cdh@mtu.edu (Chris Hooper) (05/13/91)
Yes, this is a repost. Since I received zero words of
advice my first time around, I'm hoping someone will
have something to say (flame or not) that might help
me with my situation. I'm getting desperate quick :)
My Amiga is dead.
Machine:
Amiga 2000, AmigaDOS 1.3 in ROM, Rev 4.3 motherboard
1M Agnes, Supra 8-up (with 2M), IVS Trumpcard
Problem/Symptoms:
A diode was burned out on the motherboard (actually burned clean
through). It was replaced (1N4001) located near the power supply
connector to the motherboard. A technician at the company that
sells the Megachip said diaode D400 is on the +12 power input.
He then told me that the particular revision motherboard I have
(B2000 Rev 4.3) is missing a capacitor (C908) which is needed for the
Megachip board to work. He stated this capacitor is listed as 100pf,
but recommeded 70pf. I still have yet to put this capacitor in.
My machine refuses to boot.
White screen on power up (with *no* boards in machine, no floppy
in drive). Goes through gray colors, but when it should come up
with a work bench prompt, stays white instead.
Floppy drive clicks like it is waiting for a disk.
When disk is inserted, floppy stops clicking, however, access light
does not come on drive and drive does not spin. Pull the disk out
and the drive starts clicking again.
With the Trumpcard in, the machine will attempt to Autoboot off
the hard drive. It usually does not succeed. The background color
*does* change from white to blue (which is my default background color).
Three out of the many powerup attempts let to a screen of garbage (kinda
looked like a page zero), with moving bits on the screen. Once or twice
the machine actually booted to the point where I could type a command and
it would access the disk (could not see what I typed still). The
machine usually died pretty quickly. Most of my attempts at booting,
however, ended up with the machine just stopping completely before
it finished booting. By stopping, I mean no activity; CPU halted or
waiting for something.
We tried using a logic probe, but everything looked "normal" as far
as we could tell. ie: yup, there was power to memory, chips, etc.
Ok, I'm a CS major (with enough electrical experience to be dangerous)
and my friend is a EE (both undergrads).
Since it goes through its power-up colors and booted (kinda) a couple
time, I assume the 68000 is running fine. Because the floppy drive
is also acting up, I am guessing it might have to do with a chip
around Gary that also has a hand in chip memory access (maybe two
chips are bad??)
Suspected Killers:
Megachip 2000 (The board for about $350 which allows you
to put a 2M Agnes in the 2000)
Supra 8-up (Either the Megachip, or the Megachip in
combination with this board)
How it happened (I think):
The roommate purchased this board originally for his machine,
but had to send his motherboard in for replacement since it
was a revision 6 and would not work with the Megachip.
While his machine was gone, I became interested in possibly
having a 2M board in my machine :)
So: I pulled my machine completely apart (motherboard too)
so I could get that darn original Agnes out of its socket.
I had to use the holes on the bottom as well as the chip
puller to pry that sucker from its wedge. Thinking back, I
really had to torque on the chip socket to get it out. Its
highly possible that I damaged the Agnes socket while
pulling the chip. I am leaning toward that idea now.
I put the new Agnes board in, and put the machine all back
together.
It worked beautifully (2M chip, 2M fast - Supra) for about an
hour. I then began to notice a flash occasionally. At first
I thought it might be power or the monitor acting up (dorms).
The longer I left the machine on, the worse the flicker got.
If I powered my Amiga off, then back on the flicker would go
away for a while, but always return faster than previously.
I pulled the Supra and Trumpcard out and the problem went
away. I then put the Trumpcard back in and the problem still
did not return. I assume there must have been some conflict
between the Supra and the Megachip.
I let the machine run for a whole day without the Supra board
and it did not give incident. I then put the Supra back in.
The problem did not come back immediately.
Then I did the foolish thing. I turned the monitor off and
let the machine sit for a day (I usually leave my machine on
24 hours a day). When I came back the next day, the screen
was blank (border color). I cycled power and the machine came
back on for a few seconds and then quit. I tried again;
same result. My Amiga has been unusable since.
At this point, I pulled my machine all apart and removed the
Megachip. I replaced the original Agnes back into its socket.
When I reassembled the machine, the screen did not return!
I tried pulling and reinserting Agnes; to no avail.
