#HELP! My Amiga Died!
36 messages in this thread
Last night I installed a new GVP TBC Plus board in my A4000. When I turned it
back on, the computer didn't boot. I removed the board, and the computer still
doesn't boot. Now what? I live overseas where there are no Amiga dealers or
repair centers and I would be greatful for any troubleshooting advice.
As for specific symptoms, the harddrive light flickers momentarily upon
power-up, and the hard drive spins up to speed. I also note what appears to be
an attempt to read something from the hard drive when it comes up to speed (by
virtue of the sounds), but it doesn't continue into bootup.
I inserted a floppy, but the floppy seems completely dead; the light has yet to
come on and it doesn't spin. I tried the Cnt'l-A-A to reboot, (with and
without floppy disk) but the symptoms are the same.
Questions: What voltages should I see on the drive power harness? What chip
controls the disks? Can someone give me direction?
-Bruce
PS A new motherboard with latest Buster chip was installed by a repair shop
about 2 months ago.
Bruce,
I've seen similar problems with a 4000. It was due to poor connections.
I'd reseat the Zorro slot daughterboard and check the hard drive cables.
Anything that might have been moved. On the hard drive power harness, the
two center pins are ground. The outside two have either +12 or +5v. If the
4000 is like the ECS machines, the floppy disk is controlled via Paula
(for the R/W data) and one of the 8520 CIA chips (for control lines). You
could try disconnecting the HD & see if you get the Insert Workbench
prompt. It sounds like the system is hung up on the HD and never trys to
boot from the floppy…. But I'd try reseating things first.
Jim
MOS via Autopilot & Amiga
Bruce,
I may have a similar problem as you. I installed an OpalVision board in a
2000 and the same thing happened, no boot but the HDD spins up to speed.
I took it to my dealer and he suggested that it might be one or both of
the CIAs which apparently control reboots. I don't know if this is the
same as you but if I get any more info I'll try and pass it your way.
Andrew
>> I installed an OpalVision board in a 2000 and the same thing happened,
no boot but the HDD spins up to speed.<<
PMJI, but the harddisk gets it's power straight from the powersupply,
doesn't it? If this is true, you could blow up your complete mainboard
(without the powersup ofcourse) and still got a working ('spinnig') HD.
I'm not trying to make you all depressed, but what I want to say is that
these two comparisons with your cases don't give it the same reason. Good
luck in finding the solution!
I guess suggesting a blown up CIA would be too simple? You can blow these
things up very easy. The symptoms can be almost anything too, depending on
the area damaged in the chip. Maybe you should try to switch both CIA's
and see what happens. (BTW: I don't even know how the CIA-stuff looks in a
4000, I have a 2000 which had the same symptoms. I'm just suggesting
something!)
-MaRiO- -= Intel inside! (…my washing machine) =-
Confucius say: "A bird in hand makes it hard to blow nose"
Mario,
Yes, you are correct. Having the HHD spin only indicates that the
powersupply is still working (ie, it isn't a blown fuse or lose wire). My
techy suggested that it was probably a blown CIA. Unfortunately, I still
haven't heard because he was off sick for a couple days (spring fever???)
and hasn't had a chance to look at it. I hope it is that because I would
otherwise most likely be looking at big bucks to find the problem.
Thanks for your input.
Andrew
>> My techy suggested that it was probably a blown CIA.<<
You can probably test it for yourself. If you have ever seen the inside of
the Amiga (sorry, but I lost the type of yours? was it a 2000 or a 4000?)
you should notice that there are two CIA's inside (there are in my 2000).
You can swap these two and the Amiga should come up with different
symptoms now. That's the remedie I always try when I suspect a blown CIA,
it always worked with me. That's because both CIA's have different
functions in the machine. Be shure to handle these chips with great care,
however, because you can break off a pin very easy.
-MaRiO- -= Intel inside! (…my washing machine) =-
Confucius say: "A bird in hand makes it hard to blow nose"
Not in an A4000 where the CIAs are surface mounted (or is that just the
A1200?)
—
Dale L. Larson, Intangible Assets Manufacturing
The service manual says the 4000 uses the PLCC 8520s. Some clever hardware
maker ought to design an adapter to hack a dip 8520 onto a PLCC….
-Jim
MOS via Autopilot & Amiga
No need. You can socket them. But if they are already surface mounted,
you've got trouble…
—
Dale L. Larson, Intangible Assets Manufacturing
It's been a while since I saw the inside of a 4000, but I know the 1200 is
all surface mount except for the 2 ROMS. What I was thinking of was an
adapter that fit over the PLCC and held the dip version of the 8520. After
all, I don't believe the PLCC 8520s were ever available via service parts.
