Re: 7d01 with df1 garbled display

zenith5106 <zenith@...>
 

Miroslav,

Actually I think the contacts in the TI sockets were gold plated but
it didn't seem to help in this case. See
http://w1.871.telia.com/~u87138090/TexasC95a.jpg . The flat cables I
mentioned are the standard ones you'll find everywhere in older Tek
equipment. I put a picture at
http://w1.871.telia.com/~u87138090/Flatcable.jpg but it is very
difficult to visualize the difference with a picture. You'd really
need to see them both in real life and compare. The green is the bad
one and the blue is OK. When they were new the bad ones had a much
more pale goldish color and over the years they turned darker but
still different from a good ones which stay golden.
/Zenith

--- In TekScopes@..., "Miroslav Pokorni"
<mpokorni2000@y...> wrote:
Hello Zenith,

Yes, you are absolutely right. Those low-pressure tin-to-tin
contacts tend
to fail over time due to oxidation. It is only 'gas tight' and
noble metal
contacts that make reliable sockets. I do not quite understand how a
combination of tin plate and gold works, but it does. Over past
twenty years
there were number of attacks on tin plate to gold contact, about
every four
to five years and then it turns out that arguments are short on
substance
and whole thing gets dropped quietly. Usually, argument runs that
gold and
tin in contact would cause fretting corrosion of tin. The last three
instances that story was flogged, it came out of AMP, I guess
someone in
their metallurgical department is pursuing the line of fretting
corrosion.
Is it not funny that AMP is one of few companies that have been
pushing for
tin plated edge connectors? The last instance of gold to tin
controversy
that I know of was around 1997 or 98, when Intel issued statement
that
warranty on their motherboards was void if gold plated SIMMs were
used in
it. I was working for a memory company and we had to contend with
it,
because there were customers who would demand gold plating and then
others
who were heeding Intel's edict. We requested supporting
documentation from
Intel and sure enough, they sent us an article published by an AMP's
metallurgist. Knowing that it was just another tack of an old
theme, company
decided to disregard Intel's notice and prepared a letter for
customers who
expressed concern, explaining the previous history and seesaw of
tin to
noble metal contact. Intel's notice was on their web site and their
marketing was aggressively pursuing the matter through trade
magazines, but
after couple of months whole thing was quietly dropped. The first
hint was
that notice was removed from web site.

As for flat cable, would you please describe that connector and
cable more
closely so that we can identify it. I am guessing that brand was
Spectra
Strip, a division of Amphenol: connector was black body and cable
was color
coded; I believe that Spectra Strip had color coding of flat cable
trademarked, because no one else was using that for quite long
time, just
about right timing for trade mark to expire. Spectra Strip thought
that they
were smarter than rest of the bunch and they just about sunk
because of it;
good thing was that Amphenol was doing real well and could cart
them for
awhile.
Gold price of $800/oz brought grief to all of us and Insulation
Displacement
Connectors (IDC) seem to have been hit harder then others. Their
problem was
that cable part of connector did not have enough strength to live
through
insulation displacement and still exert enough pressure to form gas
tight
contact. The T & B made a name in IDCs, and I hope fortune, with
their tulip
contact. The insulation displacement part of that contact looked
like the
tulip flower, with four edges pressing into the wire and it relied
solely on
metal to exert pressure on wire, unlike other connectors (Spectra
Strip, 3M,
Berg) which relied on plastic body to support cutting edges of
contacts. In
first iteration, all those other companies changed contact to have
four
points to cut into wire, but cutting surfaces were highly
cantilevered, so
not enough force was supplied by metal. It was the next iteration
when all
of them got it right and it was only rarely, when someone tried to
re-use
connector or did not use right tool for assembly, that problem
reappeared. I
guess, since that time there were always batches when material was
not quite
right, like GM putting out a batch of cars with gear in
transmission which
skipped heat treating.

