A2000 - resets instantly when booting?

[MarkT]

Hi,

I have a slightly strange issue with my A2000. It’s actually a Remake 1.8.5 motherboard but was built using parts from a functional board (just I didn’t trust it to last due to extensive battery leakage). The board was built for me by a reputable person and was tested before being returned to me. I’m now reassembling it and have stumbled across an unexpected problem.

I just have the motherboard connected up to a TV and an SFX power supply (a beQuiet! 300W, so the same model that sees exemplary service in my A4000). I also have an adapter board that can generate a TICK signal. When powering on, the screen flashes for a moment as if it’s about to display the Kick 1.3 “insert floppy” image then it resets straight away. It carries on going around this loop.

I’ve tried reseating chips, changing the 1.3 ROM to a 3.2 (a planned upgrade), changing the TICK source, but all to no avail. Any ideas please?

Mark

[MarkT]

An update…

I removed the CPU and Gary. There was quite a lot of corrosion on the pins (more than I remember there being - perhaps a reaction with the new sockets). Anyway, I cleaned them up but it’s not changed anything. Also cleaned Paula and Denise whilst I was at it but they were already in good condition.

Unfortunately it may be time to turf the Fat Lady out of her bed? I’m pretty new at this and the historic threads I’ve looked at don’t seem to document this exact problem, so again any suggestions welcome. Thanks!

[MarkT]

Time for an overdue update on this I think...

My ATX PSU adapter board apparently wasn't generating a TICK signal that was usable, so I switched over to the internal one and it was fine. But then something else happened and that took a lot more untangling. This is a general summary - in reality I was doing the steps below in little bursts over the period of a month.

Kick 1.3 was working so I swapped in the 3.2 ROM. Instant issues. What started as a "Green Screen" error quickly degenerated into a random Green or Yellow screen on power-up, sometimes starting as one then changing over to the other. Some internet wisdom translated that to me as a Chip RAM and/or CPU problem.

As the problem had begun with the new ROM, I put the 1.3 back in. No change. I then removed all the custom ICs and CPU and cleaned their pins with a selection of fibreglass pen, washing with IPA and sometimes using White Vinegar where there was more serious corrosion. Everything was thoroughly rinsed and dried before I tried to power up again - no change after that either.

I then embarked on a selection of things in order of convenience. I cleaned and re-seated the RAM chips, soaked the likes of Agnus' pins in Contact Cleaner, and tried to differentiate if any chips were hot/cold to the touch. Nothing obviously wrong, and still dead.

Ordered a DiagROM for the first time ever. Nothing on the screen at all. Bought a Null Modem cable and hooked that up to a Laptop with PuTTY. No output there either until I spent an age scratching my head and cleaning/re-seating things. Finally got comms going and then the Amiga display started working. Every diagnostic was suddenly passing okay, so in the excitement I put a Kickstart back in. Now getting a black screen, with an occasional green screen, at every attempt to start.

Was I being an idiot? I'm quite new at diagnosing Amiga issues so it wouldn't have surprised me.

Getting more desperate, I then started swapping out Custom Chips from a rather mangled A500 I'd bought a few years ago (not working either but individual components could be okay). Plus I ordered new RAM chips from eBay. In the end, and one of the last tries as I hate doing it, I dragged Agnus 8372A from her PLCC socket and replaced her with the one from the A500. SUCCESS AT LAST!!! [Despite both chips now having battle scars from one of those PLCC chip pulling tools]

So what's the successful part of the story so far? Booting to Kickstart with an A2630 board I picked up (68000 is still in its socket). I've actually gone for a 3.1 ROM as I'm aiming for a more period-specific feel to this machine but still want decent capabilities. I've also plugged in an A2088XT bridgeboard. Both accelerator and bridgeboard are showing as working in the Early Startup Menu.

