9 Sep
2008
9 Sep
'08
4:17 a.m.
Hey all --
I've gotten the power supply in one of my ND-812s running (thanks for
the help w/the 1103 data which helped verify the voltages) and it
appears to be within spec under load (correct voltages, no ripple in the
output, etc...) so I powered the machine on and despite having my
fingers crossed, the machine does not run. (No real surprise there).
Random lights on the front panel, and basically no response to any
toggle switches.
I started pulling out and reseating the chips on the main CPU board and
powered it up after every few rows, and every time the behavior was
slightly different, so I'm fairly sure that the sockets and chips aren't
all making good connections anymore. Seems like solving this problem
would be a good place to start.
The problem is that there are 25 rows of 13 chips each (all socketed),
in very close proximity to one another. The chips are in sockets with
wire-wrap pins underneath -- the underside of the main CPU board is a
maze of wire wrapping. Most of the chips have date stamps between 1971
and 1973, but luckily most of them aren't corroded to the point where
they're falling apart.
How would you suggest cleaning the sockets and the chips to ensure good
connections?
Thanks,
Josh
10 Sep
10 Sep
12:08 p.m.
On 9 Sep 2008 at 1:17, Josh Dersch wrote:
How would you suggest cleaning the sockets and the
chips to ensure good
connections?
I'd probably start with a good contact cleaner, such as DeoxIT. If
you've got some Freon TF kicking around, that would also be something
to try.
Cheers,
Chuck
2:22 p.m.
On Tuesday 09 September 2008 04:17, Josh Dersch wrote:
Hey all --
I've gotten the power supply in one of my ND-812s running (thanks for
the help w/the 1103 data which helped verify the voltages) and it
appears to be within spec under load (correct voltages, no ripple in the
output, etc...) so I powered the machine on and despite having my
fingers crossed, the machine does not run. (No real surprise there).
Random lights on the front panel, and basically no response to any
toggle switches.
I started pulling out and reseating the chips on the main CPU board and
powered it up after every few rows, and every time the behavior was
slightly different, so I'm fairly sure that the sockets and chips aren't
all making good connections anymore. Seems like solving this problem
would be a good place to start.
The problem is that there are 25 rows of 13 chips each (all socketed),
in very close proximity to one another. The chips are in sockets with
wire-wrap pins underneath -- the underside of the main CPU board is a
maze of wire wrapping. Most of the chips have date stamps between 1971
and 1973, but luckily most of them aren't corroded to the point where
they're falling apart.
How would you suggest cleaning the sockets and the chips to ensure good
connections?
Take it outside and blast away at it with compressed air.
Some stuff that I've dealt with has a lot of fairly find dust all over the
innards of it, and even nice gold-plated machined-contact sockets aren't
working right any more because of that stuff -- which came into the equipment
by way of forced air cooling.
--
Member of the toughest, meanest, deadliest, most unrelenting -- and
ablest -- form of life in this section of space, a critter that can
be killed but can't be tamed. --Robert A. Heinlein, "The Puppet Masters"
-
Information is more dangerous than cannon to a society ruled by lies. --James
M Dakin
10:11 p.m.
If these are standard wire wrap boards with machine pin sockets - they are sort of self
cleaning on insertion.
with a screwdriver, (lever off adjacent IC) end then other end, walk the ic 3/4 out of the
socket and re-insert it.
10 minutes to do the whole board 25 rows of 13 chips.
I bet that brings you back up.
Pencil eraser on the gold fingers on the card edge too, until they are nice and bright.
Randy
Date: Wed, 10 Sep 2008 14:22:13 -0400
From: rtellason at verizon.net
To: cctalk at classiccmp.org
Subject: Re: Suggestions for cleaning hundreds of socketed chips?
