[Tom Del Rosso[_4_]]

Getting down into the details of the booster guidance computer from IBM.

http://www.youtube.com/watch?v=vZhRb...em-uploademail


--

Reply in group, but if emailing remove the last word.

[Tom Del Rosso[_4_]]

Tom Del Rosso wrote:
Getting down into the details of the booster guidance computer from
IBM.
http://www.youtube.com/watch?v=vZhRb...em-uploademail

Pins are soldered to the hybrid substrate, and then the pins are welded to
the PC board. Somehow the welding did not melt the pins off the substrate.
(???)


--

Reply in group, but if emailing remove the last word.

[Tim Williams]

Spot welded? (I'll watch later)

Tim

--
Seven Transistor Labs
Electrical Engineering Consultation
Website: http://seventransistorlabs.com

"Tom Del Rosso" wrote in message
...

Tom Del Rosso wrote:
Getting down into the details of the booster guidance computer from
IBM.
http://www.youtube.com/watch?v=vZhRb...em-uploademail

Pins are soldered to the hybrid substrate, and then the pins are welded
to the PC board. Somehow the welding did not melt the pins off the
substrate. (???)


--

Reply in group, but if emailing remove the last word.

[Tom Del Rosso[_4_]]

Tim Williams wrote:
Spot welded? (I'll watch later)

But what temperature would that be at?


--

Reply in group, but if emailing remove the last word.

[Tim Williams]

"Tom Del Rosso" wrote in message
...

Tim Williams wrote:
Spot welded? (I'll watch later)

But what temperature would that be at?

Well.. welding is welding, yes, but you get very different results between
highly localized and loosely generalized heating methods. In order of HAZ
(heat affected zone) size, smallest to largest: e-beam, laser, spot, arc,
torch, friction, induction, forge. (Give or take various considerations
for each process, but I think this is a pretty typical ordering.)

E-beam and laser do highly localized heating (down to the micron), with
e-beam having more penetration I would think (an electron beam as high as
~1MeV can shoot, I think, a milimeter or so into solid metal?).

Spot welding is a traditional thermal method, but it's heatsinked with
solid copper electrodes, and done in a relatively short pulse (100ms for
the better ones, though you may spend seconds with a cheap unit;
obviously, YMMV). So the HAZ is small (ay from the weld, heating drops
off rapidly). Precedent: vacuum tubes are all spot welded construction;
generalized heating wouldn't go over so well on a precision grid (can you
say warpage?) or chemically treated cathode.

Arc welding, again varying by type, is still fairly sharp, but I wouldn't
guess it would be good enough for most electrical purposes. MIG (GMAW) is
generally the fastest (splat some metal on, then run), while TIG (GTAW) is
the slowest (you can sit there cooking the workpiece with the arc if you
want). (Oxyacetylene is slower still, obviously because the flame has
lower temperature than an arc, plus all the gas blowing by heats up a lot
more work than just the weld zone.)

Friction and induction welding are pretty specialized, being useful on
shapes with continuous symmetry (e.g., butt-welding circular pipes with
friction spin welding; long seam welds with friction stir, or induction,
which is common for seamed pipe).

Forge welding I suppose is the silly case of "100% HAZ", because you
generally have to put the entire workpiece into the fire to get it hot
enough (somewhere around spitting white hot) before it'll mash together
when you whack it with the hammer. Friction and induction welding really
are examples of forge welding (the metal is not melted), but because they
involve localized rather than bulk heating, and apply to specialized
processes (you can't spin-weld just anything..), they can be separately
identified.

So, when it comes to electronics, spot welding is a great way to go, if
available -- it doesn't make much EMF (yet another concern!), since
current is applied across the wire; it doesn't heat things up, and it's
fast and cheap, great for production.

Ultrasonic welding and wedge bonding are pertinent examples also from the
electronics industry, being used for very small joints of course (where
heating cannot be tolerated, so spot welding is out).

Tim "spent years on RCM"

--
Seven Transistor Labs
Electrical Engineering Consultation
Website: http://seventransistorlabs.com

[Tom Del Rosso[_4_]]

Tim Williams wrote:
"Tom Del Rosso" wrote in message
...

Tim Williams wrote:
Spot welded? (I'll watch later)

But what temperature would that be at?