I'll start a new reply for the latest problems

[MarkT]

I've cleaned the 3.5" and 5.25" floppy drives for the Amiga and Bridgeboard (a Chinon FZ-354 and FZ-502 respectively). I can't test the 502 as I've not got that far yet, but attempts to boot from the Amiga drive makes unpleasant noises and generates read errors. I've followed the method used by GadgetUK164 on YouTube, which has been fine with A600, A1200 and A4000 drives I've serviced over the past couple of years. Any suggestions please? Could it be that capacitors need replacing on the drive?

As well as the bridgeboard, the machine came with a SCSI controller when I bought it. It's a GVP Impact A2000-2B (1MB x 1) Rev 1. Unlike later GVP boards these seem to be somewhat rare and documentation isn't great. Anyway, as soon as the board is in the machine won't boot: I only get a black screen. It has what appear to be v1.0 ROMS ("ODD 69D9" and "EVEN 1A3B") on it. Does anyone know if these are compatible with Kickstart 3.1? Does anyone have any experience of faults on these boards?

[MarkT]

Further update: changed out the mechanism for the FZ-354 with one from an A500 drive, transplanting the LED and adding a 2K2 resistor at R13 as per this thread...

https://eab.abime.net/showthread.php?t=99736

Now have a working drive as DF0, with a Gotek as DF1.

Added an A2630 board with 4MB Fast RAM. Working fine and gives an era-authentic speed boost.

Gave up on the Impact A2000-2B and went fishing for something else. Now have a GVP A2000-HC Series II Rev 4 board, but now getting issues with that!

[MigaTech]

Can you cover the problem's you are experiencing with the HC Series II Rev 4.

Incidentally, in regards to your fault diagnostics, you are conducting it correctly. Sometimes working the problem yourself helps to understand better. You also find answers to faults you didn't know.

Your thread so far has offered some interesting reading. Keep up the good work and information.

Check the following and confirm if this is your board.

https://amiga.resource.cx/exp/impact2000hc8

These board's are OK apart from possible HDD size limitations. There was something about configuration where it had to be correct to function. As stated 1.3 ROM or higher to autoboot.

[MarkT]

Your thread so far has offered some interesting reading. Keep up the good work and information.

Thanks! I sometimes wonder when I write these things if anyone actually reads them, let alone finds them interesting

My board is one of these: https://amiga.resource.cx/exp/impact2000hcmk2

I believe the only difference between it and the type you linked to is the provision of Fast RAM slots. My board has an AutoBoot ROM labelled "FAAASTROM VEE.3.12/3.2A CE42 GVP (C)1991". It is detected as "Working" in the Early Startup Menu of the 3.1 ROM. That is, unless I connect my ZuluSCSI to it: then the machine never boots!

I have checked the ZuluSCSI and it seems to be configured correctly (particularly the Termination DIP switch). The log file the Zulu generates on power-up appears to confirm it detects the .hda image file I created using WinUAE on a 16GB SD card. I have set up an OS3.1 installation on that .hda image from within WinUAE.

I've been stuck for time recently but I have two new avenues of attack on this problem. The first is to upgrade the GVP ROM to version 4.15 (the hypothesis being that the ROM I have is somehow partly corrupted or incompatible with OS3.1). The second will be to try and set up the image file with GVP's FastPrep (later ExpertPrep???) rather than HDToolbox. The latter option is going to interesting as of course right now I can't even boot the A2000 with an SD card in the Zulu! Need more time to think

[MigaTech]

The ROM upgrade should help in many ways. Certainly transfer speeds etc. Have you managed to connect a CDROM to one of these yet?

Is your current ROM 3.17? Don't think I have seen that board from your link before.

My 2000 has a A2091 and GVP 030 fully populated with 3.1 ROM. It seems quite content running OS3.5 on 4GB SSCF converted to 40 Pin IDE for CDROM link. Still to install the ECS flicker fixer. Was also considering on bumping the DF1 in favour of a Gotek.

Didn't seem to have many issues if any when it came to drive limitations. The ROMs used were 6.6 at the installation, so partitions are all under 1GB. ROM 7.0 corrects this if I am not mistaken.

