2052 problem?
Joost van Vroonhoven
>Well, I bought an old C= 2052 (?) ram card for my A2000 yesterday....
> Here's the problem:
> <A2052 unreliable at bootup, read errors on hard drive disappear after reset>
I see this one too! Every day :-(
I have A2000B rev.4.1 with Kickstart 1.3 and Fat Agnus (1M chip RAM),
68020/68881 at 7.16/12 MHz or 68000 at 7.16 MHz,
A2058/2 memory board, HardFrame SCSI controller with 44MB Rodime and
A2088 BridgeBoard.
A cold boot (power up) has a 99% chance of failing with SOME guru, mostly
guru 4 with task 207aa0 which is the DH0 task.
I then reset the machine (Ctrl-A-A) and try again.
It normally :-( needs two to nine resets before it works OK.
If the machine has been on for five minutes or more NO ERRORS anymore.
>Well, I'm stumped! I noticed 2 jumpers on the 2052 (J1 & J2) but have no docs
>on this card (neither does my local dealer). Does anyone know what
>these jumpers are for?
I have the docs, but at home. E-Mail me and I'll find out.
>Could they help solve my problem?
>Anyone wish to hazard a guess as to
>why I get an error from my WorkBench(Hard drive) the first 2 power ons but not
>the 3rd? Any suggestions (other than to sell the 2052 card?)???
>ch...@zork.cc.binghamton.edu Chris Peck
Well, I'm stumped too!
Anyone wish to hazard a guess as to why I get an error from my
System_HD(first hard drive partition) the first few power ons but not later ?
Any suggestions (other than to sell the 2058 card?)???
I removed the A2058 as a test the Amiga booted fine.
No errors. So I send the card out for repairs.
It came back "repaired", ie. no visible changes and it STILL failed.
So I now think it is some wierd incompatibility between my rev.4.1, my A2058/2
and my (DMA) HardFrame.
No, CPU (68000 or 68020) does NOT matter.
No, swapping the A2058 and HardFrame does not solve it either.
No, the BridgeBoard is not started, I only run BINDDRIVERS if I need the A2088.
HELP!
--
Joost van Vroonhoven, Philips IMS CDI Authoring Support,
building SFH-6, PO.Box 80002, NL 5600 JB EIndhoven, Holland,
phone:+31 40 737184, secretary:+31 40 736299, fax:+31 40 735932
e-mail: jo...@imladris.cdi-as.ce.philips.nl or ...optimg!nlvs08!joost
Chris Peck
the number - it holds a total of 2 Meg RAM only). It's fully populated to 2Meg.
I've had an older GVP SCSI/RAM controller (can hold 0, 1 or 2 Meg RAM - it has
1 Meg on it, the jumpers are set right - I've been using this card for quite
awhile now, with no problems, until now). Also, there's a Quantum 105S connected
to this (has been working fine since I bought it 2-3 weeks ago). Let's see, the
motherboard is Rev 4.3 (so the MB has 512K chip & 512K Fast RAM)
and it's jumpered to NOT get TICK from the power supply (non commodore,
200 watt Power One PS)... Also, I have 1 floppy drive.
Here's the problem:
After I put the 2052 RAM card in my A2000, I powered it on and got a read error on
my hard drive. I turned it off, removed the 2052 and powered on again - it booted
fine. I put the card back in a different slot and powered on, same error on hard
drive, so, I powered off waited a minute & powered on - same error, so I cycled power
one more time and bingo - it booted fine! I ran QBTools & DR.AMI on my hard drive
in search of bad blocks - none were found. I ran DR. AMI on the 2052 RAM card -
no problem! The A2000 see's all 4 Meg RAM. SO I figured it was a weird one.
The next day, I powered on the A2000 - error on hard drive. I cycled power again
and the 3rd time it booted up with no errors. Same thing this AM (3rd time a charm).
