1 Apr
2005
1 Apr
'05
3:32 a.m.
Well I identified the failure, fixed it, and the disk pack is
formatting right now. Whew!
Another crapacitor -- looks like a tantalum, but it's only .01uF,
a dipped-looking bright blue axial part, about the size of a 1/8W
resistor. I think these were discussed recently; not necessarily
tants, but some short-lived process.
Anyways, the problem cap was on the read/write amp board. The bad
cap was on the -15V line. The shorted cap took out the 1-ohm, 1/4W
series resistor on the logic board (very obvious once I looked at
it; all the smoke what went out of it, and all the color-code
bands went brown).
There's only one chip (LM311) using -15, took that out, nope.
Started following copper, removing caps, finally saw one blue cap
that was... brown in the middle! Yup, shorted.
I replaced the bad one, and all the lifted caps, with ceramic
disks since I had 'em partly out anyways.
Had to take the head assembly dust cover off, to get the head
connectors plugged on, which mean clean-room again. Unnerving.
Still running diags, but no errors so far.
Man, there's hundreds of those little blue bastards in this
machine. I hope future shorts are so benign. Gulp.
1 Apr
1 Apr
4:37 a.m.
In message <20050401002020.H740 at localhost>
Tom Jennings <tomj at wps.com> wrote:
Man, there's hundreds of those little blue
bastards in this
machine. I hope future shorts are so benign. Gulp.
Probably worth keeping Murphy's Laws in mind - "if something can go wrong, it
will" and "if a component fails, it will take the most expensive or hardest
to replace part with it".
There's also Philpem's Law (<very big grin>): If you slip with a test
probe,
that probe will either short out the most expensive chip on the board, or put
a gouge into a multilayer PCB.
Anyway, I'd replace them if they had a high failure rate, if only to reduce
the likelihood of them taking something with them when they do pack up.
Later.
--
Phil. | Acorn Risc PC600 Mk3, SA202, 64MB, 6GB,
philpem at philpem.me.uk | ViewFinder, 10BaseT Ethernet, 2-slice,
http://www.philpem.me.uk/ | 48xCD, ARCINv6c IDE, SCSI
... Life. Hate it, or ignore it. You can't like it.
6:21 p.m.
There's also Philpem's Law (<very big grin>): If you slip with a test
probe,
that probe will either short out the most expensive chip on the board, or put
a gouge into a multilayer PCB.
Is this related to DrARD's law? : Any tool dropped will find the most
serious points to short-circuit.
-tony
9 Apr
9 Apr
6:54 p.m.
----- Original Message -----
From: "Philip Pemberton" <philpem at dsl.pipex.com>
To: <cctalk at classiccmp.org>
Sent: Friday, April 01, 2005 10:37 AM
Subject: Re: D.G. Nova update -- crapacitors strike again
Probably worth keeping Murphy's Laws in mind -
"if something can go wrong,
it
will" and "if a component fails, it will
take the most expensive or
hardest
to replace part with it".
Later.
--
Phil.
I seem to recall the following specific examples of Murphy's (sod's ) law :-
1. The more innocuous a design change appears , the further it's influence
will extend.
2. In any given computation , the figure that is most obviously correct will
be the source of error.
3. Any wire cut to length will be too short.
4. A device selected at random from a group having 99% reliability will
belong to the 1% group.
5. The probability of a dimension being omitted from a plan or drawing is
directly proportional to it's importance.
6. Interchangeable parts won't.
7. A circuit protected by a fast acting fuse will protect the fuse by
blowing first.
8. A failure will not appear until a unit has passed final inspection.
9. After the last of 16 mounting screws has been removed from an access
cover , it will be found the wrong access cover has been removed.
10. After an access cover has been secured by 16 screws , it will be found
that the gasket has been left off.
and my favourite :
11. Suggestions made by the value analysis group will increase costs and
reduce functionality.
Geoff.
1 Apr
1 Apr
6:23 a.m.
On Fri, 2005-04-01 at 00:32 -0800, Tom Jennings wrote:
Well I identified the failure, fixed it, and the disk
pack is
formatting right now. Whew!
Another crapacitor -- looks like a tantalum, but it's only .01uF,
a dipped-looking bright blue axial part, about the size of a 1/8W
resistor. I think these were discussed recently; not necessarily
tants, but some short-lived process.
That's one thing that bugs me about old kit for which there aren't
schematics around - tantalum caps often completely explode. If the value
was critical for some reason, then that creates a problem!
