ | T h e K i n g d o m o f S w e d e n |
Only 2 differences that are immediately obvious:
1. MOVE xx,SR instruction is ILLEGAL in user mode on a 68010+
you need to use MOVE xx,CCR instead.
MOVE xx,SR is legal in user mode on a 68000. A patch is required for
1.2,1.3 Kickstart AMIGA's which the CDTV probably will be in your case?
2. There are some slight optimizations for tight timing loop sequences,
which shouldn't really be used on a multitasking machine anyhow.
Can't think of anywhere in Kickstart (with the exception of
Trackdisk.device)
where any instruction based timing loops are employed.
Red.
> 1. MOVE xx,SR instruction is ILLEGAL in user mode on a 68010+
> you need to use MOVE xx,CCR instead.
> MOVE xx,SR is legal in user mode on a 68000. A patch is required for
> 1.2,1.3 Kickstart AMIGA's which the CDTV probably will be in your case?
I don't think so. All kickstarts from 1.2 and up know the 68010 and
the 68020 processors which both have the MOVE SR,xxx as priviledged.
Execbase contains a function (GetCC()) which keeps care of this.
Move xx,SR is a priviledged instruction all over the mot family,
including the 68000.
> 2. There are some slight optimizations for tight timing loop sequences,
> which shouldn't really be used on a multitasking machine anyhow.
A one-word instruction followed by an immedate "dbxx" does not re-
fetch the loop and the dbxx instruction and hence is a bit faster.
Somewhat more optimizations have been made concerning the instruction
set giving slightly more speed, but the overall picture is that it
isn't worth it.
So long,
Thomas
Hi Thomas,
Quite correct.
The instruction I was thinking of was:
MOVE[.w] SR,<adea> ; MOVE from SR
* priviledged on 68010/20/30/40/60; non-priviledged on 68000
*68000 <adea>.w = SR
*68010 etc. If S=1 then <adea>.w = SR
ELSE TRAP #8 and S (supervisor flag) = 1
I know about the GetCC() call.
If I remember correctly, the startup code on KS2.04 or later installs
a patch on TRAP #8 to catch any programs which do a MOVE SR,xx.
Whereas KS1.2/1.3 DO NOT, they need a patch program called
"decigel" or similar.
Some early Amiga programs do contain the MOVE SR,xx instruction
since one of the early versions of the C compilers (can't remember
whether it was Lattice or Aztec) generate that instruction sequence.
Really flogging my memory now, but I think WordPerfect4.1 for the
Amiga was an example of a program that bombed on a 1.2/1.3
machine fitted with a 68010. Anyone confirm?
Cheers,
Red
> If I remember correctly, the startup code on KS2.04 or later installs
> a patch on TRAP #8 to catch any programs which do a MOVE SR,xx.
> Whereas KS1.2/1.3 DO NOT, they need a patch program called
> "decigel" or similar.
AFAIK there is no Os supplied method to catch the situation where
a program executes a MOVE SR,xx. There is a "priviledged instruction"
trap, but that's used by the Os itself to execute code in supervisor
space which is forced by a MOVE xx,SR. However, this trap does not
try to attack the problem caused by the former instruction, and it
exists since the beginnings of exec.
> Some early Amiga programs do contain the MOVE SR,xx instruction
> since one of the early versions of the C compilers (can't remember
> whether it was Lattice or Aztec) generate that instruction sequence.
It's mainly an instruction relevant for assembly language and even
there it's rarely used except by the exec kernel itself; I've no
idea why a C compiler might want to generate it as there are clearly
more direct methods to make use of the condition codes of the processor.
> Really flogging my memory now, but I think WordPerfect4.1 for the
> Amiga was an example of a program that bombed on a 1.2/1.3
> machine fitted with a 68010. Anyone confirm?
I don't know, sorry.
So long,
THomas
And a 14mhz 68000 is even slower. It is really not worth upgrading to
anything less than a 68020 or better, but I don't know if you can do
that with a CDTV.