On Tuesday 09 September 2008 04:17, Josh Dersch wrote:
_________________________________________________________________
See how Windows Mobile brings your life together?at home, work, or on the go.
http://clk.atdmt.com/MRT/go/msnnkwxp1020093182mrt/direct/01/
Hey all --
I've gotten the power supply in one of my ND-812s running (thanks for
the help w/the 1103 data which helped verify the voltages) and it
appears to be within spec under load (correct voltages, no ripple in the
output, etc...) so I powered the machine on and despite having my
fingers crossed, the machine does not run. (No real surprise there).
Random lights on the front panel, and basically no response to any
toggle switches.
I started pulling out and reseating the chips on the main CPU board and
powered it up after every few rows, and every time the behavior was
slightly different, so I'm fairly sure that the sockets and chips aren't
all making good connections anymore. Seems like solving this problem
would be a good place to start.
The problem is that there are 25 rows of 13 chips each (all socketed),
in very close proximity to one another. The chips are in sockets with
wire-wrap pins underneath -- the underside of the main CPU board is a
maze of wire wrapping. Most of the chips have date stamps between 1971
and 1973, but luckily most of them aren't corroded to the point where
they're falling apart.
How would you suggest cleaning the sockets and the chips to ensure good
connections?
Take it outside and blast away at it with compressed air.
Some stuff that I've dealt with has a lot of fairly find dust all over the
innards of it, and even nice gold-plated machined-contact sockets aren't
working right any more because of that stuff -- which came into the equipment
by way of forced air cooling.
--
Member of the toughest, meanest, deadliest, most unrelenting -- and
ablest -- form of life in this section of space, a critter that can
be killed but can't be tamed. --Robert A. Heinlein, "The Puppet Masters"
-
Information is more dangerous than cannon to a society ruled by lies. --James
M Dakin
11:19 p.m.
My take on this:
On 10 Sep 2008 at 21:11, Randy Dawson wrote:
with a screwdriver, (lever off adjacent IC) end then
other end, walk
the ic 3/4 out of the socket and re-insert it. 10 minutes to do the
whole board 25 rows of 13 chips.
I've found that stuff stored for a long time builds up a coating of
dust and gunk (oil, grease?). Better to use a solvent cleaner on the
chips and sockets than shove the crud back into the socket.
I wouldn't use a screwdriver--take some scrap steel or aluminum sheet
metal and use it to make an "L" shape. The length of the horizontal
part of the "L" should be nearly the length of the chip and about
2/3rds the width of the chip between pins; the vertical part should
be long enough for you to use it as a handle. Slip the leg of the L
under the chip and gently rock it from side to side using the
vertical handle until the chip pops out. You're much less likely to
bend the chip pins this way.
Pencil eraser on the gold fingers on the card edge
too, until they are
nice and bright.
If you have to use an eraser, use a gum-rubber one without the "grit"
most pencil erasers have. Better just to wipe the edge connectors
with a contact cleaning solution.
My .02 anyway,
Chuck
11:52 p.m.
The aluminum sheet idea is great, except that on this board there's
really no room to slide it in -- the chips are close enough together
that I can barely get a small flat-blade screwdriver between them to pry
(gently).
And I reseated all the chips the other night (very carefully). Going
through with a continuity tester on the first couple rows reveals that
about every other chip has at least one pin that's not making good
contact... this is going to be an ordeal :). The sockets themselves are
going to be a problem -- they have gold contacts (I _think_) but the
actual socket doesn't expose any accessible metal surface -- the pins go
through a hole (that's just barely larger than a pin) in the plastic
sheath and beneath that are the contacts. So getting contact cleaner in
there is going to be an interesting problem.
I've started taking the chips out one by one and cleaning the legs by
running them over a strip of scotch-brite. This clears off the
corrosion (and some of these are VERY corroded). I've then been (again,
carefully) bending the pins out slightly to allow them all to make good
contact with the socket. I did a row and a half last night, verifying
the continuity of the connections after each chip. It only took an hour
and a half :). At this rate, I'll have all 25 rows done in a couple of
weeks :).
I'd take some pictures to show you what I'm up against, but I lent my
camera out...
Thanks for the suggestions...
