Files
-----
RemAPollo package consists of the following files:
- 68060.pch - patch file for 68060.library, version 60.02 to version 60.10

5.Ram Disk:> list libs:68060.library

Directory "libs:" on Monday 21-Oct-13
68060.library 64632 ----rwed Today 21:07:00
: patched by MM for RemAPollo
1 file - 128 blocks used

5.Ram Disk:> version libs:68060.library

68060.library 60.10

workbench/c/setpatch

Patch for mathieeesingbas.library Mul/Div FPU handling.
Patch for the OS copyright.
Patch for WritePixelLine8() trashing the source.
Froze 43.4.
Fixed V39 ScreenQuiet patch.
Froze 43.5.
Added BOOL CheckIO() patch to help WB and others
Froze 43.6.

Make a copy of your old SetPatch before testing the V43 SetPatch.
The V43 SetPatch should work well with any OS >= V37. Please send
in a short report if it WORKS WELL, too!

The CheckSetPatch.c code fragment may help some software developers
to internally turn off workarounds for OS bugs conditionally.
***NOTE***: Prior to the December 1997 release of this archive, the
CheckSetPatch.c function had a bug in the version check.

Heinz Wrobel

Important

You must remove all other patches (e.g. "FASTIEEE", "ffppatch", "msbpatch", math-patch from "MCP") of mathlibraries.

The patches from "SetPatch", "DraCoSetPatch", "CyberPatcher", "OxyronPatcher", "DraCo060Emu" and "68060.library" are OK.
Information about "mathffp-Patch" and "mathieeesingbas-Patch" - if item 3.1.0 ("mathffp-Patch") and/or item 4.1.0
("mathieeesingbas-Patch") was chosen during installation:

It is not so important where the patches ("mathffp-Patch", "mathieeesingbas-Patch") are located. They may be called from any place
in "user-startup", "draco-startup" or the "startup-sequence", after "SetPatch" (even after "LoadWB"). During the installation
of "HSMathLibs" it is checked whether "draco-startup" is present. If so, the patches are inserted into "draco-startup". If
"draco-startup" does not exist, the patches are inserted into "user-startup".

9. What is neccessary to start DirOpus when booting without Startup-Sequence?

If your Computer is equipped with a MC 68060 Processor, you should first start setpatch (simply
type SETPATCH in the CLI), otherwise problems may occur (see next question).

10. Why is there no version of DirOpus compiled for a 68060?

Compiling the complete Directory Opus Package takes quite much time and there would be no speed
gain for 68060 users. The benefits simply wouldn't justify the efforts. The only drawback is
that SetPatch has to be started before DirOpus on a 68060 equipped Amiga.

SetPatch97 By Gareth Murfin
---------------------------

Here it is! A new SetPatch for your Workbench. Its got all the features of the
most recent one and NO more ;)) But this SetPatch is an Unofficial GateWay 2000
one! For instance when you goto 'About' on your Workbench it will say 'Gateway
2000 AmigaOS 1985-1997', it will also bring up a similar message in the title
bar when your WB first loads. So basically copy it into your C: dir and then
reboot. You wont look back ;)) And since its the latest version of SetPatch
anyway it uses all the new things coded into the last one such as Memory Pooling
and lots of 040 enchanced routines (which really have been helpful to my 040!).

Wont it be nice to have a recent date on your Workbench and a new and excellent
company name? Yes I knew it, thats what made me decide to make this new version
of setpatch ;) (its actual version is 43.7)

Release 42.3:
mmu.library: Added another new command "ClearMMU" which can be used to
re-run the MMU table setup procedure for some computer resources, mainly to
ignore a previously loaded MMU setup at least partially if it is
sub-optimal. This seems to be required for some Apollo boards.

Q:      I own an Apollo 68040 or 68060 board and installed your
library, followed the above question but the system is still slower than
usual.


A:      The Apollo boot ROMs seem to provide an option to remap the
Kickstart ROM to RAM, using the MMU. The MMU setup used for this procedure
will, by the very design of the mmu.library, remain used even after the
library has been loaded, even though this setup is sub-optimal.