I've started taking lots of photos of any rare boards I have these days,
just in case. Once a cap explodes or a resistor burns out it might be
impossible to tell what the right part should be (well, without serious
circuit analysis and/or guesswork)
6:25 p.m.
I've started taking lots of photos of any rare
boards I have these days,
just in case. Once a cap explodes or a resistor burns out it might be
impossible to tell what the right part should be (well, without serious
circuit analysis and/or guesswork)
I prefer to draw out annotated schematics of anything remotely odd. That
way I have a good chance of being able to fix it.
Unless your photographs are seriously good, you will have trouble reading
component values, and maybe even recognising colour bands 'is that violet
or brown, is that one red or orange?'
-tony
12:57 p.m.
New subject: D.G. Nova update -- crapacitors strike again reprise
On Fri, 1 Apr 2005, Tom Jennings wrote:
Well I identified the failure, fixed it, and the disk
pack is
formatting right now. Whew!
That was at midnight. At about 1am, initializing (OS image) the
last sectors of the last surface -- really -- it crapped out
again. Similar symptom; infinite read errors. Seeks, appears to
format OK) can't read sector address headers.
I *hope* it's the same problem, which means shotgunning all those
@#_(*&#_(#*&@_#@# G.D.M.F. blue-death caps.
(The Read/Write Amplifier Board in this 6070 drive is an ugly
thing; lots of ECOs, a .25" hole drilled in it, to hold an added
trimpot! with lots of wiggly wires. ALso the PCB layout doesn't
match the drawings (schematic at least mostly does) so it's likely
an early rev or something. I hope this means that it's the only
thing with blue-death caps on it. I can hope can't I?)
2:18 p.m.
Upon the date 03:32 01-04-05, Tom Jennings said something like:
Well I identified the failure, fixed it, and the disk
pack is
formatting right now. Whew!
Another crapacitor -- looks like a tantalum, but it's only .01uF,
a dipped-looking bright blue axial part, about the size of a 1/8W
resistor. I think these were discussed recently; not necessarily
tants, but some short-lived process.
Oh yeah, these are likely monolithic ceramic caps. We used them in various
products as logic supply bypass caps where I used to work. I personally
have not seen any failures like this in my own experience. I have a bunch
of 0.1 uF parts on reels around here somewhere.
However, disc ceramic capacitor failures were a recent topic of discussion
on one of the old radio email lists. Failure mode was migration of the
capacitor plate material through the disc because of flaws or cracks. High
leakage current or outright shorts were the problems observed. I can see
that happening with axial monolithic caps if there were manufacturing
quality problems with the ceramic materials used to make the parts.
Man, there's hundreds of those little blue bastards
in this
machine. I hope future shorts are so benign. Gulp.
May be just the only one you'll ever see but at least you've learned by
experience a failure mode of these parts.
Regards, Chris F.
NNNN
Christian Fandt, Electronic/Electrical Historian
Jamestown, NY USA cfandt at netsync.net
Member of Antique Wireless Association
URL: http://www.antiquewireless.org/
7:57 p.m.
On Fri, 1 Apr 2005, Christian R. Fandt wrote:
Oh yeah, these are likely monolithic ceramic caps. We
used them in various
products as logic supply bypass caps where I used to work. I personally have
not seen any failures like this in my own experience. I have a bunch of 0.1
uF parts on reels around here somewhere.
Well this one sure failed -- turned brown from heat! It was just an
ordinary monolithic, but I replaced them all just the same :-)
However, disc ceramic capacitor failures were a recent
topic of discussion on
one of the old radio email lists. Failure mode was migration of the capacitor
plate material through the disc because of flaws or cracks. High leakage
current or outright shorts were the problems observed. I can see that
happening with axial monolithic caps if there were manufacturing quality
problems with the ceramic materials used to make the parts.
As tolerances tighten in general parts are run closer to their
ratings; my solution is to use 100V and up disk ceramics. I stock
only so many parts, this makes life easier...
My guess is that all the tants that fail, lower-voltage ratings
fail more often than higher-voltage... just guessing...
7040
days inactive
7048
days old
8 comments
6 participants
participants (6)
-
ard＠p850ug1.demon.co.uk -
cfandt＠netsync.net -
geoffreythomas＠onetel.com -
julesrichardsonuk＠yahoo.co.uk -
philpem＠dsl.pipex.com -
tomj＠wps.com