[comp.sys.cbm] 4MHz C-64's

cyaa01@mixcom.UUCP (Chris Klausmeier) (11/14/90)

In article <5527@uafhp.uark.edu> bbs00068@uafcseg.uucp (Joel Kolstad) writes:

/I'm not.  I'm running a C-64 at 4MHz using a Schnedler Systems Turbo 
/Accelerator.  Availiable from Schnedler Systems directly... about $200.
/
/And boy, is it great!  (Too bad you can't use a 1764 with it, though!)

Just how compatible is it? Does anyone know if Image BBS will run under
such a beastie?

/						---Joel Kolstad
/						kolstad@cae.wisc.edu

-- 
[][] Chris Klausmeier [] "For the words of the profits were written on the
[] cyaa01@mixcom.UUCP []  studio walls... concert halls" - Neil Peart - RUSH

treesh@vangogh.helios.nd.edu (11/15/90)

I have seen it work quite nicly with the Image BBS program.  The system was
probabaly the sweetest system I have ever called running on a C64.  It
was running Jiffy Dos with the CMD hard disk and the 4Mhz!  Very quick!

ctfm

cyaa01@mixcom.UUCP (Chris Klausmeier) (11/17/90)

In article <1990Nov15.131943.1979@news.nd.edu> treesh@vangogh.helios.nd.edu writes:

/ I have seen it work quite nicly with the Image BBS program.  The system was
/ probabaly the sweetest system I have ever called running on a C64.  It
/ was running Jiffy Dos with the CMD hard disk and the 4Mhz!  Very quick!
/
/ ctfm

What is the name/number of that board? I'm curious enough to actually
waste some LD money on checking it out. (wow! heh) 


-- 
[] Chris Klausmeier -- cyaa01@mixcom.UUCP []
[] if I could wave my magic wand... [][][][]

bbs00068@uafcseg.uucp (Joel Kolstad) (11/19/90)

Since  I  posted  my  message about Swiftlinks and accelerated C-64's,
there's been a flurry of activity on the subject, so I figured I'd try
and tell you something about the system here.
 
The  accelerator  is  called  a  "Turbo  Master  CPU,"  and is made by
Schnedler  Systems,  of Asheville, North Carolina.  Their phone number
is  (704) 274-4646, and they'd be glad to send you lots of information
on it.
 
The  Turbo  Master  contains  a  65C02 CPU operating at 4.09MHz.  When
power  is  applied, the Turbo Master takes over the C-64 and your C-64
basically  serves as a keyboard and a VIC/SID/6526 I/O box!  The Turbo
Master  continues updating the C-64 RAM at the old 1MHz frequency, and
updates  it's  own  RAM  (yes,  it has it's own 64K CMOS RAM) at 4MHz.
This  makes  some  interesting  things  possible  --  you can use your
assembler  and  assemble  a program at 4MHz, "switch out" (more later)
the  accelerator,  and  be  left  with  a  1MHz  machine that has your
assembled  program in it, ready to go.  Run your program, and find out
if  it  works.  As is often the case, it may well crash...  but no big
deal!   Just  switch  the accelerator back in, and you can change your
program,  reassemble,  and  try  again.   This scenario emphasizes the
"separate but combined" nature of the C-64/Turbo Master interface.
 
On  to the switches...  There are four swtiches on the Turbo Master --
Disable/Enable,  Rom2/Rom1,  Manual/Soft, and 1MHz/4MHz.  As discussed
above,  disable/enable  switches the whole device in and out.
 
The Turbo Master contains a 32K EPROM storing two sets of Basic/Kernel
ROM  images.  Rom2/1 selects between the two.  As shipped, you get the
standard  ROM  and Schedler's "feature" ROM which contains a DOS wedge
and  high  speed  disk  loader/saver.   If  you  own JiffyDOS, you can
request  that  they  send  you  the  standard ROM/Jiffy DOS ROM option
instead.   (NOTE:   The Jiffy DOS ROM option is Jiffy DOS version 5.0.
The  newest  Jiffy DOS is version 6.0.  CMD, who makes Jiffy DOS, says
that  they  allowed Schnedler Systems to use version 5.0 "as a favor."
They  say that it took them a lot of time to get it to work on a Turbo
Master,  and that the extra sales they made from it didn't justify the
time they put into it.  Therefore, CMD says they will NOT be releasing
JiffyDOS  6.0  for Turbo Accelerators.  I'm personally a little peeved
about  this,  and have called them up to discuss it.  They say if they
knew  enough  people  would buy JiffyDOS for Turbo Masters, they might
release  it.  So if you have a Turbo Master and Jiffy DOS, or might be
getting  such a setup, CALL THEM UP!  LET THEM KNOW YOU WANT JIFFY DOS
6.0 FOR TURBO ACCELERATORS!)
 