To avoid these problems, please add the following line to the
"ENVARC:MMU-Configuration" file:


ClearMMU Memory Blank Motherboard Expansion


This will re-initialize and override the sub-optimal settings of the Apollo
boot ROM.

____________________________________________________________________________


Q:      I own an Apollo 68040 or 68060 board and installed your
library, but some strange software seems now to crash with a "Bus Error",
the Guru number 80000002.


A:      Please use the same procedure described in the FAQ entry above.

ClearMMU        MEMORY/S,MOTHERBOARD/S,EXPANSION/S,ROM/S,BLANK/S,ALL/S


Resets or overrides a previously made MMU setup completely or partially.
This is helpful in case the MMU library has to be loaded on top of an
already active MMU setup which is sub-optimal, and adjustment of this
configuration by hand would be too painful. This command will therefore
ignore parts of the MMU setup which was found active before and will re-run
parts of the automatic setup procedure of the mmu.library.


MEMORY/S        Reconfigure MMU setup for the memory in the system. This
                setup marks FAST memory as "copyback" and CHIP memory as
                "cacheinhibit imprecise nonserial".

                NOTE: Do not try this on certain P5 boards as
                some on-board hardware will expect that parts of the
                memory are not in "copyback" mode.


MOTHERBOARD/S   Reconfigure the address space of all motherboard resources,
                as for example the CIAs, the custom chips, the IDE and
                SCSI ports, the PCMCIA slot etc...
                This should hardly be necessary.


EXPANSION/S     Reconfigure the address space of all expansion boards. Their
                caching mode is set to "cacheinhibit" completely.


ROM/S           Reconfigure the Kickstart ROM space and the "Debug ROM"
                area. The ROM area is set to "ROM", the "Debug ROM" or
                "F-Space" area is set to "cacheinhibit" since some boards
                provide non-autoconfiguring resources here.

                NOTE: Do not use this setting in case your board
                depends on a MMU-remapped ROM for booting, i.e. some
                Apollo boards.

BLANK/S         Reconfigure all blank areas to "blank". Some boards as for
                example the Apollo might leave these blank areas as
                "invalid" which means that faulty software will cause a
                crash immediately. This can be worked around by re-
                configuring unused areas of the memory map by this option.

ALL/S           Ignore the previous setup completely and re-run the
                automatic configuration from start.

                NOTE: This will cause heavy problems on boards
                that require a custom setup, as for example some P5
                boards.

AMIGA ROM Operating System and Libraries
Copyright © 1985-1997 Amiga International, Inc. 
All Rights Reserved.

SetPatch (V43.6) has been installed.  Patch list:
     68040 Support Code Loaded
     BOOL CheckIO() workaround
     Enabled Advanced Graphics Modes
     AmigaOS Copyright
     HWG IEEE Single Precision Mul/Div/FPU Fix
     Graphics WritePixelLine8
     graphics/WriteChunkyPixel() patched
     Memory Pools
     Enabled data cache

Count Process Name       Action     Target Name                 Options Res.
----- ------------       ------     -----------                 ------- ----
1     ramlib             Load       LIBS:68040.library                  OK  
2     ramlib             Load       LIBS:68060.library                  OK

AMIGA ROM Operating System and Libraries
Copyright © 1985-1997 Amiga International, Inc. 
All Rights Reserved.

SetPatch (V43.7) has been installed.  Patch list:
     BOOL CheckIO() workaround
     Enabled Advanced Graphics Modes
     AmigaOS Copyright
     HWG IEEE Single Precision Mul/Div/FPU Fix
     Graphics WritePixelLine8
     graphics/WriteChunkyPixel() patched
     Memory Pools
     Enabled data cache

Count Process Name       Action     Target Name                 Options Res.
----- ------------       ------     -----------                 ------- ----
1     ramlib             Load       LIBS:680x0.library                  Fail
2     ramlib             Load       680x0.library                       Fail

For the 68060 processor to operate properly the following files must be installed:

• the 68060.library in the LIBS:-directory of your boot partition
• the new 68040.library in the LIBS:-directory of your boot partition
• the 68o4o.library in the LIBS:-directory of your boot partition (not mandatory)
• the setpatch command in you startup-sequence
• the CPU60 command in the C:-directory of your boot partition

Please use the 060-lnstall tool of the Apollo-68060 disk for installation.