Josh
Chuck Guzis wrote:
My take on this:
On 10 Sep 2008 at 21:11, Randy Dawson wrote:
with a screwdriver, (lever off adjacent IC) end
then other end, walk
the ic 3/4 out of the socket and re-insert it. 10 minutes to do the
whole board 25 rows of 13 chips.
I've found that stuff stored for a long time builds up a coating of
dust and gunk (oil, grease?). Better to use a solvent cleaner on the
chips and sockets than shove the crud back into the socket.
I wouldn't use a screwdriver--take some scrap steel or aluminum sheet
metal and use it to make an "L" shape. The length of the horizontal
part of the "L" should be nearly the length of the chip and about
2/3rds the width of the chip between pins; the vertical part should
be long enough for you to use it as a handle. Slip the leg of the L
under the chip and gently rock it from side to side using the
vertical handle until the chip pops out. You're much less likely to
bend the chip pins this way.
Pencil eraser on the gold fingers on the card
edge too, until they are
nice and bright.
If you have to use an eraser, use a gum-rubber one without the "grit"
most pencil erasers have. Better just to wipe the edge connectors
with a contact cleaning solution.
My .02 anyway,
Chuck
11 Sep
11 Sep
8:49 a.m.
Josh Dersch wrote:
I've started taking the chips out one by one and
cleaning the legs by
running them over a strip of scotch-brite. This clears off the
corrosion (and some of these are VERY corroded).
Ugh. I've seen early ICs where the legs essentially rot from the inside-out;
the pins become extremely fragile and prone to just falling off, and no amount
of cleaning makes a good connection. I've brought machines back to life where
it's just been easier to ditch any IC that looks suspect and replace it with a
good part - it's far less trouble than trying to judge how good every single
IC leg is.
I know what a headache densely-packed boards can be. Sometimes I've found that
bending sewing pins right at the tips, hooking them under the ends of the IC
and then using a pair of pliers to pull on them is the best way of getting ICs
out.
cheers
Jules
1:06 p.m.
On Sep 11, 2008, at 8:49 AM, Jules Richardson wrote:
I've seen that too! WTF causes that?
-Dave
--
Dave McGuire
Port Charlotte, FL
I've
started taking the chips out one by one and cleaning the legs
by running them over a strip of scotch-brite. This clears off the
corrosion (and some of these are VERY corroded).
Ugh. I've seen early ICs where the legs essentially rot from the
inside-out; the pins become extremely fragile and prone to just
falling off, and no amount of cleaning makes a good connection.
3:43 p.m.
Dave McGuire wrote:
On Sep 11, 2008, at 8:49 AM, Jules Richardson wrote:
I've seen that too! WTF causes that?
I know there's been discussion here about it a few years ago. Possibly
something along the lines of the inside of the IC leg being a different
material to the outside plating, and in some circumstances it can decay much
faster than the outer material.
I'm not sure if there any wisdom as to whether any particular kind of
conditions cause it, or whether it's simply an age-related issue (possibly in
conjunction with a particular manufacturing procedure - I don't recall ever
seeing it* in anything newer than 1980 or so)
* Oddly, I don't think I've come across it in larger chips, either. It's
always been in ICs with smaller pin counts; was the production of smaller ICs
done in a different way during the '70s to ones with higher numbers of pins?
cheers
Jules
I've
started taking the chips out one by one and cleaning the legs by
running them over a strip of scotch-brite. This clears off the
corrosion (and some of these are VERY corroded).
Ugh. I've seen early ICs where the legs essentially rot from the
inside-out; the pins become extremely fragile and prone to just
falling off, and no amount of cleaning makes a good connection. 
4:08 p.m.
On Sep 11,
2008, at 8:49 AM, Jules Richardson wrote:
I've seen that too! WTF causes that?
I've started taking the chips out one by one and cleaning the legs by
running them over a strip of scotch-brite. This clears off the corrosion
(and some of these are VERY corroded).
Ugh. I've seen early ICs where the legs essentially rot from the
inside-out; the pins become extremely fragile and prone to just falling off,
and no amount of cleaning makes a good connection. 