So we're going to have some dead 4000's and 1200's lying about in an
unfixable condition. Unless someone starts collecting fried MB's and
saving them for parts.
-Jim
MOS via Autopilot & Amiga
Again, the buffering on the A1200s and A4000s should prevent this. I seem
to recall trying moderately hard to blow one of those CIAs and not being
able to do it…
—
Dale L. Larson, Intangible Assets Manufacturing
>> After all, I don't believe the PLCC 8520s were ever available via service
parts.
So we're going to have some dead 4000's and 1200's lying about in an
unfixable condition. Unless someone starts collecting fried MB's and
saving them for parts.
When you think about it, we're all going to have some dead computers sooner or
later that are quite un-fixable unless somebody gets rights to manufacture the
custom chips. Even if nobody buys Commodore per say, it would be nice if
someone developed either clone chips or bought the rights so that repair shops
will be able to fix the computers in the future. Of course future repairs
would probably have to be done at the component, rather than board, level.
I think of the fact that I finally have the computer system that will do what I
want (including all those expensive periferals) and now I'm almost afraid to
use it for fear of it breaking. I've shopped many years before I settled on
the Amiga. I tried an IBM painting package and actually had to read the manual
just to get started <g>.
And the current solution to repairs seems diametrically opposed to the future
of Amiga. If enough people bail out, leaving their Amiga's behind, there will
be plenty of spare parts around for the rest of us. But if that happens, the
customer base is down the drain and it's not a very attractive situation for
developers and a prospective Commodore buyer. So we would have working
computers with no future<g>.
-Bruce
True,
There is a grand fear amoung many that in case of failure there
will be few who can handle the repair of our Amiga's.
However, there are a few interested groups trying to purchase
the rights to the Amiga. Some on the other hand are just interested
in the technology.
My fear is if Samsung actually buys Amiga that they will decide
not to continue with this platform. Of course this would be a
loss to us IF this happens but I am certain that most here will
agree that if that does it does not necessarily spell doom.
There may arise alternatives and other conduits for getting viable
parts.
As of now its still TOO quite out there and I wish someone could
give us the inside scoop as to what the heck is going on.
Shawn
Silent Paw Productions
MaRiO,
Thanks. That's useful information to keep on hand. Unfortunately, my
dealer has had my system for the past week (he keeps saying, "Tomorrow")
but I will try and remember that for the next time.
Andrew
I second Jim's advice. I often must transport my 4000. 50% of the time
something loosens up in transport and it won't boot. Opening the cover and
"wiggling" and reseating everthing solves the problem.
…AutoPilot Engaged…
-Paul-
Hi everyone, thanks for the advise. It's encouraging to know that it may just
be a loose connection. Unfortunately, we've been having some major thunder and
lighting so I don't feel comfortable working on the computer with so much
electricity about <g>. I'll try re-seating everything and let you know when
the storm blows over. Stay tuned….
-Bruce
Hi Bruce
This is a bit late perhaps but the first rule of troubleshooting a DEAD pc
of any kind is to strip it to the lowest configuration that will work. In
the Amiga's case that is floppy DH0: (the OS insists), all peripheral
cards out of the machine, and a monitor on the video out. If you don't get
the friendly "stick a disk in me" on the screen you are in deep dodo.
In that case check the power first with a meter on the big plug where the
PS plugs into the Motherboard. It is in the documentation. There is a fuse
inside the PS. Make sure all cables are plugged in well, especailly power
and Floppy cables. Still no go, you may have blown a chip but double check
that nothing is shorting the traces, turn it upside down and shake if
necessary to get anything out of the machine. Any pluggable chips should
be at least pressed to make sure they didn't "walk" out of the sockets.
The most likely to fail chips are the infamous 8520's, the Agnes and 5719,
but since the TBC is a video board and you aren't getting video to the
monitor I would look hard at agnes if you can't get anything on monitor.
It may not be a chip if all this fails, a cracked motherboard or cracked
solder joint can show up when you jam a new card in a slot too. This is
very hard to prove and to find, but fortunately is rare too.
Of course if you get video and can boot a WB diskette in DF0: start
putting things back until it fails again and that will guide you.
Sorry I can't give you the power pinout for the 4000, all I have are the
2000 and 3000 diagrams and I don't know that the PS is that similar.
Good luck, I hope you find something easy to fix.
Tom
alternate: jonest@hfsi.com
I don't make jokes, I just report what the government does…Will Rogers
Tom (and everybody else that responded), thanks for the assistance. I checked
the drive voltages and they were okay. I then removed the disk drives and the
daughter board with the Zorro slots. I started just gently pushing around on
the boards and heard a "click" as the board with the CPU seated down onto the
motherboard. I reassembled and horray, all works well!!