This program is no use for any normal user which does not
know about the 060 and all the modes and if he would have
the 68060 manual it's still no use for him because the
68060.library activates the optimal settings.
In general...DON'T touch this thing.

If anybody thinks his system should still work with
the command -nf after the 68060.library was loaded
he should read the upper sentence again.
If you disable the fpu you should know what to expect
from a system that needs a fpu.

This tool is a direct 060 interface...it doesn't care for
the OS and is dedicated to people who know what the operations
exactly do.

You were warned.


USE c:cpu...it allows you to disable/enable the caches and that's
all you need. NO..c:cpu doesn't speed up the 68060.library either.

o NOSTOREBYPASS
  Some old 060 bugs can be "fixed" by using this arg.

6.3 The Jumper CLK
The 68040 processor has to be supplied with the double external clock frequency.
A 68040RC40 (40MHz) CPU is operated with an 80mhz crystal oscillator.
The 68060 processor (with 50MHz) is operated with ist exernal bus frequency.
The CLK jumper is used to adjust the corresponding clock distribution as following:

Jumper CLK (settings):
68040-CPU Installed:            position '040
68060-CPU (50Mhz) Installed:    position '060
68060-CPU (66Mhz) Installed:    position '040

• Jos CLKEN on GND asennossa, on väylänopeus 1/1. CLK jummperin asennolla ei merkitystä.
• Jos CLKEN on CLK asennossa, väylänopeus valitaan CLK jumpperilla.
  • '060 asennossa väylänopeus on 1/1.
  • '040 asennossa väylänopeus on 1/2.

Here are the parts I used, I got them all at Radio Shack:
Diode - 1N914 Radio Shack 10 pack US$1.19 part# 276-1122
Battery Clip - Radio Shack US$1.99 part# 270-430
Battery - 3.6V coin/watch battery, clip can take either CR2025 or CR2032 type.

LithBatt.gif (530x362) 3.34kt

With the CPU60 command the 68060 processor can be configured:

C	(CACHE)	Turns all Caches on (INST, DATA, BRANCH)
NC	(NOCACHE)	Turns all Caches off (INST, DATA, BRANCH)
IC	(INSTCACHE)	Turns the Instruction Cache on
NIC	(NOINSTCACHE)	Turns the Instruction Cache off
DC	(DATACACHE)	Turns the Data Cache on
NDC	(NODATACACHE)	Turns the Data Cache off
BC	(BRANCHCACHE)	Turns the Branch Cache on
NBC	(NOBRANCHCACHE)	Turns the Branch Cache off
SB	(STOREBUFFER)	Turns the Storebuffers on
NSB	(NOSTOREBUFFER)	Turns the Storebuffers off
A	(ALLOCATE)	Normal Cache function (INST, DATA)
NA	(NOALLOCATE)	Caches are freezed
IA	(INSTALLOCATE)	Normal function of the INST Cache
NIA	(NOINSTALLOCATE)	INST Cache is freezed
DA	(DATAALLOCATE)	Normal function of the Data Cache
NDA	(NODATAALLOCATE)	Data Cache is freezed
S	(SUPERSCALAR)	CPU in superscalar mode (two Integer Units parallel)
NS	(NOSUPERSCALAR)	second Integer Unit deactivated
F	(FASTROM)	FASTROM option activated (uses 512KB FastRAM)
NF	(NOFASTROM)	FASTROM option deactivated
B	(BEST)	best possible configuration
w	(WORST)	worst possible configuration

Cpu060.doc:

o NOSTOREBYPASS
  Some old 060 bugs can be "fixed" by using this arg.

PG wrote:
The fact is that most Apollo:s are not stable at 50Mhz FSB, thats why for example eyetech always used the 040 position for the FSB (FSB =1/2x cpu clock), regardless what cpu they used, 040 or 060.

http://www.amiga.org/forums/showpost.php?p=257508&postcount=10

Note that for all '040 & '060 accelerators supplied by Eyetech the 'CLK' jumper should be
set on the '040 position - even for the '060/50MHz board. Ignore any instructions to the con-
trary which are given in the accelerator-specific instruction sheet.