Well, I'm stumped! I noticed 2 jumpers on the 2052 (J1 & J2) but have no docs
on this card (neither does my local dealer). Does anyone know what these jumpers
are for? Could they help solve my problem? Anyone wish to hazard a guess as to
why I get an error from my WorkBench(Hard drive) the first 2 power ons but not
the 3rd? Any suggestions (other than to sell the 2052 card?)???
-thanks
chris
--
ch...@zork.cc.binghamton.edu Chris Peck
ch...@bingvaxu.cc.binghamton.edu SUNY Binghamton, NY
ch...@bingvaxa.bitnet
--"Any opinions expressed above are mine, ALL MINE!"--
Chris Peck
it's labeled C= 25000 2MB RAM (I don't see 2052 on it anywhere). I also
noticed some type of resistor on the back with a wire going somewhere
(nice and specific, eh? well, I'm a software kind of guy...). I have
taken the card out for now, since I can't get it to boot reliably with
it in... Still wondering about the jumpers (J1 & J2)....
Peter Wemm
get random bit-cell failures AT ANY TIME! My system is an OLD A2000 (with
a 48-pin Agnus), A2090, and a 1 MEG ram card in the CPU slot. It WONT run
a CPU accelerator card. Up until a few weeks ago, it ran with rock solid
reliability, but now, it won't even load the great Anti-Lemmings demo from
floppy's from a hard reset. It needs to have something (anything) running
to change the memory allocation layout. At least the 2090 has never killed
my Hard drive like it has done for some. Also, it seems to be running at
very high temperatires inside the case - some of the chips are almost too
hot to touch! (That has been happening for years, it's not a recent
development.)
I'd like to know what the jumpers J1 and J2 do too.
-Peter
PS: I am looking for some gullible person to buy this machine.. Any offers?
--
Peter Wemm
------------------------------------------------------------------------------
pe...@cs.curtin.edu.au (Home) +61-9-450-5243
Curtin University of Technology, Perth, Western Australia.
Nuke the Simpsons!
Chris Peck
Well - here's some of the responses I got..
-power supply problem - the 2052 is a power hog! (more on this later)
-upgrade the motherboard to Rev 4.5 (no explanation why)
-RFI - try using an insulated shield between 2052 and GVP SCSI/RAM (or shielding the cable).
well, I tried putting a RFI shield in place (and shielding the SCSI drive cable), no help.
I did not upgrade to rev 4.5 (got this suggestion last nite). Wish I knew what going
from 4.3 to 4.5 would do???
However, my dealer (after talking with C=) indicated that the 2052 is a power hog. Since I
have a "homebrew" A2000 (with a Power One 200 watt PS) I figured I tried "shortening" the
wires from the PS (although, I used to have a 80286 BB which had no problems). This worked!
My dealer also said that the 2 jumpers are for "diagnostics"...
It appears to be working just fine now - I made the change Sat nite, but, didn't get to
run it too much Sunday - however - it boots up without any problems....
-thanks for the info folks
-chris
Dave Haynie
>Add me to the list too, I have a sick A2052 as well.. I am lucky enough to
>get random bit-cell failures AT ANY TIME! My system is an OLD A2000 (with
>a 48-pin Agnus), A2090, and a 1 MEG ram card in the CPU slot. It WONT run
>a CPU accelerator card.
A couple things to know about the A2000-a boards. First off, there was a bug
in the bus control PAL of many of these, which causes major problems if you
have DMA to an expansion bus memory card. Technically speaking, the bus buffers
between the motherboard local bus and expansion bus are supposed to be off, or
pointed toward the motherboard, during an on-expansion-bus DMA transaction. On
the original A2000-a, they're actually pointed at the expansion bus, so you get
data bus buffer fights between the DMA device or RAM card and motherboard
buffers on every cycle. They released a PAL that fixes this, which was shipped
with A2090s once they knew about the problem.
As for accelerator cards, the old A2000-a's don't have a full fledged
Coprocessor slot. You can't run any accelerator board without removing the
68000 from the motherboard, and some accelerator cards aren't designed to work
in this system at all. The A26x0 boards do work, though you need to change
three jumpers on the A2620, one jumper on the A2630, and of course remove the
68000.
