honp9@menudo.uh.edu (Jason L. Tibbitts III) (03/09/91)
The ATonce is a hardware IBM AT emulator similar to the AT bridgeboard,
except it is designed for the Amiga 500 instead of the Amiga 2000. It
performs the emulation quite well, but problems with its setup software
and possible problems with various Amiga configurations detract from its
usefulness except on a bare-bones A500 system.
Lately a lot of people have been wondering about the ATonce 80286
emulator for the Amiga 500. Last week I had the opportunity to test
one of these and unfortunately did not have much success. Here I will
explain what the thing is (if you don't know), how it's supposed to work,
and the problems I had. All testing was done on a Revision 5 Amiga 500
with the new Agnus, 1 meg chip ram, 42ms 48Meg Seagate st157-n hard drive,
Trumpcard controller and 2 meg fast ram. I also have both AmigaDos 1.3 and
2.0, but since 2.0 is loaded to ram it's pointers seem to be reset when
the ATonce is used. For this reason testing was done using Kickstart
and Workbench 1.3. The product was tried both with and without the A500
expansion hardware and on a 2000, but the latter two configurations were
tried only briefly while first was used for about a week.
[Ed. Note: If you have used the ATonce on other configurations, or have
any other information which you wish to add, please send mail to
HONP9@menudo.uh.edu with your experiences.]
At the end of this review are benchmarks and test results obtained
with Norton Utilities and CheckIt.
----------------------------------------------------------------------
The ATonce is an 80286 AT emulator from a German company 'Vortex'.
Their address:
Falterstrasse 51-53, D-7101 Flien bei Heilbronn, Germany.
phone : 07 131 58 72-0
e-mail: 100016,2545@compuserve.com
I tested a unit from:
Computers Plus,
Ralph Doncaster, 880274d@aucs.AcadiaU.Ca
It is a very small board containing a Motorola 68000. This is important
because many Amigas have third party 68000's and Vortex claims they need
the stability of a real one. (I don't know if this is so, but at least
it is easier to install; you simply remove the existing 68000 and plug in
the card) The board also contains an Intel 80286 and a large custom gate
array containing the BIOS and other custom circuitry. Along with this is
a special "double socket" to put under the Gary chip and initialization
and setup software.
Once I had the ATonce I immediately installed it following Vortex's manual.
This caused some problems. I had the CPU card and the Gary module installed
correctly, and I'd run the setup program and told it to use default settings,
but I could not get the program to work. When you start the initialization
software it normally sets some pointers to the ATonce code, then reboots
the Amiga. At this point you get both an Amiga startup screen and a
standard MS-DOS screen complete with memory and BIOS test. On my
system I was losing the memory required for the ATonce, but was not getting
that MS-DOS screen. I tried with and without my HD, expanded memory,
even tried an old Agnus, but no luck. It turned out that on my system
it only works WITHOUT the Gary module. The manual says that the Gary module
may cause problems finding an A590 HD, but apparently causes problems on
older A500's as well. I should mention here that the purpose of the Gary
module is to avoid timing differences between internal and external
memory expansion in the 500, and is therefore unnecessary in a 2000.
The manual says that it will not work with the 2000, but someone else here
tried it in his and it seemed to work about the same as on my 500. (As you
will see this should be read FLAKY.) On the 2000 ATonce would initialize
about half of the time and would run MS-DOS software once it was up.
The extent of our tests was to boot DOS and exit so perhaps someone else
(or maybe myself if I get time) could do more in depth tests on the 2000.
[Ed. Note: If anyone has done this, please sent mail to HONP9@menudo.uh.edu.]
While attempting to get the ATonce working the first time I noticed that
the manual mentioned problems with older revision motherboards. There
are two files on the program disk that have the extension .dsg. Only one
of these is used by the ATonce, but apparently they are modifications
for different revision mother boards. The files are named "atonce.dsg"
and "other.dsg". The manual says something to the effect of:
"If your motherboard is an older revision than 6A you
may need to rename other.dsg to atonce.dsg to obtain full
performance from your ATonce. Do NOT ever use the file
other.dsg on a revision older than 6A."
Yes, I know this contradicts itself, but it is in the manual (not
a direct quote as I have returned the product, but close enough).
To confuse things even more their Erratum listed the passage as:
"If your motherboard is newer than revision 5 you may need
to rename other.dsg to atonce.dsg to obtain full performance.
NEVER use the file other.dsg on a revision newer than 5."
Again this is not a direct quote, but it is just as confusing and explains
the problem. Not knowing which to use I tried with both and found
they both work, but other.dsg was faster according to Norton's SI.