Over the next few days, I had the opportunity to try another
1M Agnes and also a 512K Agnes. Each produced exactly the
same results.
What I have already tried:
Replacing the diode that was burned clean through.
Swapping every pullable chip on the motherboard
[68000, Agnes, Paula, Denise, Gary, ROM, the two 8520
chips, etc].
Probing around randomly with a logic probe.
Praying to the Magic Montra
Sacrificing the neighbor's cat. (Just kidding, all you
animal lovers out there)
Anyone with any ideas or suggestions, please help.
Thanks.
-Chris
- Chris Hooper Computing Technology Services Consultant
cdh@mtu.edu Michigan Technological University
Also: cdh@mtus5.mtu.edu CDH@MTUS5.Bitnet and cdhooper@symmetry.cs.mtu.eduburkert1@platon.fmi.uni-passau.de (Klaus Burkert) (05/15/91)
In article <1991May13.031522.28822@mtu.edu> cdh@mtu.edu (Chris Hooper) writes: >My Amiga is dead. > >Machine: > Amiga 2000, AmigaDOS 1.3 in ROM, Rev 4.3 motherboard > 1M Agnes, Supra 8-up (with 2M), IVS Trumpcard > My configuration: Amiga 2000, AmigaDOS 1.3 in ROM, Rev 4.3 motherboard 1M Agnes, GVP-Series I Hardcard, 4 Meg Fast-RAM >Problem/Symptoms: > A diode was burned out on the motherboard (actually burned clean > through). It was replaced (1N4001) located near the power supply > connector to the motherboard. A technician at the company that > sells the Megachip said diaode D400 is on the +12 power input. > He then told me that the particular revision motherboard I have > (B2000 Rev 4.3) is missing a capacitor (C908) which is needed for the > Megachip board to work. He stated this capacitor is listed as 100pf, > but recommeded 70pf. I still have yet to put this capacitor in. > I don't know, how or why You burned that diode. Also I don't anything special about the MegaChip, I simply saw it at the Amiga'90 expo in Cologne, Germany , but it was too expensive (DM 600.- (german currency, ~350.- US$) _without_ 2-Meg-Agnus), so I can't help you with the capacitor. [ stuff deleted ] >What I have already tried: > Replacing the diode that was burned clean through. > > Swapping every pullable chip on the motherboard > [68000, Agnes, Paula, Denise, Gary, ROM, the two 8520 > chips, etc]. > > Probing around randomly with a logic probe. > | > Praying to the Magic Montra | > | > Sacrificing the neighbor's cat. (Just kidding, all you | > animal lovers out there) | does this help ? :-) > >Anyone with any ideas or suggestions, please help. > Due to the high pricing mentioned above, I decided to do a hack by my own. (Too much to post or mail, a lot of trace-cutting, de- and re-soldering, piggy-backing, etc. I took me about three or four days before working perfectly). During this hack, I ran into some similar problems: Sometimes, the machine booted, sometimes not, or it booted perfectly from floppy or HD but screen remained white. I got the moving-bits-display, too. The reason was a floating (i.e. bad contact) Address-Line at Agnus. I think,there are only few possible defects: - Agnus is dead (you swapped it already?) - some RAM-chips are dead (*) - the RAM-buffers are dead (**) - Agnus' socket is damaged (*): the RAM-area is split into two 512K banks. If the lower bank would be defective, you'd get only a _green_ screen (error in first 256K of chip-ram). If the upper bank would be flaky, you should try switching the jumper J101, when it helped then there is/are one/several bad RAMs in the upper bank. (**): when the buffers are dead, you'd get only the green screen, too. I think it's the last possibility, because you already mentioned how hard it has been to pull agnus out of it's socket. So here's my advice: replace the socket and all should work well. >Thanks. > -Chris I hope this helps and you can "reanimate" your Amiga! >- Chris Hooper Computing Technology Services Consultant > cdh@mtu.edu Michigan Technological University > >Also: cdh@mtus5.mtu.edu CDH@MTUS5.Bitnet and cdhooper@symmetry.cs.mtu.edu ------------------------------------------------------------- Klaus Burkert email: burkert1@platon.fmi.uni-passau.de Brandweg 11 voice: +49-851/83993 D-W-8390 Passau / Federal Republic of Germany
daveh@cbmvax.commodore.com (Dave Haynie) (05/15/91)
