5 Jul
2019
5 Jul
'19
11:38 a.m.
I have an Apple /// that I've had for many years; it's never worked.
When you power it up, you get a checkerboard screen, where half the squares
are solid white, and the other half have a little mosaic pattern in them.
Looks like this:
https://share.icloud.com/photos/0NHNkEG9ssPsi65ojivBteKaQ
Does this failure mode ring any bells? Obviously the video signal is being
generated well enough to sync a composite output. Any idea whether I
should start by replacing the CPU or the ROMs?
Adam
5 Jul
5 Jul
2:21 p.m.
Adam Thornton wrote on Fri, 5 Jul 2019 11:38:56 -0700
I have an Apple /// that I've had for many years;
it's never worked.
When you power it up, you get a checkerboard screen, where half the squares
are solid white, and the other half have a little mosaic pattern in them.
Looks like this:
https://share.icloud.com/photos/0NHNkEG9ssPsi65ojivBteKaQ
Does this failure mode ring any bells? Obviously the video signal is being
generated well enough to sync a composite output. Any idea whether I
should start by replacing the CPU or the ROMs?
You have 40 by 24 very crisp characters there, so I would guess
everything about the video is ok. DRAMs are supposed to have random data
in them when power is first applied but that is not normally the case.
Instead half of the bits tend to always come up as 0 and the other half
as 1. These are normally in alternating rows inside the DRAM (all 1s in
one row and all 0s in the next row) which often leads to a checkerboard
pattern when used as a frame buffer. Normally the various DRAMs in a
byte don't have the exact same pattern so you might get different colors
or characters than 00h and FFh.
This indicates that the processor never wrote anything to the
framebuffer, so a bad CPU or ROM is likely though other chips could also
cause the software to not execute properly. Given the Apple /// design
it is very likely that all the clocks are good.
Another defect that leads to a similar visual effect is if the character
ROM is bad, but then you tend to have a uniform screen and not the
checkerboard.
The Apple /// is infamous for having chips work their way out of their
sockets due to heat, so you might start off checking for that.
-- Jecel
5:47 p.m.
Bad video memory?
Our? Lisa? does? that? too...
I? just? leave it? shit? off (grin!)
Ed#In a message dated 7/5/2019 2:28:31 PM US Mountain Standard Time, cctech at
classiccmp.org writes:
Adam Thornton wrote on Fri, 5 Jul 2019 11:38:56 -0700
I have an Apple /// that I've had for many years;
it's never worked.
When you power it up, you get a checkerboard screen, where half the squares
are solid white, and the other half have a little mosaic pattern in them.
Looks like this:
https://share.icloud.com/photos/0NHNkEG9ssPsi65ojivBteKaQ
Does this failure mode ring any bells?? Obviously the video signal is being
generated well enough to sync a composite output.? Any idea whether I
should start by replacing the CPU or the ROMs?
You have 40 by 24 very crisp characters there, so I would guess
everything about the video is ok. DRAMs are supposed to have random data
in them when power is first applied but that is not normally the case.
Instead half of the bits tend to always come up as 0 and the other half
as 1. These are normally in alternating rows inside the DRAM (all 1s in
one row and all 0s in the next row) which often leads to a checkerboard
pattern when used as a frame buffer. Normally the various DRAMs in a
byte don't have the exact same pattern so you might get different colors
or characters than 00h and FFh.
This indicates that the processor never wrote anything to the
framebuffer, so a bad CPU or ROM is likely though other chips could also
cause the software to not execute properly. Given the Apple /// design
it is very likely that all the clocks are good.
Another defect that leads to a similar visual effect is if the character
ROM is bad, but then you tend to have a uniform screen and not the
checkerboard.
The Apple /// is infamous for having chips work their way out of their
sockets due to heat, so you might start off checking for that.
-- Jecel
6 Jul
6 Jul
3:18 a.m.
Opps....that was supposed to say shut off!
In a message dated 7/5/2019 11:51:16 PM US Mountain Standard Time, cctech at
classiccmp.org writes:
Bad video memory?
Our? Lisa? does? that? too...
I? just? leave it? shit? off (grin!)
1:05 p.m.
In a message dated 7/5/2019 11:51:16 PM US Mountain
Standard Time, cctech at classiccmp.org writes:
Bad video memory?