Now at this point I have the ATonce program loading from my HD, but
DOS booting from floppy and have only run MS-DOS and SI.
The biggest problems start when trying to install ATonce so that MS-DOS
can use the Amiga HD. ATonce allows for HD support as either a file
on an AmigaDos partition or a full MS-DOS partition. First I tried
to create a file. The manual and setup program are straight forward.
You select 'Hard Drive' and tell it what you want following prompts AND
the manual. The format for file is:
HD:path/filename,first_dos_track,last_dos_track
with 50 tracks being about 5 megs. For a partition you simply give the
partition name (it must be a mounted device).
After creating a file the right size I ran ATonce. No luck. It used
memory when I rebooted, but didn't initialize the card. The next time
I rebooted, AmigaDos couldn't validate the hard drive containing the
partition file. I managed to validate the drive with AmigaDos 2.0 (not
sure why 1.3 wouldn't) and the tried again after doing a cold boot.
This time the ATonce found the file and used it as C:. I was able to
make this a boot partition using Zenith's PART program, format it, and
install Zenith Dos 3.30 plus. I rebooted and tried again. This
time the ATonce worked correctly, and even booted from the HD file.
Great, now it was working! WRONG!
Now every time I changed something in the ATonce configuration my HD
wouldn't validate. Eventually it was beyond even FixDisk and was
reformatted. Since I had to reformat I tried creating a MS-DOS partition
in addition to a file. This did not go well either. I could get the
partition recognized, but when I tried to format or partition it
the system would crash. Amiga-Amiga-S is a hardware reset for the ATonce,
and even this did not work at this point. (One thing I did like was
that it never crashed or even affected the Amiga side once it was running.
The setup and initialization programs often did, but MS-DOS didn't.)
OK, enough problems with the hardware. Now I created a partition file
and formatted it and decided to test the thing's MS-DOS compatibility.
After great pains with disk drive being trashed I got ATonce working.
With my memory configuration (3 megs) I was able to get ONLY 640K base
and 512K extended. This left over 1.5 megs for the Amiga Side, but
the way the ATonce software operates it couldn't access it.
My configuration is:
$8c2 - $7e7ff : original chip ram
$80000 - $fffff : extra 512K chip ram
$280000 - $2fffff : 1st Meg fast ram
$380000 - $3fffff : 2nd Meg fast ram
ATonce would let me use $80000-$fffff and either Meg of fast ram, but
it would not allow me to use both megs of fast ram, even when I specified
it with an option "MEMMODE" which is provided to allow user specified
memory usage.
Now, after figuring out the memory problem I got ATonce running. It 'SI'ed
at :
other.dsg atonce.dsg
---------------------------------------------
CI 5.9 4.6
DI 0.2 0.2
PI 4.0 3.1
It never reached the 6.1 that Vortex advertises, and I was able to slow it
down to 1.6 by having high activity on the Amiga Side. (Moving screens
and graphics demos, etc.)
Once running the ATonce performed very closely to what is advertised.
It had printer support, RS-232 support (although CheckIt reported errors
on COM2), CGA, Hercules, and Olivetti graphics and text support. There
was even the pitiful MS-DOS sound. There are only 2 real problems:
1. The Amiga mouse is supposed to be emulating a Microsoft mouse.
This was not a very good emulation. MOUSE.COM was the only program
that would find my mouse. None of the Microsoft utilities would
find it, a graphics program I've written (which uses a Microsoft
mouse fine on PC's) wouldn't, CheckIt wouldn't, and neither would
any other program I ran.
2. Apparently certain access to interrupts can crash the ATonce. When
CheckIt tested interrupts it crashed, and when I tried to use the
program FastBack Plus it crashed. (These are, however the only two
cases I had it crash from the MS-Dos side other than formating the
HD.)
I don't have much to say about the MS-DOS emulation except that it is
good when it works. I ran games and utility software with equal success.
Here is a list of tested programs, all of which worked exactly as on
a real AT, except for mouse support and FastBack Plus (which crashed).
Norton Utilities
CheckIt
Turbo C 2.0
Turbo Assembler (From C 2.0)
Turbo Debugger (From C 2.0)
Turbo Pascal 5.0
Word Perfect 5.1
Procomm
Firepower
Silpheed
Mouse
FastBack Plus
This is not a long list, but I think that it is a good indication of
what the ATonce is capable of running. I think that the only problem
is that there are some configurations of hardware that were not thoroughly
tested prior to the ATonce's release. When running on a stand-alone,
A500 with only one meg of ram and with the Gary module uninstalled, the
ATonce seems to work fine. I think it is a good emulator if you want
stand alone AT in an Amiga case and are happy with only 640K memory
(and 86K on the Amiga side). I was attracted to the ATonce by the promise
of support of hard drives, expanded memory and the mouse.