In article <1991May13.031522.28822@mtu.edu> cdh@mtu.edu (Chris Hooper) writes: >My Amiga is dead. >Problem/Symptoms: > A diode was burned out on the motherboard (actually burned clean > through). That's a real bad sign. It takes a lot of current to burn through a power diode. What the heck was connected to this machine? > It was replaced (1N4001) located near the power supply connector to the > motherboard. A technician at the company that sells the Megachip said What's a megachip? > diaode D400 is on the +12 power input. D400 is a "protection" diode between "+12V", the internal +12V supply, and "+12V_USER", which is the +12V supply that goes to your video and serial connectors. Damage to this diode indicates that you did something horrible to either the video or serial port. Often, plugging in a connector while the system is on can cause pins to short. There's no certainty that this diode was the only thing zapped, of course. > He then told me that the particular revision motherboard I have > (B2000 Rev 4.3) is missing a capacitor (C908) which is needed for the > Megachip board to work. He stated this capacitor is listed as 100pf, > but recommeded 70pf. I still have yet to put this capacitor in. On the B2000 motherboard, 900-series capacitors were additions made for FCC purposes. Sometimes they would be required for one revision of something, not for another. They should not have any effect on any proper expansion device (which I already gather this "Megachip" isn't). > White screen on power up (with *no* boards in machine, no floppy > in drive). Goes through gray colors, but when it should come up > with a work bench prompt, stays white instead. > Floppy drive clicks like it is waiting for a disk. > When disk is inserted, floppy stops clicking, however, access light > does not come on drive and drive does not spin. Pull the disk out > and the drive starts clicking again. Whatever zapped your +12_USER supply could very well have zapped something else. If you zapped it through the serial port, it's possible, though not real likely, that the 8520 CIA chip at U301 was also damaged. That would affect floppy and parallel port operation, along with some timing functions (this is the CIA timed off of HSYNC). If you zapped it through your video port, you could have damaged the Agnus or Denise chips. Agnus, of course, is the RGA/Chip bus master, and nothing works if there's a real problem with Agnus. It's really impossible to tell what actually is wrong over the net, you need a good repair technician. Also, since I gather this "Megachip" thing is plugging down into chip sockets, not a proper expansion bus slot, check out the damage that may have caused. Check any chips that you removed and make sure they are in correctly and don't have any pins bent. Check out the sockets, most "tower" type plug ins use Augat type sockets as their "feet". These sockets can bend the pins of the socket they go into to the point where the smaller pins of an IC no longer make good contact. > Most of my attempts at booting, however, ended up with the machine just > stopping completely before it finished booting. By stopping, I mean no > activity; CPU halted or waiting for something. From what you said, I expect that the CPU itself is OK. It's probably one of the custom chips, either Agnus, Denise, or that 8520, having a problem. My best guess would be Agnus. This is just a guess, though. To check out Agnus activity, you could probe around and look for RGA bus activity during non-CPU cycles, but you really need a logic analyzer to get real far with that, or at least a 'scope; a logic probe isn't going to help. > At this point, I pulled my machine all apart and removed the > Megachip. I replaced the original Agnes back into its socket. > When I reassembled the machine, the screen did not return! Either your pulling on the socket, or the presence of this Megachip thing, may have messed with the Agnus socket. These sockets are designed to hold chips, not tower assemblys, so as I mentioned previously, it could be that the tower bent back the socket's pins far enough to no longer allow the use of the socket as a normal chip carrier. Or perhaps you bent the pins in removing the original Agnus. In either case, that could explain some of your problems, but it doesn't have anything directly to do with D400 getting zapped, since Agnus doesn't use +12V, and in fact, all motherboard use of +12V would come straight from the +12V supply, not +12V_USER. > Sacrificing the neighbor's cat. (Just kidding, all you > animal lovers out there) Nope, cat sacrifices aggrevate Amiga problems, though they have been known to fix IBMs. You need to sacrifice a Marketroid to fix an Amiga, if you're looking for shortcuts. Actually, though, it's already too late, that only works within a week of the vernal equinox, as I recall... >- Chris Hooper Computing Technology Services Consultant -- Dave Haynie Commodore-Amiga (Amiga 3000) "The Crew That Never Rests" {uunet|pyramid|rutgers}!cbmvax!daveh PLINK: hazy BIX: hazy "That's me in the corner, that's me in the spotlight" -R.E.M.