Damion wrote:
Looking at the pinouts, CLKEN doesn't exist at all on 68040. (Also, 68040 BCLK is missing on 68060.) CLKEN seems to be the signal on 68060 that enables/modifies the divider - full, half, or quarter speed operation.

http://eab.abime.net/showpost.php?p=611189&postcount=19

rkauer wrote:
The SMD 74F74 (the TTL chip near the battery, marked as "F74") got some battery leakage. You must have someone near you to replace it.

Don't use any other type of TTL series chip (LS, L, S and even ALS will not work), just the F series work in this board.

http://www.amibay.com/showpost.php?p=339575&postcount=2

Apollo ROM versions:

- V3.55: Boots now quickly disposed of at the A1200 (error)
- V3.56: Runs now with SQ3105-AT-removable disk
- V3.59: RAM installed involvement!
- V3.60: runs with a first CD-Rom
- V4.00: Standardization of the version numbers 1994 - expanding RAM
- V4.01: Fixed bug
- V4.02: RAM also runs out at HD
- V4.03: SCSI reset completely rausgenommen
- V4.04: BUG in wait for the SCSI 2.Controller if AT plate as
- V4.50: switches affect only the boot time
; Caches turned off at the beginning
- V4.51: caches at the end restored (not working)
- V4.52: as above, but functional
- V4.55: RAM-involvement improved ED-RAM support
- V4.56: SCSI and AT decoupled, it no longer have to both devices
; Be on each plate on it, one well enough.
- V4.57: RAM-involvement improved upper 2GB are no longer
, Scan, scanned at 4040 only in the range $ 08 - $ 10
; Assignment of "serial" in Autokonfig has changed to
- V4.58 bug when Speichereinbinden in 4040 resolved
- V4.60 Ram integration moved forward in the Diag-entry routine
- V4.61 Adding support for> 2GB disks
- V4.62 support of the ED gebugfixt RAMS!
- Version V5.00 extradition to XX40 Card
- V5.00a 3.10.95 fixed bug when ED RAM
- V5.00b 6.10.95 Bug fixed in the absence of a device
- V5.00c 8.10.95 Copying 60SP into RAM and start it.
- V5.00d 1.11.95 CacheClearU afternoon copying, so that code in memory
- V5.01 28.11.95 Again ED RAM bug solved.
- V5.02 29.11.95 Fixed bug in the SCSI switch
- V5.10 02.12.95 transformation of 8KB ROMs into a 16KB ROM
- V5.11 02.12.95 A3000 Kick3.1 bug fix
; ****** FROM NOW, THEREFORE NO LONGER SUITABLE FOR OLD 4BIT ROMS *******
- V5.12 04.12.95 A3000 Kick3.1 bug fix Fix (ROM checksum)
- V5.14 09.12.95 A3000 Kick2.0 bug fix
- V5.15 10.12.95 BugFix in A3000 Kick2.04 patch
- V5.21 27.12.95 integration of SCSI devices, if present at EPROM
; 68060.library integration of the system
- V5.22 16.01.96 involvement of the SCSI devices bug fix
; Involvement 68060.library back into the system out
- V5.23-V5.40 test the RAM involvement by MMU, but abandoned
; Because DMA is not so runs, so back to V5.22
- V5.50 to V5.22 REGENERATIVE 01.03.96 !!!!!
; Reloc routine improved to also run with AFS
- V5.51 08.03.96 AKTIVAT-ED rausgenommen provisional restoration,
, Since problems with the 2030 does (from $ 08000000)
- V5.52 10.03.96 AKTIVAT-ED back in, gebugfixt for 2030,
; Reloc routine gebugfixt
- V5.55 08.04.96 Apollo neutralized in the ROM
- V5.60 10.12.96 support of 2/8MB-Rams and all assignments
- V5.61 28.03.97 Special A600 version without SCSI

Jens Schönfeld wrote:
Hi,

I learned about this petition just now, and I have to say that it's way out of line to ask for this stuff for free.