8:53 a.m.
From: derschjo at msu.edu
---snip---
I've started taking the chips out one by one and cleaning the legs by
running them over a strip of scotch-brite. This clears off the
corrosion (and some of these are VERY corroded). I've then been (again,
carefully) bending the pins out slightly to allow them all to make good
contact with the socket. I did a row and a half last night, verifying
the continuity of the connections after each chip. It only took an hour
and a half :). At this rate, I'll have all 25 rows done in a couple of
weeks :).
Hi
Get some DoweCorning #4 silicon grease. Put this on the leads.
It will help improve the contacts and help keep future corrossion
from being a problem.
Many electrical shops have
it. I recall seeing it in the
McMaster Carr cat. as well.
Dwight
_________________________________________________________________
See how Windows connects the people, information, and fun that are part of your life.
http://clk.atdmt.com/MRT/go/msnnkwxp1020093175mrt/direct/01/
10:48 a.m.
On 10 Sep 2008 at 20:52, Josh Dersch wrote:
The aluminum sheet idea is great, except that on this
board there's
really no room to slide it in -- the chips are close enough together
that I can barely get a small flat-blade screwdriver between them to pry
(gently).
You could also try making an extractor that's modeled after forceps--
a short horizontal "L" section on each leg, just long enough to clear
the space between chips. Some of the pins on these old ceramic
packages are very fragile.
And I reseated all the chips the other night (very
carefully). Going
through with a continuity tester on the first couple rows reveals that
about every other chip has at least one pin that's not making good
contact... this is going to be an ordeal :). The sockets themselves are
going to be a problem -- they have gold contacts (I _think_) but the
actual socket doesn't expose any accessible metal surface -- the pins go
through a hole (that's just barely larger than a pin) in the plastic
sheath and beneath that are the contacts. So getting contact cleaner in
there is going to be an interesting problem.
DeoxIT comes in an aerosol form, which should allow for easy
penetration.
The concern here is that with abrasive methods, you're also rubbing
off the gold plating with the corrosion. That could mean some misery
down the road.
None of this may help you if the sockets have degraded to the point
where the contacts have deteriorated and cannot make good contact
even after cleaning. But a chemical cleaner such as DeoxIT is
probably the best way to approach the problem of dirty/corroded
contacts.
Cheers,
Chuck
12 Sep
12 Sep
3:12 a.m.
Chuck Guzis wrote:
None of the chips in question have gold plated pins, so I can't make
things too much worse there :). I'll get some DeoxIT and see how that
goes.
On 10 Sep 2008 at 20:52, Josh Dersch wrote:
I may just do that. I'm finding more and more of these chips are
leaving pins in the sockets no matter how carefully I extract them. I
have two of these ND-812 machines, one of which is basically good for
parts only (the power supply's been stripped and the front panel looks
like it was left in the sun too long (the toggle switches are actually
melted!) so I have a set of spare chips... I hope I can manage to get
one working machine out of the two.
The aluminum sheet idea is great, except that on
this board there's
really no room to slide it in -- the chips are close enough together
that I can barely get a small flat-blade screwdriver between them to pry
(gently).
You could also try making an extractor that's modeled after forceps--
a short horizontal "L" section on each leg, just long enough to clear
the space between chips. Some of the pins on these old ceramic
packages are very fragile.
And I reseated all the chips the other night
(very carefully). Going
through with a continuity tester on the first couple rows reveals that
about every other chip has at least one pin that's not making good
contact... this is going to be an ordeal :). The sockets themselves are
going to be a problem -- they have gold contacts (I _think_) but the
actual socket doesn't expose any accessible metal surface -- the pins go
through a hole (that's just barely larger than a pin) in the plastic
sheath and beneath that are the contacts. So getting contact cleaner in
there is going to be an interesting problem.
DeoxIT comes in an aerosol form, which should allow for easy
penetration.
The concern here is that with abrasive methods, you're also rubbing
off the gold plating with the corrosion. That could mean some misery
down the road.