In these three areas it is sadly lacking, but if there is a new release
of the support software this will be a very useful and long awaited
package.
One last thing that may be interesting is the quality of translation
from German to English in the manual. I found it humourous that the
new Agnus was referred to as the "REALLY BIG FAT AGNUS" rather than
new or super or fatter Agnus.
Benchmarks:
-----------
Norton SI Results With atonce.dsg file being used:
SI-System Information, Advanced Edition 4.50, (C) Copr 1987-88, Peter Norton
Computer Name: IBM AT
Operating System: DOS 3.30
Built-in BIOS dated: Friday, November 30, 1990
Main Processor: Intel 80286 Serial Ports: 2
Co-Processor: None Parallel Ports: 1
Video Display Adapter: Color/Graphics (CGA)
Current Video Mode: Text, 80 x 25 Black and White
Available Disk Drives: 3, A: - C:
DOS reports 640 K-bytes of memory:
73 K-bytes used by DOS and resident programs
567 K-bytes available for application programs
A search for active memory finds:
640 K-bytes main memory (at hex 00000-0A000)
128 K-bytes display memory (at hex 0A000-0C000)
512 K-bytes extended memory (at hex 10000-18000)
ROM-BIOS Extensions are found at hex paragraphs: C400
Computing Index (CI), relative to IBM/XT: 4.6
Disk Index (DI), relative to IBM/XT: 0.2
Performance Index (PI), relative to IBM/XT: 3.1
-----------
Norton SI results with other.dsg file being used:
SI-System Information, Advanced Edition 4.50, (C) Copr 1987-88, Peter Norton
Computer Name: IBM AT
Operating System: DOS 3.30
Built-in BIOS dated: Friday, November 30, 1990
Main Processor: Intel 80286 Serial Ports: 2
Co-Processor: None Parallel Ports: 1
Video Display Adapter: Color/Graphics (CGA)
Current Video Mode: Text, 80 x 25 Black and White
Available Disk Drives: 3, A: - C:
DOS reports 640 K-bytes of memory:
73 K-bytes used by DOS and resident programs
567 K-bytes available for application programs
A search for active memory finds:
640 K-bytes main memory (at hex 00000-0A000)
128 K-bytes display memory (at hex 0A000-0C000)
512 K-bytes extended memory (at hex 10000-18000)
ROM-BIOS Extensions are found at hex paragraphs: C400
Computing Index (CI), relative to IBM/XT: 5.9
Disk Index (DI), relative to IBM/XT: 0.2
Performance Index (PI), relative to IBM/XT: 4.0
_______
CheckIt Results:
CheckIt 2.1 Activity Log
3/2/1991 22:36:18
=== CONFIGURATION INFORMATION ===
DOS Version: 3.30
ROM BIOS: (Standard) BIOS Date: 11/30/90
Processor Type: 80286 AT Machine
Math Coprocessor: Not Present
Base Memory: 640K Available: 566K
Extended Memory: 512K Available: 512K
EXPANDed Memory: No EMS driver installed
Video Adapter: CGA
Video Address: B800h Video RAM Size: 16K
Hard Drive(s): Drive 0 (C:) = 18M
Floppy Drive(s): A:720K(3.5")
Clock/Calendar: CMOS Clock
Parallel Port(s): LPT1=378h, LPT2=278h
Serial Port(s): COM1=3F8h, COM2=2F8h
Mouse: None Joystick(s): None
=== CLOCK/CALENDAR TEST ===
Compare Current Time............................................FAILED ***
DOS: 22:38:03.28 Real-Time Clock: 22:38:26.00 (22.72 apart)
Compare Current Date............................................Passed
DOS: 03/02/1991 Real-Time Clock: 03/02/1991.
Real-Time Clock Alarm...........................................FAILED ***
Compare Elapsed Time............................................FAILED ***
DOS: 18.84 Seconds Real-Time Clock: 20.00 Seconds(1.16 apart)
*** END TESTS: 3 ERRORS ENCOUNTERED ***
=== COM1 SERIAL PORT TEST ===
Loopback Plug...............................................Not Present
Test data register..........................................Passed
Test interrupt enable.......................................Passed
Test interrupt id...........................................Passed
Test line control...........................................Passed
Test modem control..........................................Passed
Test line status............................................Passed
Test modem status...........................................FAILED ***
Expected 05h, received 15h.