Check this link on other device ROM upgrades:

https://bigbookofamigahardware.com/b...t.aspx?id=1765

[MarkT]

I have some happy news: the ROM upgrade to version 4.15 on the GVP board (A2000-HC Series II) did the trick! Partitioning is very slightly more convoluted than I've encountered on the A1200 etc. but these were my steps...

1) Created "HD0.hda" file using WinUAE on the SD card. Image size is 1GB.
2) Took it to the A2000 and used GVP ExpertPrep to create 100MB and 900MB DH0 and DH1 partitions.
3) For convenience, took the SD Card back to my PC and used WinUAE to install Workbench 3.1.
4) Returned once more to the A2000 and found, with some relief, that she finally sees and boots from the ZuluSCSI!

If you set up images using HDToolbox then the GVP card won't recognise them: I'm not sure why, but there's some vague memory I have of there being a difference in the RDB created on the drive. I'm sure someone can correct me.

To make life easier in future I have mounted the ZuluSCSI on a 3D printed bracket that makes the SD Card accessible via a Zorro Slot backplate.

[MarkT]

Next small conundrum: misbehaving Floppy Drive and Gotek. Now I have Workbench up and running I can see that, even with no disk in the drive, DF0 (the mechanical unit, a Chinon FB-354) is convinced there is one. I see a "DF0:????" icon in Workbench. I'll try and break the full configuration and behaviour down...

CONFIGURATION
1) DF0 is a Chinon FB-354. It is connected at the end of the original floppy cable, at the connector with a twist in the ribbon, and is configured by jumper as DS0. No other jumpers are used, and this is exactly the same as the original drive that I replaced.
2) DF1 is not *actually* a Gotek. Some forced research just now (because it was advertised as a Gotek when I bought it) reveals that it's really an "OpenFlops" board made by Simulant Systems. I've got to say the documentation for this is uniquely awful, but I think that it's configured as DS1. Moving it to "SEL1" and other positions (except "SEL2") just prevents either drive being seen in Workbench. No other jumpers are in place.
3) Motherboard jumper J301 is closed, as it should be for an A2000 with 2 internal floppy drives.

BEHAVIOURS
1) If a floppy is present in DF0 from boot, it reads fine. I can even get Amiga Test Kit up and running from DF0 like this, and read tests for both DF0 and DF1 are passed.
2) If there's no disk in DF0, it will make 2 attempts to read something at boot and check what's in DF1 before booting from the Hard Drive.
3) After that initial activity, DF0 doesn't "click" as it should. I have no software installed that prevents floppy clicks.
4) Disk changes on either drive are not detected in Workbench.
5) Once DF0:???? then can't open anything in Workbench, even if a disk is inserted (something about the icon not being able to be opened).

I'm still researching of course, but does anyone have any ideas?

[MigaTech]

The DF0:???? usually means it cannot detect DF0: properly. What is the condition of your FDD cable? Try changing the cable to see if it solves DF0 issue. The reason for it trying DF0 before the other drives, is boot priority.

RDB is more likely the reason. They supplied a series of drive set up disks for that GVP, which I have here. It has a different way of setting up the drive. Something along the lines of ASCII. Your set up of 1GB should be OK. What file system are you using?

Sometimes it is better to use WB2.0 as it has few disks and overriding is easy in set-up. It is quite fast on set-up and runs quick on the A2000.

[vk3heg]

Originally Posted by MarkT

If you set up images using HDToolbox then the GVP card won't recognise them: I'm not sure why, but there's some vague memory I have of there being a difference in the RDB created on the drive. I'm sure someone can correct me.

Going from memory, the GVP fastprep software does things slightly differently so it's software is required with the GVP products.


I had a Combo Accelerator/memory/HD controller in my A200 back in the day.

[MarkT]

Your set up of 1GB should be OK. What file system are you using?

I'm just using FFS.

Going from memory, the GVP fastprep software does things slightly differently so it's software is required with the GVP products.

I had a Combo Accelerator/memory/HD controller in my A200 back in the day.