I've paid good money for these rights in order to let the Apollo boards rest in peace. I will not let go of any code or schematics, because these boards make an Amiga unstable. They are the prime cause for Amigans leaving the market: An unstable Amiga is no fun, and hardly anyone will accept that it's the Apollo card causing the trouble. If you're interested, I can dig up one of the many explanations that I have posted online and in private eMails, so you can post that in the amiga.org thread (you know I'm not there any more).

That said, I'm willing to change my policy if serious money is involved. Here's the model:

- I'll change my policy from "if you have an Apollo, it's your choice and if it works, fine", to the new policy: No support for any Apollo-equipped Amiga whatsoever". This includes any and all products, old and new.

- I'll sell MACH131 chips programmed for Apollo 1260 and Eprom chips with the required code, combined for 150,- EUR each, plus shipping. This price does not guarantee anything, just proper function of the MACH131 in a system with no promises of any improvement over the MACH130 that people had before. Should the MACH131 not work as expected, I'll replace it once. Should the replacement still not work, I will refund the chip for 4,- EUR. The remaining 144,- EUR have been paid for the code license in the Eprom.

- once 267 chip-upgrades are sold (40k EUR barrier crossed), I will release the jedec file of the MACH131 of the Apollo 1260 for non-commercial use. Should anyone start offering MACH131 for more than 4,- EUR each, I will go after him. The jedec file of the MACH210 and schematics will remain closed.

ciao,
--
Jens Schönfeld

Jens Schönfeld wrote:
If the cards work fine for you, good for you. However, having insight into the inner workings, I know that the MACH chips are operated way out of their specifications. To be more specific, the design is full of syncronous flipflops with async set/clear equations, which are not specified to work at these high frequencies. As a result, the whole design may or may not work, depending factors that nobody can influence.

And yes, please quote me, but please quote whole eMails, not just single sentences.

so short,
--
Jens Schönfeld


http://www.amiga.org/forums/showthread.php?t=61787
http://moobunny.dreamhosters.com/cgi/mbthread.pl/amiga/expand/217922

Directory "DF1:" on Monday 25-Nov-13
MBRTest.Readme.info          961 ----rwed 06-Jun-89 20:37:43
MBRTest                    40660 ----rwed 01-Jun-89 17:20:06
MBRTest.Readme             14425 ----rwed 11-Apr-89 01:48:12

Directory "DF1:" on Monday 25-Nov-13
MBRTest-2                 110236 ----rwed 26-Jan-78 00:11:58
MBRTest-2.prefs               36 ----rwed 01-Jan-91 10:02:25

Directory "DF1:MBRTest-2" on Monday 25-Nov-13
MBRTest-2.doc.info           617 ----rw-d 01-Jan-91 11:13:29
MBRTest-2.info               522 ----rw-d 01-Jan-91 11:13:30
MBRTest-2.doc              28094 ----rwed 01-Jan-91 11:13:33
RamTests.info                632 ----rw-d 01-Jan-91 11:13:34
MBRTest-2                 110236 ----rwed 01-Jan-91 11:13:44
RamTests                     Dir ----rwed 01-Jan-91 11:13:55

2.5 Memory Tests

These gadgets control which memory tests are done on each selected memory
area. When the specific test is running, its name will be highlighted next
to its button gadget. Here is a general description of each test:

MSCAN -
This test writes a pattern to the entire area being tested, then reads and
checks it. Then it writes an inverted pattern, then reads and checks it. The
MSCAN test checks DRAM memory behaviour and installation verification.

RSCAN -
This writes a pattern to the entire memory area being tested, waits 15
seconds, then reads and checks it. Due to the nature of this test, it can
only be done destructively. The RSCAN test checks DRAM refresh circuits and
DRAM data integrity.

MARCH -
This writes a pattern to the entire memory area being tested, and reads each
location, checks it, inverts it, and moves on to the next location. Then it
repeats the read, check, and invert pass. The MARCH test checks partial
address uniqueness (even address boundaries), crosstalk, and address/data
line noise immunity.

DTADD -
This writes a unique data value to each location in the entire memory area
being tested, then reads each location and checks it. The DTADD test checks
address uniqueness.