Our? Lisa? does? that? too...
I? just? leave it? shit? off (grin!)
On Sat, 6 Jul 2019, ED SHATNER via cctalk
wrote:
Opps....that was supposed to say shut off!
Why?
"shut off" doesn't convey the mood of a bad video display.
5 Jul
5 Jul
11:55 p.m.
<snip>
Aside from the chips/sockets fix, verify that the keyboard light comes on.
If the bulb is dead the system will not boot (someone can "2nd" this).
Verify that the keyboard reset key is not stuck. Overall I'd find a
working system and compare/contrast voltages at key obvious points. Verify
your boot media and disk drive are OK by comparing with a known-working
system.
Where are you located? Maybe someone near you can offer to assist.
Bill
Looks like this:
https://share.icloud.com/photos/0NHNkEG9ssPsi65ojivBteKaQ
Does this failure mode ring any bells? Obviously the video signal is
being
generated well enough to sync a composite output.
Any idea whether I
should start by replacing the CPU or the ROMs?
<snip>
The Apple /// is infamous for having chips work their way out of their
sockets due to heat, so you might start off checking for that.
-- Jecel
4:28 p.m.
The first 15,000, or so, Apple ///s had a problem where the very large,
dense, poorly mechanically-supported motherboard would warp as it heated up
after power-on. That resulted in ICs popping up in sockets enough to break
contact, as well as micro-cracks in printed circuit board traces. The
warping would occur after about 10 - 15 minutes badly enough to cause the
problems. Apple quietly replaced those systems when people brought them
in, including mine.
If any of the ICs on your board have popped up enough to cause problems,
reinserting them may not be enough to fix the problem. High humidity and
fine dust can cause socket contacts to become corroded or dirty, making
operation of the computer flaky.
Micro-cracks occur as the board warps while heating up and stretches the
traces on the expanding side of the board. They open when expansion is
sufficient and close when the board cools down enough, generally causing
operation to cease and resume after a fairly constant amount of of time
being powered on and powered off, dependent on the ambient temperature.
The warmer it is, the shorter the time, and vice versa when it?s cooler.
One potential solution, if the warping is occurring, is to remove the board
and drill the mounting holes out to slightly enlarge them. Then, when
remounting the board, don?t fully tighten the screws, allowing the board to
?float? and expand without pushing against the screws and warping.
It?s possible that at least one hole is being used as a connection to the
metal base as a frame ground, which helps reduce RF emissions that might
interfere with very sensitive devices, especially nearby and/or on the same
AC power circuit. Not tightening associated screws might prevent that
interference-reduction from occurring, but, that?s secondary to solving the
failure to boot.
ICs can be swapped out one at a time with those in a known working board,
if one IC has failed. However, extreme care in not bending any pins when
reinserting them must be performed. If more than one has failed, the
combinatorics of the number of swaps needed increase exponentially with the
number of failed components. Use of an o scope, logic analyzer, etc.,
along with extensive digital signal troubleshooting knowledge (especially
without documentation, including schematics and timing diagrams) may be the
only recourse.
Should you or any of the IM Force be caught or killed, The Secretary will
disavow any knowledge of your actions. This device will self-destruct in
five seconds. Good luck, Mr. Phelps.
FFFFFFSHSHSHSHSSSSSSSSKIKIKIKIK!!!
On Fri, Jul 5, 2019 at 12:39 PM Adam Thornton via cctech <
cctech at classiccmp.org> wrote:
I have an Apple /// that I've had for many years;
it's never worked.
When you power it up, you get a checkerboard screen, where half the squares
are solid white, and the other half have a little mosaic pattern in them.
Looks like this:
https://share.icloud.com/photos/0NHNkEG9ssPsi65ojivBteKaQ
Does this failure mode ring any bells? Obviously the video signal is being
generated well enough to sync a composite output. Any idea whether I
should start by replacing the CPU or the ROMs?
Adam
1779
days inactive
1780
days old
6 comments
6 participants
participants (6)
-
athornton＠gmail.com -
billdegnan＠gmail.com -
cisin＠xenosoft.com -
couryhouse＠aol.com -
jecel＠merlintec.com -
jim.manley＠gmail.com