Thanks. I'm just recalling this from somewhere too.

Only had an expanded A1200 back in the day, and regretted selling it. After a 20 year gap I've now amassed a small collection of various models and couldn't be happier

[MarkT]

Success! After what felt like an age this morning of fiddling with jumper settings, swapping drives and checking things like cable orientations I decided to swap the floppy cable out. All the work I've been doing in there must've broken it, as everything is now working fine. MigaTech was right all along!

For clarity: -

• Chinon FB-354 is DF0. Its jumpers are set at DS0 and it's and the end of the floppy cable, past the twist.
• The OpenFlops drive is DF1. The 'SEL1' jumper is connected to the '0' side and it's connected at the mid-point of the floppy cable with no twist.
• J301 on the motherboard is closed.

[MarkT]

Well, I thought it would be plain sailing from here but that's clearly too much to ask for. Having now got a machine that boots from a ZuluSCSI image via the GVP A2000 HC Series II, has access to both floppy drives, detects its 68030 and sees 1MB Chip and 4MB Fast RAM, I felt that counted as fully functional.

I installed Monkey Island 1 and 2 onto my 'Work' partition using WinUAE. Back in the Amiga, these worked fine and loaded into each game. Then, I thought I'd try setting up my Bridgeboard (https://amiga.resource.cx/exp/a2088xt) and that's when the problems began.

Installing Janus 2.1 (I downloaded the DMS image from the Amiga Hardware Database page and converted it to ADF) in WinUAE understandably failed as it couldn't detect the board. "No problem", I thought: I'll just read the ADF using my Gotek on the real Amiga.

The installation started okay, but as soon as it gets past copying a few files it crashed spectacularly. After multiple attempts at doing this it sometimes gurus, but mostly just locks up for a bit before blanking the screen and just sitting there. This generally trashes the HDA image on the SD card and prevents the machine from booting (weirdly even when trying to boot off a floppy image - as soon as it reads the drive it throws a tantrum) and gets a Guru.

To recover, I have to restore an image file from an earlier version or just blank everything using FastPrep and start over. To say this is infuriating is an understatement!

I'm going to try and see if this happens when writing files to the hard drive in other installs, or if it's just a Janus thing.

[Jope]

Do you have enforcer running? The bridgeboards need that if you have an 030 system.

[MarkT]

Do you have enforcer running? The bridgeboards need that if you have an 030 system.

Not yet, as Janus wasn't installed at the time. Installer literally copies a few files and then everything falls over; although I noted it seemed to do so at slightly different points.

I went to try and install Monkey Island [yet again!] after that. I didn't even get that far as when I tried to make a new "Games" drawer on DH1 the same lockup happened! Some frantic Googling has revealed this stonking discovery from comp.sys.amiga.hardware all the way back in April 1993...

https://groups.google.com/g/comp.sys.../c/-acTIBqR_Nk

By amazing coincidence, you'll see in the last message from June this year that someone has been having similar problems. It all looks like an interaction between the A2630 and a DMA SCSI controller (a GVP one in my case rather than an A2091). I'm going to try and replicate some of the behaviours described to see if I need to buy some ICs!

[MarkT]

For posterity, I've copied several of the original posts below. I wouldn't want gems like this to get lost, especially 30 years after they were written!

Klaus Weber
unread,
Apr 23, 1993, 2:59:37?PM
to
Dear netters,
For the last two years I used a Nexus PIO-SCSI-Host-Adapter
(Polling, i.e. not even interrupt-driven IO) and had no problems
with it. Finally, I decided to buy a DMA-Host-Adapter (A2091) to
improve overall system performance. After installing the A2091, I
experienced system crashes during Harddisk writes every now and
then. At first, I thought the RDB format of the Nexus H-A was
slightly incompatible with the A2091, so I backed up my Harddisk,
reformatted/-partitioned it with the A2091 and restored all files.
Even during the restore process I experienced 10-15 system crashes
(with about 70-80 Megs of data to be restored, to give you an idea
about the frequency of the crashes). This was strange, because I
had booted from my original 2.04 Workbench disk, and everything
should have worked fine.