BYW, I have a revision 11 buster chip on a new motherboard (received it only
about 2 weeks before Commodore shut down <whew>). When I look at it, it
doesn't appear to be socketed. Should the sockets be obvious, or are the
surface mount sockets rather descrete?
-Bruce
Glad you're up & running again. Now that you remind me, I did see that
sensitive CPU card seating on at least one 4000 that I've serviced.
-Jim
MOS via Autopilot & Amiga
Every A4000 I've put a CPU card in (probably more than a dozen) has been
sensitive to how well the CPU card is seated. Probably 25% of the time it
wouldn't boot on the first try after getting a new CPU card.
—
Dale L. Larson, Intangible Assets Manufacturing
You've seen more 4000's than I. I only noticed it on one system out of
maybe 10. But then, the one system was one where I'd added a lot of
internal devices. So I probably moved it accientally.
-Jim
MOS via Autopilot & Amiga
Most of the 4000s (maybe all but my personal system) were prototypes
and/or first production run. This might explain.
—
Dale L. Larson, Intangible Assets Manufacturing
Bruce
Glad to hear you got it up ok. Statistically and based on experience many
of the problems are connectors…not seated, corroded etc.
>it doesn't appear to be socketed. Should the sockets be obvious,
Yeh, pretty obvious. The surface mounts look like a bug with his legs
glued to the board, much more low profile. You probably have surface
mount.
The reason for having surface mount was to reduce these connector related
problems, but if the chip fails it gets to be a problem in itself. You are
lucky indeed to get the new MB when you did. Rev 11 busters are rare today.
alternate: jonest@hfsi.com
I don't make jokes, I just report what the government does…Will Rogers
Tom,
not only does it reduce connector related problems, it is easier to
manufacture. Simply glue the chips down to the board and wave solder them
using a robot. Of course the soldering and glue makes it a pain to
replace them, as you have to cut the leads, break the chip off the board,
and then desolder each lead from the board. Not fun. I had to do it
exactly twice, and will never, EVER, do it again.
The Four Seasons of Southern California:
Fire, Flood, Earthquake, and Riot.
Brian
I've never even attempted it, my eyes aren't good enough for one thing.
What kind of soldering iron tool did you use to put the chips back on? One
of those bars that gets all the leads at one swipe?
Tom
alternate: jonest@hfsi.com
I don't make jokes, I just report what the government does…Will Rogers
Tom,
nope, I wish I were that lucky. Just yea old standard micro-mini
soldering station. Standard fare in the Navy.
The Four Seasons of Southern California:
Fire, Flood, Earthquake, and Riot.
Brian said:
>soldering station. Standard fare in the Navy.
But nevertheless you were able to hand solder SMT chips with it? Pretty
good, many people swear it is neigh to impossible to do that. But you did
say you would never do it again I believe 🙂
I just successfully repaied a chip with two leads broken off flush with
the chip body with an Iso-Tip rechargeable iron by running wirewrap wires
to the stubs and I thought I was doing something, but SMT is a lot more.
Tom
alternate: jonest@hfsi.com
I don't make jokes, I just report what the government does…Will Rogers
Tom,
I've been run through several courses which incorporated training to
the same standards that NASA requires. I'm fairly good at it, I just
don't have the patience to put in the required time on any manual labor.
I'll trouble-shoot something for days, without sleep, just don't expect me
to do detailed soldering <g>.
The Four Seasons of Southern California:
Fire, Flood, Earthquake, and Riot.
SMT doesn't just reduce connector-related problems, it also dramatically
reduces costs (the sockets are expensive, stuffing sockets is more
expensive than wave-soldering).
—
Dale L. Larson, Intangible Assets Manufacturing
Hi Dale
re:
>SMT doesn't just reduce connector-related problems, it also dramatically
reduces costs (the sockets are expensive, stuffing sockets is more
expensive than wave-soldering).
I know, all the textbooks say that. So why is a A1200 cost more than a
A500, and a A4000 cost more than a A3000 which even has a SCSI bus and a
built in flicker-fixer too? AGA chips cost that much more than ECS? Hmmm.
Tom
alternate: jonest@hfsi.com
I don't make jokes, I just report what the government does…Will Rogers
Because "selling price" and "production cost" are two different things.
—————————– YAMAGATA Hiroo
TJ >>I know, all the textbooks say that. So why is a A1200 cost more than
a A500, and a A4000 cost more than a A3000 which even has a SCSI bus and a
built in flicker-fixer too? AGA chips cost that much more than ECS?<<
While the A1200/4000 may not seem less expensive, they do offer more for a
cheaper price than the A500/2000/3000 did.