OK,  now that I'm done yelling at CMD, back to the switches.  The next
switch,  Manual/Soft  determines  whether  software  can  control  the
Acclerator's speed.  In the "Soft" position, bit #7 or memory location
0  swtiches  the  accelerator  between 1MHz and 4MHz.  In the "manual"
position,  the  final switch, 1MHz/4MHz, forces the accelerator to one
speed or another.
 
That's  the general overview of it.  It plugs into the expansion port,
and  works  well  with  cartridge expansion ports.  This allows you to
have  Swiftlink carts, MIDI carts, I/O boards, etc.  going at the same
time  as  your Turbo Master.  (As I mentioned in the original article,
with a Swiftlink and Turbo Master, scrolling high res screens at those
high  bauds  ratesi s no problem!) Note, however, that devices that do
direct  memory  access  with  the  C-64  won't  work  with  the  Turbo
Acclerator.   As  far  as I know, the only device that many people out
there  might  have  that  does  this  is a Commodore REU -- a 1764, or
equivalent.  I'm not sure if GeoRAM uses DMA or not.
 
In  gerneral,  software  compatbiliity is excellent.  There are really
only  two  "gotchas:"  Unimplmented Op-codes, and timing difficulties.
As  I mentioned, the Turbo Master uses a 65C02.  A 65C02 takes many of
the  "undefined" 6502 opcodes, and defines them.  However, the defined
instructions  are  what Rockwell thought would be useful, not what the
"undefined"  6502 instructions did anyway.  The new instructions allow
clearing registers, better register swapping, better bit handling, and
many  other  improvements.   (And  Schnedler Systems does sell a 65C02
assembler, by the way.) However, copy protection routines that use the
undefined  6502  opcodes have about a 0 probability of working.  Since
games  are  by  far the programs that usually fall into this category,
you  may well be forced to play your games at 1MHz.  (Oh well!) As far
as  I  know,  no  "serious" programs suffer from this problem.  (GEOS,
PaperClip  III,  PaperClip  Publisher  all  work  fine.  And PaperClip
Publisher verges on being bearable at 4MHz!)
 
The  other  problem  relates to timing:  Routines that rely on PRECISE
timing  will  obviosuly  get messed up.  For example, since the serial
bus  still  only  goes  the same speed, the ROMs have "switch to 1MHz"
instructions  placed  before  the  Kernel  LISTEN, TALK, UNTALK (etc.)
routines.  The serial routines are about the only built in 64 routines
that  were  patched.   The  RS-232  routines only need to be "tweaked"
(since  the  6526  still only interrupts the processor based on a 1MHz
clock, up to about 1200 baud nothing changes).  Tweaking the vlaues is
a  simple  matter,  especially  for terminal programs such as NovaTerm
that  come  right  out  and  ask  you for timing values.  As you might
expect,  at  4MHz  a  64 can keep up with 9600 baud, although it takes
some  fiddling to get it to work.  (...and then you have to figure out
what you're going to do with 1K/sec...)
 
Programs  that  use  "busy-wait loops" also run in 1/4 the usual time.
This means that simple programs like BASIC "play a tune" programs will
run  fast.   It  should  be noted that almost all commercial games use
interrupt driven music and movement, and rarely exhibit this problem.
 
If  you're  making your own programs, it's very easy to detect a Turbo
Accelerator, and change any "critical" timing values that you have to.
Detection  takes anywhere from 5-50 bytes, depending on whether you're
just  detecting  "Normal"  vs.  "65C02" system (and then assume it's a
Turbo   Accelerator)   or   "Normal"   vs.    "Fast"  system  (and  do
whatever...).
 
That's  about  all  I can thing of on it at the moment.  I believe the
current  price  is  $199.  It also has a reset switch.  Call Schnedler
Systems  and  they'll give you all the answers you want (if you have a
really  technical  question, ask for Steve Schnedler, who designed the
unit).
 
If  you  have  relatively  straightforward  questions, I'll attempt to
answer them if you send me e-mail at kolstad@cae.wisc.edu.
 
Hope  this  answers  a  lot of Usenetter's questions!
 
								 	 ---Joel Kolstad
					---Joel Kolstad
					kolstad@cae.wisc.edu