To make a long story short, it seems that the crashes-during-write
problem is a hardware problem.

Symptoms:
- System crashes when writing to the Harddisk, if the Host-Adapter
uses DMA. System works fine if the H-A uses PIO.
- The A2091 works fine in another A2000 with an Rev 9 A2630, so I
would say the A2091 is O.K.
- The System works fine (and slow :-<) if the A2630 is removed.
- The System works fine if the A2630 is present, but the A2630 RAM
is disabled (by setting J303).
- The System _seems_ to work fine if the Rev 6 A2630 is replaced by
a Rev 9 A2630 (I could not test that as thoroughly as I wanted
to, so I would say I'm only 90% sure, but maybe I was just lucky).

- The system _seems_ to work if
* The Rev 6 A2630 and the A2091 are the only cards in the system.
_and_
* The A2091 is put in the first Zorro-II slot.

The system crashes again if
* Other Zorro-II cards are present.
and/or
* The A2091 is put in another slot (i.e. more to the left). The
higher the slotnumber, the more crashes I get (it's hard to
measure though :-<)

From what I have tried, I would say this is a timing or noise problem
with DMA to the Rev 6 A2630 RAM, but then, I'm not a technician.


Configuration:

A2000:
- Rev 6.2 Motherboard
- V37.175 ROMs
- J900 is set.

A2091:
- PCB Rev 4.0
- V6.6 ROMs
- Serial #: 0100892312769001111833
- Memory jumper set to 0 Meg.
- JP 5 (Options): all jumpers removed.
- Harddisk is correctly terminated.

A2630:
- Rev 6 Board
- Rev 6 ROMs (2.0 compatible, both MouseButtons for Bootmenu, AMIX
Option)
- populated with 4 Meg RAM
- Backside says: EXCEL M 94VO 21-89 [next to the RAM]
3 Stickers: * RWK 8
* RWK 7
* RWK 4 (?, not sure)
- Frontside:
* Sticker which says "West Chester build 0611" (what does that
mean, anyway ?) [next to U702]

* "ZD 6" (?) stamped on it. [next to U704]
- J200 is moved in direction of the RAM (away from the CPU).
I think this is what the A2630 manual calls "on pins 2-3",
although I tried both settings (did not cure the problems).


Other cards (the system crashes even if they are not installed, so
this should not matter):

Nexus: Rev 1.1 board (8 Meg option).

A2060 ARCNet Card.
---


If there is an easy solution for my problem (throwing my A2630 out
and buying a new one is not my preferred solution), or even just an
explanation what is going on here, I would appreciate to hear it.

Please send your answer (also) via email, to make sure I get it.

Thanks in advance,
Klaus

Klaus Weber
unread,
Apr 26, 1993, 10:36:15?PM
to
Hello again,
In <1993Apr23....@arbi.Informatik.Uni-Oldenburg.DE> I wrote:
[I'm quoting that stuff, because i made some additions and clari-
fications. Sorry to waste some bandwidth...]

!!! I'm not quite so sure anymore that the A2630 is to blame, see
!!! notes below.

>
>
> Configuration:
>
> A2000:
> - Rev 6.2 Motherboard
> - V37.175 ROMs
> - J900 is set.
>
> A2091:
> - PCB Rev 4.0
> - V6.6 ROMs
> - Serial #: 0100892312769001111833
> - Memory jumper set to 0 Meg.
> - JP 5 (Options): all jumpers removed.
> - Harddisk is correctly terminated.
>
> A2630:
> - Rev 6 Board
> - Rev 6 ROMs (2.0 compatible, both MouseButtons for Bootmenu, AMIX
> Option)
> - populated with 4 Meg RAM

- Board is _not_ overclocked. :->


> - Backside says: EXCEL M 94VO 21-89 [next to the RAM]
> 3 Stickers: * RWK 8
> * RWK 7
> * RWK 4 (?, not sure)
> - Frontside:
> * Sticker which says "West Chester build 0611" (what does that

> mean, anyways ?) [next to U702]


>
> * "ZD 6" (?) stamped on it. [next to U704]

> - J200 is moved towards the RAM (away from the CPU).