--
Mike Leavitt ac...@lafn.org
All I can confirm is that I ran it from Floppy on a 68030/16 on OS 2.1
and 2.04 chips in an A3000, but I could never get it to install to
hard-disk. I liked ProWrite better anyway, and later Final Copy and
of course Final Writer, so I did not pursue it.
--
Mike Leavitt ac...@lafn.org
>"Mike Leavitt" <ac...@lafn.org> ha scritto nel messaggio
>news:3D6C85A9.MD-...@lafn.org...
>
>> You could try dropping in a 68010/14 processor, no software needed.
>
>Uhm... it seems the 68010 isn't really 100% compatible with the 68000 CPU...
Nowadays with the 020 and better CPUs having been around for so long, the
compatiblity problems really aren't an issue. I've been running an 010 in my
"serious" A500 for ages and no problems.
Of course some old games and demos are bound to fail, but then again, they'd
fail on an 020 or better too.
--
CBM, Amiga and PEZ-nut, Scout, Glider pilot, Administrator
Email: vjo...@sci.fi, URL: http://www.sci.fi/~vjouppi/
GSM: +358-40-5679999, IRCNet: Jope
**** COMMODORE BASIC V2 ****
You need a bit more than just a faster rated chip .. You need to halve the
clock line from the motherboard main clock (around 28MHz) and feed that to the
clock pin on the CPU which you've isolated from the socket.
Hairy, not recommended, better get a real turbo card. :-)
DG>
DG> I would like to try on my CDTV the 14MHz overclock project called
DG> "14MhzA500.lha" which I have found on Aminet. The project
DG> requires a 16MHz 68000 CPU but I've only found an old Motorola
DG> MC68000P12 CPU, which I think is the 12.5MHz version: do you
DG> think it can be overclocked to 14MHz or it won't work at all?
DG>
The ICD ADSPEED is still made and sold by icd i think. check
www.icd.com. its slightly faster than adding just a faster 68000 as i
understand it,it has a small cache onboard. Its no speed demon in any
case.
Theoretically the CSA derringer 030/50 might work in it(it can hold
32M of fast 32bit ram). The MegaMidgetRacer is an old 030 that may
also work. VXL-030 might be one to check too. The Phase5 Viper520 and
Viper530 may also. i have the viper 520 in my A1000 atm,planning on
testing if it will even fit in the cdtv soon as i have a chance.
If your looking for a homemade accelerator,there was the old LUCAS
board. plans are still available for it on the net with a 32bit addon
ram board. it worked in the 1000,and many adapted it to the A500.i
dont see why it wont work in the cdtv.The lucus looked like a decent
design and well thought out. i always wanted to try one.
good luck,
Mike
--
------------------------------------------------------------------------
Posted with Amiga NewsRog
------------------------------------------------------------------------
Not exactly correct. It depends on the instructions the program executes.
The improvement can be up to 10 to 15 %. As an example run Gravity Wars,
from (I think) Fish disk 105. It is unbearably slow on a 68000, but becomes
playable on a 68010.
--
Jerry wa2rkn darkstar at idsi dot net
I think any speed test I ran only showed about a 3% to 5% increase. Just
barely enough to be occasionally noticable. The 68010 also allows you to
move the vector base register (VBR). Supposedly if you have fast RAM, by
moving the VBR into fast RAM you can improve serial port reliability, and
software utilities are available to do it.
My 68010 has worked flawlessly for me in my CDTV for years. I never
noticed any incompatibilies with CDTV titles. Any software that would
have problems with the 010 would not work on newer processors either.
Is it worth it? That's up to the buyer. Check around and see what other
accellerators work with CDTV. There are possibly some other 14MHz 68000
ones you could pick up on eBay just as cheaply.
Craig
In australian Dollars.
http://cgi.ebay.com.au/ws/eBayISAPI.dll?ViewItem&item=2057395202&rd=1