Test different baud rates...................................Passed
*** END TESTS: 1 ERROR ENCOUNTERED ***
=== COM2 SERIAL PORT TEST ===
Loopback Plug...............................................Not Present
Test data register..........................................FAILED ***
Received unexpected character: 00h.
Test interrupt enable.......................................Passed
Test interrupt id...........................................Passed
Test line control...........................................Passed
Test modem control..........................................Passed
Test line status............................................FAILED ***
Received unexpected character: 00h.
Test modem status...........................................FAILED ***
Expected 05h, received 15h.
Test different baud rates...................................Passed
*** END TESTS: 3 ERRORS ENCOUNTERED ***
=== LPT1 PARALLEL PORT TEST ===
Loopback Plug...............................................Not Present
Test parallel data register.................................Passed
Test with external loopback.................................Not Tested
*** END TESTS: NO ERRORS ENCOUNTERED ***
=== VIDEO RAM TEST ===
Test Results:
Video Memory................................................Passed
Video Page Test (4 Pages)...................................Passed
=== VIDEO TEXT TEST ===
Character Set:
Mode 00h (CGA)..............................................Passed
Mode 01h (CGA)..............................................Passed
Mode 02h (CGA)..............................................Passed
Mode 03h (CGA)..............................................Passed
Character Attributes:
Mode 00h (CGA)..............................................Passed
Mode 01h (CGA)..............................................Passed
Mode 02h (CGA)..............................................Passed
Mode 03h (CGA)..............................................Passed
=== VIDEO GRAPHICS TEST ===
Mode 04h (CGA)..............................................Passed
Mode 05h (CGA)..............................................Passed
Mode 06h (CGA)..............................................Passed
*** END TESTS: NO ERRORS ENCOUNTERED ***
=== MAIN SYSTEM BENCHMARK ===
+- CPU SPEED -+ +--- VIDEO SPEED ----+ +- MATH SPEED-+
2.0K+ | 2.0K+ | 50K+ |
1.8K+ Model 60| 1.8K+ | 45K+ |
1.6K+ | 1.6K+ | 40K+ |
1.4K+ AT 339 | 1.4K+ | 35K+ |
1.2K+ | 1.2K+ @@@ | 30K+ |
1.0K+ ___ Model 30| 1.0K+ @@@ | 25K+ |
0.8K+ @@@ Turbo-10| 0.8K+ @@@ | 20K+ ___ |
0.6K+ @@@ | 0.6K+ @@@ | 15K+ @@@ |
0.4K+ @@@ PC, XT | 0.4K+ @@@ | 10K+ @@@ |
0.2K+ @@@ | 0.2K+ @@@ | 5K+ @@@ |
0 +-------------+ 0 +--------------------+ 0 +-------------+
80286 CPU, 0.63 Mhz CGA Video Adapter No Coprocessor
+-+
1023 Dhrystones |/| Determine CPU Speed
+-+
1270 Characters/Second |/| Determine Video Speed
+-+
20.4K Whetstones |/| Determine Math Speed
+-+
=== HARD DISK BENCHMARK ===
+--- SEEK TIME ---+ +- XFER SPEED -+
| | | |
1.0sec+ ___ | 20K+ |
0.9sec+ @@@ | 18K+ |
0.8sec+ @@@ | Drive: 0 (C:) 16K+ |
0.7sec+ @@@ | Disk Type: 0 14K+ ___ |
0.6sec+ @@@ | Cylinders: 160 12K+ @@@ |
0.5sec+ @@@ | Heads: 8 10K+ @@@ |
0.4sec+ @@@ | Sectors/Track: 27 8K+ @@@ |
0.3sec+ @@@ | Total Bytes: 17,694,720 6K+ @@@ |
0.2sec+ @@@ | 4K+ @@@ |
0.1sec+ @@@ | 2K+ @@@ |
0 +-----------------+ 0 +--------------+
Average Track Block Size: 110K
+-+
1389.6ms |/| Determine Average Seek Time
+-+
2.9ms |/| Determine Track to Track Seek Time
+-+
14.0K/Second |/| Determine Transfer Speed
+-+
[Ed. Note: The following was added by Mr. Lowe recently.]
I neglected to mention that the A501 clock is used to store CMOS info for
the MS-DOS side, and I had my clock do really strange things (like reading
the date and getting seconds) three times.
-----------
-Todd Lowe
871579l@aucs.AcadiaU.Ca
--
Jason L. Tibbitts III | Moderator: comp.sys.amiga.reviews
"Blob Shop Programmers:| Send submissions to HONP9@menudo.uh.edu
Because We're Bored!" | Check comp.sys.amiga.reviews for submissions
Disclaimer: Opinions...| guide, disclaimers, etc. Fnord.