> I think this is what the A2630 manual calls "on pins 2-3",
> although I tried both settings (did not cure the problems).
>
>
> Other cards (the system crashes even if they are not installed, so
> this should not matter):
>
> Nexus: Rev 1.1 board (8 Meg option).
>
> A2060 ARCNet Card.

- Rev 2 Board.
> ---
>

Barry Tigner <tig...@paun01.pa.msu.edu> sent me file called
"ATechTips", which contains several do-it-yourself repair
instructions. I did the following modifications to my mother-
board (they were described in that file):

First step:
- "Replace U602 and U605 (74LS245) with 74ALS245."
-> ** NOTE ** My 6.2 board had the ALS types already installed.
see below...

- "Make sure that 74XX24(4|5)'s between Agnus & Ram are F types
the symptoms include spots appearing on screen"
-> they are F types.

- "put 3300 ohm resistor between #20 & #11 of U605"
-> did that.

As I said, U602 & U605 already were 74ALS245's. I replaced them
anyway (soldering sockets in of course) and added a 3K3 ohm
resistor between #20 & #11 of U605. I then tried if the problem
would appear again. Much to my surprise (and dislike, I might
add :->), the problem was much worse than before! With the old
chips installed (see below for what I mean with "old" and "new"
cheips) the problem showed up randomly and was not easily repeat-
able. With the new chips, the computer will crash within the first
two seconds after starting the write operation. Further investi-
gation revealed that U605 is the critical chip:

U602 U605 crash easily repeatable ?
old old no
old new yes
new old no
new new yes

I exchanged the chips several times to make sure that the change
of behaviour is really dependant of the type of chip installed
as U605.


Second step:
After taking another close look at the "ATechTips" file, I found
some more modifications mentioned later in the file. So I made
these to see if they would cure the problem:

- "For the A2000-A2500, put four 3.3K ohm pull up resistors on U605
(74ALS245)
pin11 to pin20; pin12 to pin20; pin13 to pin20; pin14 to pin20"
-> did that, too.

This did not seem to change the behaviour described above (Computer
crashes fast if new chips installed, etc).

SUMMARY OF MODIFICATIONS:
- U602 and U605 socketed. Tried several combinations of 74ALS245's.
- added 3.3k ohm pull-up resistors to pin 11, pin 12, pin 13 and
pin 14 of U605 (using Pin 20 as Vcc source)

SPECIFICATIONS of the 74ALS245's used:
"old" (as installed by CBM):
MALAYSIA 9001DF
* SN74ALS245AN [* is a strange company logo :->]

"new" (the chips I bought today):
_
| / 74ALS245 A - 1N [that's a large S between the
| \ FHK6883 vertikal bar and the text :->]
| _/ 9035VB

Perhaps someone knows if there is a difference between those two
types. I know they _should_ be identical, but since they behave
differently.


As a change on the motherboard clearly make a difference in behaviour
I'm not sure anymore if the A2630 is to blame.

In <1993Apr23....@cgc.com> Mike Armstrong <armst...@cgc.com>
wrote:


> Quite odd, I have the identical configuration, minus the Nexus, 2060 and my
> drive is teh quantum lp240s. I have had no problems at all. All the rom
> levels are identical also. Have you tried placing your level 6 roms in the
> other 2630(the one with 9) and tried it?

Not yet. I will try that as soon as I can though.

> It might be the memory in your card
> that is causing the failure. Since the transfer is DMA, it might just be
> hitting a bad location. (Like I know anything about it...)

Well, IMHO that's unlikely, because the system is working stable
without a DMA-Hostadapter. If the memory would be faulty, I would
expect errors even if the CPU is accessing a bad location.