The A500 was priced around $599, when released. It was 7.14Mhz, used a
68000 CPU, had only 512K of 16-bit memory, had a external 86-pin 24bit
local expansion buss, and an internal A501 slot for a single 512K RAM
expander/clock card. The A501 card was an extra $199. The adaptor (A520)
for COLOR composite, and RF modulator was also extra.
The price did come down *later*, dropping as low as $199 for the A500, and
$50 for the A501. An the A1200 price will (would have?) come down.
For the same price, the A1200 was 14Mhz, used a 68ec020 CPU, offered FPU
support had 2-megs of 32-bit memory, a internal 150-pin 32-bit local
expansion buss, and a PCMCIA interface. It added in IDE interface. It
not only had a Video port, but a RF modular and color composite built-in.
Then it also had the upgrade graphic chip design (AGA).
The A1200 had (at least) twice the speed, four times the memory, hardware
for HDD support, support for TVs, composite monitors, and IBM/Clone
expansion cards (PCMCIA), and better graphics — straight out of the box.
—
The A4000 & A3000 are in a simular position with changes in the CPU used
(68030 vs. 68040 price). The Current A3000 may be sold cheaper, but the
origionally they may have not been.
Ron.
– ap'ing with AP. (Eagles may fly, but weasels aren't sucked
into jet engines.)
PMFJI
>> For the same price, the A1200 was 14Mhz, used a 68ec020 CPU, offered
FPU support had 2-megs of 32-bit memory… <<
That should be: 2-megs of 16-bit memory.
-MaRiO- -= Intel inside! (…my washing machine) =-
Confucius say: "A bird in hand makes it hard to blow nose"
MvG >>That should be: 2-megs of 16-bit memory.<<
The A1200 uses a 32-bit CPU, with 32-bit RAM access. The 68ec020, does
have a 24-bit address space, limiting address space to 16-megs, but using
a full 68020 or higher in the CPU expansion slot removes this limit.
The A3000's CPU had 32-bit access to CHIP RAM, but the ECS had only 16-bit
access.
Both the CPU & AGA chips used in the A1200/4000 have 32-bit access to CHIP
RAM. And the FAST RAM used in the A1200 CPU expansion (trapfdoor) is
32-bit.
The PCMCIA slot is a 16-bit standard, limiting it to 16-bit RAM, and
16-bit access from the CPU.
Ron.
– ap'ing with AP. (Eagles may fly, but weasels aren't sucked
into jet engines.)
Well I checked it out and you're right! I was getting mixd up with the
PCMCIA-mem.
BTW: why is FastRAM really so much faster than CHIP?
-MaRiO- -= Intel inside! (…my washing machine) =-
Confucius say: "A bird in hand makes it hard to blow nose"
MvG >>why is FastRAM really so much faster than CHIP?<<
There are two reasons.
—
The first reason is CHIP RAM, is a shared RAM area.
The CPU is not the only processor accessing CHIp RAM, and altering the
data held in this memory area. The Custom Chips (Paula, Agnus, & Denise)
— we'll called them the PAD — also access this RAM and preforms changes
on the data. Since both can't access the RAM at the same time, they have
to take turns.
While the CPU is accessing CHIP RAM, the PAD are doing internal processing
on the data it already has accessed,
While the PAD is accessing CHIP RAM, and the CPU is doing internal
processing on the data it just accessed.
The two processors share an interleaved accessing that allows both to run
at full speed, and *normally* have full access.
However, there are times when the job the PAD needs to do requires more
time than its allotted time slot. The PAD has the power to have a 'wait'
signal sent to the CPU, putting its access the CHIP RAM on hold.
The PAD can put the CPU on hold for upto three cycles, then must allow the
CPU to access CHIP RAM. The CPU is accessing time can be reduced to
one-fourth of its normal speed.
FAST RAM, is memory placed only on the CPU buss, and not the shared
CHIP-RAM buss. The CPU never has to wait for access to FAST RAM, so
access is faster, when compared to the CHIP RAM during heavy graphic use.
—
The second reason FAST RAM is faster, is that the CHIP RAM is limited to
the speed of the motherboard. If the FAST RAM is placed on a fast
CPU-card, it canbe as fast as the CPU on card allows.
While the CHIP RAM on an A1200 is "14-Mhz", a 40Mhz 68030 CPU card in the
trapdoor, can have "40Mhz" RAM plugged onto the card for the 50Mhz CPU, so
the CPU does so down while accessing the CHIP across the 14Mhz buss.
Ron.
– ap'ing with AP. (Eagles may fly, but weasels aren't sucked
into jet engines.)