HSCAN -
This writes a pattern to the entire memory area being tested, then reads and
checks it. The read pass is repeated as many times as is specified by the
Test Cycles gadget. When this test is selected, all other tests are
automatically deselected. Due to the nature of this test, it can only be
done destructively. The HSCAN test checks DRAM refresh circuits and DRAM
data persistence.

Both the RSCAN and HSCAN tests are marked by asterisks to indicate that they
may only be done as destructive tests. Destructive testing can only be
performed on those memory areas which have not been added to the system
memory list. The other tests may be done either destructively or
non-destructively. Note that "destructive" in this context only refers to
the contents of the memory being tested: you cannot harm or permanently
affect the working of any of your Amiga's memory simply by running MBRTest.

Non-destructive testing is done when a memory area is partially (or fully)
in use by other tasks on the Amiga. This type of testing is done in a single
pass, and is designed not to interfere with those other tasks. Note that
while this type of testing is done, all mounted disk volumes are temporarily
disabled to prevent DMA interference. Also, while doing non-destructive
testing of CHIP memory, you may see 2 or 3 display flashes, a mouse pointer
flash, and the screen's bitplane memory being tested. These temporary
effects will not interfere with the integrity of the system.

Destructive testing is done when a memory area is fully available for the
exclusive use of MBRTest-2 for the duration of the test. This is the case
for memory boards that have not been added to the memory list, and for
memory areas that can be fully allocated by MBRTest-2 when the area is
selected by the user for testing. This type of testing is done in multiple
passes, and usually runs somewhat faster than non-destructive testing,
because the data in the memory being tested need not be saved elsewhere
while testing is done and restored after the test. Destructive testing is
also better at detecting certain types of addressing errors in the memory
hardware.

The type of testing to be done is determined automatically by MBRTest-2, and
all the memory installed in the Amiga may be tested.

Tantaalikondensaattorin toimintaperiaate on likimain sama ( kuin elektrolyyttikondensaattorien), eli anodina toimivan tantaalipalan pintaan tehdään elektrolyyttisesti tantaalioksidikalvo (Ta2O5), joka siis toimii eristeenä. Sen päälle muodostetaan katodi hopeoidusta grafiittikerroksesta. Käyttämättömänä olleen tantaalikondensaattorin oksidikerros ei vaadi korjausta.

http://fi.wikipedia.org/wiki/Kondensaattori

• MACH131-15JC on juotettu suoraan piirilevyyn.
• MACH210-15JC on juotettu suoraan piirilevyyn.
• EPROM versio 5.60 ( uusin on V5.61 "28.03.97 Special A600 version without SCSI" ).
• Tantaalikondensaattori ( saattaa olla alkuperäinen 040:ssa ).
• Akku korvattu Panasonic VL2020 3v litium-akkulla, eli ei patteri/diodi-viritystä.
• Kide liimattu kantaansa ( hyvä, koska omassakaan se ei ole kovin tiukka )
• Itse tehty 3.3V power converter
• Väylänopeus on 1/2, eli 40MHz (Jumperit CLK: 040, CLKEN: CLK)
• 60ns SIMM kammat

Single sided 72pin SIMMs are: 4, 16, 64MB
Doubled sided units: 1, 8, 32 & 128MB

Kuvat Kreikkalaisen mfilosin blogista: http://www.mfilos.com/search/label/a4000

SMD Capacitor pad (Radial Can):

[1], [2], [3]

PCB track:

[1], [2], [3]

PLCC Socket pad:

[1], [2]

Freescale Support wrote:
Dear ShK,

Thank you for your inquiry.

The MC68060RC50 device is still manufactured, however recently (21 Nov 2013) has been announced start of 2-year OEL process (see Product Discontinuance notification http://cache.freescale.com/files/shared/doc/pcn/PCN15939.htm ).

Last Order Date is 21-Nov-2014 and Last Ship Date 14-Nov-2015.
This gives you time for placing orders and build inventory stock, or re-design your product, if necessary.

Should you need to contact us with regard to this message, please see the notes below.

Best Regards,
Technical Support
Freescale Semiconductor

stachu100 wrote:
Tantalum capacitors have several advantages: longer lifetime, lower impedance, higher operation temperature.