>
> Mike Armstrong
> armst...@cgc.com

Thanks for your suggestions.

This is again quoted from my previous posting:


>
> If there is an easy solution for my problem (throwing my A2630 out
> and buying a new one is not my preferred solution), or even just an
> explanation what is going on here, I would appreciate to hear it.
>

> Please send your answer (also) via email, to make sure I get it.
>

Thanks in advance,
Klaus

Klaus Weber
unread,
Apr 29, 1993, 8:09:26?PM
to
Hello net,
it seems that I have solved my problems with my A2091 and rev 6 A2630. :-]

SUMMARY OF MY PROBLEM:
- A2000, Rev 6.2 motherboard, V37.175 ROMs
- Rev 6 A2630 W/ 4 MB and V6 ROMs
- Rev 4 A2091 W/ 0 MB and V6.6 ROMs

I experienced random system crashes when _writing_ to the harddisk.
Reading worked fine. The system works fine, if a non-DMA-host-adapter
(e.g. PIO) is used. This is _not_ a lockup of the SCSI-Bus, but a "real"
system crash.

SOLUTION (see disclaimer below)
I replaced U605 (74ALS245, on A2000 motherboard) with an 74HCT245. The
chip I used is labeled:

TOSHIBA 9131H
74HCT245AP JAPAN

U605 is located left of the A2000 coprocessor slot. I _strongly_
recommend the use of an IC-socket.

The computer has written continuously to the harddisk for several (>=6)
hours now (and, in fact, is still writing while I type this :->). I have
not experienced a single system crash since I swapped the chips, to
I think the problem is solved.

DISCLAIMER
This modification has worked for me. It may or may not work for you.
In fact, since U605 is specified as an 74ALS245 part, there is a good
chance that it won't work (you have been warned!). I disclaim all
responsibility for loss of data or damage done to any parts of your
equipment, as a direct or indirect result of the modifications described
above. You make these modification entirely on your own risk.


If you have questions about the problem or its solution, feel free to
email me.

Thanks to all who responded.

Regards,
Klaus

Michael Mickatroid
unread,
Jun 9, 2023, 6:46:14?AM
to

Dear Klaus

I just soldered together the A2000 remake v6.2 I then added a R4.0 A2091. Alles gut. I then added a Rev 6 2630 and was greeted with what passes for Guru Meditation these days. HDD access was so garbled I never made it to workbench (I am using a Zulu scsi HDD replacement so replacing the scrambled mess on DH0 was easy.)

I too became suspicious of an interaction between the cards and searched for them by model number plus 'crash' to immediately find your post. I could not find the HCT chips you used easily and used HC. I believe they are slightly faster but that both are significantly slower than the original ALS part.

Replacing U605 did nothing but replacing the second ALS245 right next to it as well fixed my problems too. Hopefully it has not messed up the bus for other cards. Looking at the bandwidth of the HC though I cannot imagine how.

Thanks so much for sharing the details 30 years ago.

M

I think this illustrates the value to the whole community of documenting our hobby!

[MarkT]

Right, final update for today as after four rounds I'm tired

I've moved the GVP SCSI to Slot 1 (I think; the one nearest the 68000) and left the Bridgeboard in. A fresh WB3.1 install booted as expected and then I've successfully installed Monkey Island 1/2 and Janus 2.1 from the Gotek! In the excitement I've ordered some 74HCT245s too!

The Bridgeboard seems to work and I can fire up the software to get as far as the BIOS displaying.

Getting carried away with this victory, I fired up Monkey Island to find the sound apparently doesn't work. A quick blind guess at using NoFastMem didn't help, but I know Paula is fine as AmigaTestKit will bang out all 4 channels no problem.

I've got to say that this is up there with the most difficult bits of kit I've ever repaired...

[hooverphonique]

Regarding 68030 and Zorro DMA, there is one or more threads on EAB which discusses problems with this due to a cache-related bug in the 030. Make sure it's not something which "just" needs a software fix that is giving you grief