I guess, in this instance you can not blame Tektronix accountants
for trying
to unreasonably pinch few pennies, not that I would put that past
them. The
whole IDC industry went to selective gold plating, whatever
marketing name
they gave to it, so getting full gold plating without custom order
was just
not possible ever since it became clear that gold price was to stay
high.

Regards

Miroslav Pokorni


----- Original Message -----
From: "zenith5106" <zenith@t...>
To: "Miroslav Pokorni" <mpokorni2000@y...>
Sent: Thursday, May 29, 2003 9:38 AM
Subject: Re: 7d01 with df1 garbled display


Miroslav,
I won't argue with you on that but I'm pretty sure age is also a
factor. The way the contacts were designed in combination with low
tension it was easy for dirt it enter or more likely oxide to
build
up between the contacts an IC pin.
While we're at the topic there were another similar source of
intermittent faults. I guess at about the same time as the TI
sockets
Tek was trying to save money on the ribbon cable connectors that
were
in virtually every instruments. They used to be all gold plated
but
they changed it to be goldplated only at the contact areas and the
rest of the connector had some other cheaper plating. I think the
name of the connectors were "spot on gold" or "spot of gold". They
also were susceptable to oxide building up at the point it had
been
crimped to the cable. If you carefully wiggle intermittent cable
you
can watch the fault come and go. A simple solution is to carefully
solder the cable to the connector. They can be identified by
color,
compared to a real gold connector they were much more pale in
color.
The bottom line is, if you have an intermittent instrument which
contains these sockets and/or the these ribbon cables they should
be
the prime suspects.
/Zenith

--- In TekScopes@..., "Miroslav Pokorni"
<mpokorni2000@y...> wrote:
Hello Zenith,

Those sockets that you showed in picture look like famous TI
sockets that
wicked up flux and got fouled up. In most cases, they would be a
problem
from the start, on production line. Later on, TI tried to
correct
problem by
placing a rectangular piece of paper of some kind at the bottom
of
the
socket. The paper was pierced to let pins through and idea was
that
paper
would hug outside surfaces of pin and keep flux from wicking.
That
was
improvement, but was not good enough and that socket died. The
AMP
had a
cheap socket, but they learned from TI's experience. Pin in
AMP's
socket was
going all the way up the socket and was bent down to form one
side
of
contact with IC and then was bent upward again to make the other
side of
contact. With such a pin flux would wick up the pin but could
not
reach
contact area.

Regards

Miroslav Pokorni

----- Original Message -----
From: "zenith5106" <zenith@t...>
To: <TekScopes@...>
Sent: Wednesday, May 28, 2003 12:10 PM
Subject: [TekScopes] Re: 7d01 with df1 garbled display


--- In TekScopes@..., Joshua Van Tol <josh@s...>
wrote:
I've got a 7D01 with a DF1 data formatter that exhibits the
following
symptoms:

On power up, all the front panel lights come on. Sometimes
most
of
them
come on and then a few seconds later the rest come on.
When the display comes up, it's all garbled. Usually a whole
screen
of
the same character.

Any ideas what this might be? I'm thinking the processor is
either
resetting, or reading incorrect data from its ROM.
---
Miroslav is right. The IC sockets on early 7D01's were really
bad. If
I remember correctly they were Texas brand and type C95. They
were
installed in many Tek instruments at the time but seemed to
create
the most problems in 7D01. The 7D01 problems were so severe
that
Tek
even replaced several of the boards on warranty. It could be
easily
identified by its unique profile, looking at the short end. I
put
a
picture at http://w1.871.telia.com/~u87138090/TexasC95.jpg if
anybody
is interested. Another problem with early 7D01's was that the
cursor
switch and position pots were very sensitve to even small
static
discharge which would cause the instrument to reset when the
knobs
were touched by the operator. This was later corrected with a
new
type of pots with grounded shafts.
/Zenith

Join TekScopes@groups.io to automatically receive all group messages.