None of this may help you if the sockets have degraded
to the point
where the contacts have deteriorated and cannot make good contact
even after cleaning.
I'm thinking that this may be the case. In which case
I'm not sure how
much it'd actually be worth replacing every socket... what a pain. :)
Thanks for the suggestions...
But a chemical cleaner such as DeoxIT is
probably the best way to approach the problem of dirty/corroded
contacts.
Cheers,
Chuck
11 Sep
11 Sep
12:24 p.m.
On Wednesday 10 September 2008, Josh Dersch wrote:
good contact... this is going to be an ordeal :). The
sockets
themselves are going to be a problem -- they have gold contacts (I
_think_) but the actual socket doesn't expose any accessible metal
surface -- the pins go through a hole (that's just barely larger than
a pin) in the plastic sheath and beneath that are the contacts.
It sounds like the sockets you have are of the style that never really
worked well. You may want to consider removing and replacing all of
the sockets, if you want the machine to work reliably, either with
turned-pin (machined) sockets, or cheap "wiper" blade sockets, both of
which are more reliable than what you have.
Pat
--
Purdue University Research Computing --- http://www.rcac.purdue.edu/
The Computer Refuge --- http://computer-refuge.org
12:41 p.m.
Patrick Finnegan wrote:
On Wednesday 10 September 2008, Josh Dersch wrote:
You mustn't have read the thread.
This machine is wire wrapped. Replacing the sockets would be tantamount
to building the boards from scratch.
good contact... this is going to be an ordeal :).
The sockets
themselves are going to be a problem -- they have gold contacts (I
_think_) but the actual socket doesn't expose any accessible metal
surface -- the pins go through a hole (that's just barely larger than
a pin) in the plastic sheath and beneath that are the contacts.
It sounds like the sockets you have are of the style that never really
worked well. You may want to consider removing and replacing all of
the sockets, if you want the machine to work reliably, either with
turned-pin (machined) sockets, or cheap "wiper" blade sockets, both of
which are more reliable than what you have.
Pat
2:05 p.m.
On Thursday 11 September 2008, Jim Battle wrote:
Patrick Finnegan wrote:
You're right that I haven't been following this thread... Still, my
comment stands. I don't think I'd consider it any more work to replace
WW ones and re-wrap them than if they were soldered onto a PCB
(especially if it has multiple layers), as long as you have some sort
of schematice/wire list. If you wanted to be creative, you could
probably cut the pins off the old sockets (or disassemble the sockets
to get the pins out), and figure out how to replace the sockets
(soldering new ones onto the old pins) without having to re-wrap
everything.
Pat
--
Purdue University Research Computing --- http://www.rcac.purdue.edu/
The Computer Refuge --- http://computer-refuge.org
On Wednesday 10 September 2008, Josh Dersch
wrote:
You mustn't have read the thread.
This machine is wire wrapped. Replacing the sockets would be
tantamount to building the boards from scratch. good contact... this is going to be an ordeal :).
The sockets
themselves are going to be a problem -- they have gold contacts (I
_think_) but the actual socket doesn't expose any accessible metal
surface -- the pins go through a hole (that's just barely larger
than a pin) in the plastic sheath and beneath that are the
contacts.
It sounds like the sockets you have are of the style that never
really worked well. You may want to consider removing and
replacing all of the sockets, if you want the machine to work
reliably, either with turned-pin (machined) sockets, or cheap
"wiper" blade sockets, both of which are more reliable than what
you have.
Pat 
5:11 p.m.
From: "Patrick Finnegan"
On Thursday 11 September 2008, Jim Battle wrote:
This is similar to a thought that I had. Couldn't you just pop off
the plastic bit, insert a machine pin socket into the WW socket, and
add a bit of solder to secure electrical and mechanical connection
between the the machine pins and the WW socket? (You'd have to make
sure the WW pins are anchored firmly to the board first.)
You could try just inserting the machine pin socket (without the solder),
but I don't know that it would make much better contact than your chips
are doing.
Vince
You mustn't have read the thread.