Those are the problems I know about with that system.
--
Dave Haynie Commodore-Amiga (Amiga 3000) "The Crew That Never Rests"
{uunet|pyramid|rutgers}!cbmvax!daveh PLINK: hazy BIX: hazy
"This is my mistake. Let me make it good." -R.E.M.
Craig Lemon
[..talk about 2058 causing HD read errors deleted...]
>I removed the A2058 as a test the Amiga booted fine.
>No errors. So I send the card out for repairs.
>It came back "repaired", ie. no visible changes and it STILL failed.
>
>So I now think it is some wierd incompatibility between my rev.4.1, my A2058/2
>and my (DMA) HardFrame.
I have :
REV 4.2 B2000 w/ 68010
2x2052/2 = 4MB RAM
Hardframe/2000 v1.5C firmware
A2088 Bridgecard
[...other stuff...]
and have never had any problems like either of you have described.
My 2052s are a little strange though. They have writing in magic marker
all over them and have A25000 printed in the lower corner where the model
number usually goes (I believe that the number). Dave H., did I get some
factory test boards by accident or something?
>
>No, CPU (68000 or 68020) does NOT matter.
>No, swapping the A2058 and HardFrame does not solve it either.
>No, the BridgeBoard is not started, I only run BINDDRIVERS if I need the A2088.
>
>HELP!
>--
>Joost van Vroonhoven, Philips IMS CDI Authoring Support,
>building SFH-6, PO.Box 80002, NL 5600 JB EIndhoven, Holland,
>phone:+31 40 737184, secretary:+31 40 736299, fax:+31 40 735932
>e-mail: jo...@imladris.cdi-as.ce.philips.nl or ...optimg!nlvs08!joost
--
Craig Lemon - Kitchener, Ontario. Amiga B2000 UUCPv1.13D.
cle...@lemsys.UUCP lemsys!cle...@xenitec.on.ca | Please Mail any binaries
xenitec!lemsys!cle...@watmath.waterloo.edu | to 'files' at this site
..!uunet!watmath!xenitec!lemsys!clemon | instead of 'clemon'
Dave Haynie
>My 2052s are a little strange though. They have writing in magic marker
>all over them and have A25000 printed in the lower corner where the model
>number usually goes (I believe that the number). Dave H., did I get some
>factory test boards by accident or something?
The A2052 was actually built before it was named. Since at the time of its
design, the A2000 was being called the A2500, Jeff Boyer (the card's designer)
had A2500, or some-such, written on the board.
Jeff Dickson
>
>[..talk about 2058 causing HD read errors deleted...]
>
>>I removed the A2058 as a test the Amiga booted fine.
>>No errors. So I send the card out for repairs.
>>It came back "repaired", ie. no visible changes and it STILL failed.
>>
>>So I now think it is some wierd incompatibility between my rev.4.1, my A2058/2
>>and my (DMA) HardFrame.
>
> I have :
>REV 4.2 B2000 w/ 68010
>2x2052/2 = 4MB RAM
This original poster never got this right and now further discussion
is practically meaningless. Is it an A2052 or A2058?
>>Joost van Vroonhoven, Philips IMS CDI Authoring Support,
Bernd Felsche
[ major A2000-a hardware problems revealed ]
>Those are the problems I know about with that system.
Gee thanks Dave. How the heck do you expect me to sell this box now?
Give it away?
It would have been much nicer if I'd been told about these problems
before I bought the machine (it cost me almost 2 months' salary). I
would have waited for a real Amiga which I could expand, the original
reason for buying an A2000 instead of an A500.
Strewth!