This machine is wire wrapped. Replacing the sockets would be
tantamount to building the boards from scratch.
You're right that I haven't been following this thread... Still, my
comment stands. I don't think I'd consider it any more work to replace
WW ones and re-wrap them than if they were soldered onto a PCB
(especially if it has multiple layers), as long as you have some sort
of schematice/wire list. If you wanted to be creative, you could
probably cut the pins off the old sockets (or disassemble the sockets
to get the pins out), and figure out how to replace the sockets
(soldering new ones onto the old pins) without having to re-wrap
everything.
12 Sep
12 Sep
3:16 a.m.
Patrick Finnegan wrote:
Whew. I dunno if I want to tackle that just yet. I have a schematic
set but no wire-list... not sure it'd really be worth the effort to
completely rebuild the board... I don't even have any actual software to
run on this thing (except for what I feel like toggling into the front
panel) once I get it running :).
But I'll keep it in mind. Thanks!
Josh
You mustn't have read the thread.
This machine is wire wrapped. Replacing the sockets would be
tantamount to building the boards from scratch.
You're right that I haven't been following this thread... Still, my
comment stands. I don't think I'd consider it any more work to replace
WW ones and re-wrap them than if they were soldered onto a PCB
(especially if it has multiple layers), as long as you have some sort
of schematice/wire list. If you wanted to be creative, you could
probably cut the pins off the old sockets (or disassemble the sockets
to get the pins out), and figure out how to replace the sockets
(soldering new ones onto the old pins) without having to re-wrap
everything.
Pat
11 Sep
11 Sep
12:53 p.m.
On 11 Sep 2008 at 12:24, Patrick Finnegan wrote:
It sounds like the sockets you have are of the style
that never really
worked well. You may want to consider removing and replacing all of
the sockets, if you want the machine to work reliably, either with
turned-pin (machined) sockets, or cheap "wiper" blade sockets, both of
which are more reliable than what you have.
ISTR that Josh mentioned that this was a wire-wrap board. I recall
some of the WW sockets of the day, ones marketed by TI, for example,
were miserable things that also had the unhappy tendency for the
contact "fork" to detach from the wire-wrap pin, if the pin got
wiggled too much.
I don't want to think of the effort involved in replacing them.
Probably easier to construct a new board.
Cheers,
Chuck
1:07 p.m.
--- On Thu, 9/11/08, Chuck Guzis <cclist at sydex.com> wrote:
ISTR that Josh mentioned that this was a wire-wrap
board.
I recall
some of the WW sockets of the day, ones marketed by TI, for
example,
were miserable things that also had the unhappy tendency
for the
contact "fork" to detach from the wire-wrap pin,
if the pin got
wiggled too much.
Ugh, I think I know the sockets of which you speak. Horrible.
It's funny, it seems that in the 70's, TI made simply the worst quality parts
imaginable. TI chips from this era are really prone to have corroded legs.
-Ian
1:39 p.m.
It's funny, it seems that in the 70's, TI made
simply the worst quality
parts imaginable. TI chips from this era are really prone to have corroded
legs.
Not that much different nowadays. Have you seen how thin and fragile are
new IC legs (at least in the TTL/CMOS line) from TI and Maxim? :oO
3:46 p.m.
Not that
much different nowadays. Have you seen how thin and
fragile are new IC legs (at least in the TTL/CMOS line) from TI and
Maxim? :oO
I wonder if this is because of new process technology for RoHS?
5793
days inactive
5796
days old
22 comments
14 participants
participants (14)
-
alexandre-listas＠e-secure.com.br -
cclist＠sydex.com -
derschjo＠msu.edu -
dkelvey＠hotmail.com -
frustum＠pacbell.net -
ian_primus＠yahoo.com -
josefcub＠gmail.com -
jules.richardson99＠gmail.com -
mcguire＠neurotica.com -
pat＠computer-refuge.org -
rdawson16＠hotmail.com -
rtellason＠verizon.net -
vrs＠msn.com -
wdonzelli＠gmail.com