--
Bernd Felsche, _--_|\ #include <std/disclaimer.h>
Metapro Systems, / sold \ Fax: +61 9 472 3337
328 Albany Highway, \_.--._/ Phone: +61 9 362 9355
Victoria Park, Western Australia v Email: ber...@metapro.DIALix.oz.au
Dave Haynie
>In <22...@cbmvax.commodore.com> da...@cbmvax.commodore.com (Dave Haynie) writes:
>[ major A2000-a hardware problems revealed ]
>>Those are the problems I know about with that system.
>Gee thanks Dave. How the heck do you expect me to sell this box now?
>Give it away?
Well, as they say, be careful what you ask for, you just might get it.
>It would have been much nicer if I'd been told about these problems
>before I bought the machine (it cost me almost 2 months' salary).
I wish they had told me about them before I went and built a custom chip with
the same problems (the expansion bus problem, which I found, not while
designing the Buster chip, but three months later during pre-production on the
A2000-b in Taiwan).
I don't really think of these problems as all that major. The only real killer
is the expansion bus bug. That's easily fixed by replacing one socketed PAL.
And keep in mind most A2000-a's don't have that bug; I found it in June of '87,
they had only been shipping for a short time at that point, and since the fix
was a PAL, they were capable of reworking any units still in stock instantly.
The problem with accelerators depends on which accelerator you're using. As
I mentioned, all the C= accelerators work just fine, you just have to remove
the 68000. I haven't booted in 68000 mode myself except to prove that it works,
and I've been using these thing longer than most anyone (I went 68020 full time in
1987, 20MHz 68030 in 1988). Most of the 3rd party accelerators either fit in
the 68000 socket itself, or have no software driven means to get the 68000 back
anyway, so this isn't something that most folks will miss.
>I would have waited for a real Amiga which I could expand, the original
>reason for buying an A2000 instead of an A500.
The A2000-a really can be expanded, and in most cases you aren't going to run
into troubles with it. It does have some nice characteristics, like the 4
layer board, which makes it much less sensitive to noisy PICs (eg, marginally
designed boards, there are some out there) than an A2000-b.
--
Dave Haynie Commodore-Amiga (Amiga 3000) "The Crew That Never Rests"
{uunet|pyramid|rutgers}!cbmvax!daveh PLINK: hazy BIX: hazy
"I looked for it and I found it" -R.E.M.
Bernd Felsche
>In article <1991Jun29....@metapro.DIALix.oz.au> ber...@metapro.DIALix.oz.au (Bernd Felsche) writes:
>>Gee thanks Dave. How the heck do you expect me to sell this box now?
>>Give it away?
>Well, as they say, be careful what you ask for, you just might get it.
[ some stuff deleted ]
>The A2000-a really can be expanded, and in most cases you aren't going to run
>into troubles with it. It does have some nice characteristics, like the 4
>layer board, which makes it much less sensitive to noisy PICs (eg, marginally
>designed boards, there are some out there) than an A2000-b.
WOW. I can sell it! :-) Just let me tear off this bit and stick it
inside the box when it comes to upgrade.
You wouldn't by any chance be having problems with noisy PICs in your A3000?
BTW: The A3000 is really hoopy. Any idea if/when a hardware tech ref
manual will be produced?
Dave Haynie
>You wouldn't by any chance be having problems with noisy PICs in your A3000?
A3000? No, if anything, a marginal PIC is more likely to work in the A3000
than the A2000. There are a couple big mistakes that will cause a PIC to fail
in the A3000; forgetting to qualify board select and DBOE* with AS* (CCS*) is
The Big One. The A3000's backplane is quieter than that of either A2000,
though. And the bus controller is smart. It snoops for Zorro II PIC activity
after a cycle is supposed to be over with, and if it sees any, it delays the
start of the next cycle.
>BTW: The A3000 is really hoopy. Any idea if/when a hardware tech ref
> manual will be produced?
I really don't know. We've had many of the kinds of things you'd like to see
in a Tech Ref manual in the last few sets of Developer's Conference notes. The
Zorro III Bus Specification, an example Zorro III memory board, and an A3000
System Specification were included in the notes from last year's Atlanta
conference. That's not far from what I'd have in a complete Tech Ref book.