Well.. welding is welding, yes, but you get very different results
between highly localized and loosely generalized heating methods. In
order of HAZ (heat affected zone) size, smallest to largest: e-beam,
laser, spot, arc, torch, friction, induction, forge. (Give or take
various considerations for each process, but I think this is a pretty
typical ordering.)

Thank you. Friction is the one whose existence I learned of most recently.

I'm sure we can narrow it down to the first 3.


Ultrasonic welding and wedge bonding are pertinent examples also from
the electronics industry, being used for very small joints of course
(where heating cannot be tolerated, so spot welding is out).

I assumed even spot welding would have melted the solder 2mm away.

So how do you think those pins were welded by IBM in 1968?


Tim "spent years on RCM"

I spent a year there one day.


--

Reply in group, but if emailing remove the last word.

[legg]

On Sun, 27 Oct 2013 02:53:20 -0400, "Tom Del Rosso"
wrote:


Tom Del Rosso wrote:
Getting down into the details of the booster guidance computer from
IBM.
http://www.youtube.com/watch?v=vZhRb...em-uploademail

Pins are soldered to the hybrid substrate, and then the pins are welded to
the PC board. Somehow the welding did not melt the pins off the substrate.
(???)

Pins are brazed within the ceramic form. Internal bonds are
stitch-bond gold or Al wire to pad, high temperature solder between
substrate and sputter/print patterned ceramic.

Ceramic carrier is simply reflowed to printed wiring.

RL

[Tom Del Rosso[_4_]]

legg wrote:

Pins are brazed within the ceramic form. Internal bonds are
stitch-bond gold or Al wire to pad, high temperature solder between
substrate and sputter/print patterned ceramic.

Ceramic carrier is simply reflowed to printed wiring.

See the video. IBM welded the pins to the board without solder, but the
pins are soldered to the hybrid 2mm away.


--

Reply in group, but if emailing remove the last word.

[Jon Elson]

Tom Del Rosso wrote:


Tom Del Rosso wrote:
Getting down into the details of the booster guidance computer from
IBM.
http://www.youtube.com/watch?v=vZhRb...em-uploademail

Pins are soldered to the hybrid substrate, and then the pins are welded to
the PC board. Somehow the welding did not melt the pins off the
substrate. (???)

If this is the same technology as some other stitch-bond gear, it is
done by ultrasonic welding, and pretty fast. I've also seen
boards that were a bit like wire-wrap, the chips were welded
to fat, round pins that stuck above the board on both sides.
The pins were maybe .050" diameter. Then, on the other side,
the wires were ultrasonically welded to the pins, right through
the insulation.

Jon

[legg]

On Sun, 27 Oct 2013 11:38:13 -0400, "Tom Del Rosso"
wrote:


legg wrote:

Pins are brazed within the ceramic form. Internal bonds are
stitch-bond gold or Al wire to pad, high temperature solder between
substrate and sputter/print patterned ceramic.

Ceramic carrier is simply reflowed to printed wiring.

See the video. IBM welded the pins to the board without solder, but the
pins are soldered to the hybrid 2mm away.

Didn't watch the video. This machine doesn't even have sound.

You wouldn't solder pins to the ceramic without mech support. 'Modern'
commercial hybrids count on c-clip attachment for inline side attach -
older ones were brazed.

I suppose you could attach an IC to the board any way you pleased.
Spot welding would be good, if you could do it right every time and
had the budget.

The combination you suggest was used isn't impossible, or bad, it's
just uncoordinated. As an FRU, it's a throw-away item anyways, if it
shows signs of misbehavior.

RL

[Tom Del Rosso[_4_]]

legg wrote:
On Sun, 27 Oct 2013 11:38:13 -0400, "Tom Del Rosso"
wrote:

See the video. IBM welded the pins to the board without solder,
but the pins are soldered to the hybrid 2mm away.

Didn't watch the video. This machine doesn't even have sound.

You wouldn't solder pins to the ceramic without mech support. 'Modern'
commercial hybrids count on c-clip attachment for inline side attach -
older ones were brazed.

I suppose you could attach an IC to the board any way you pleased.
Spot welding would be good, if you could do it right every time and
had the budget.

The combination you suggest was used isn't impossible, or bad, it's
just uncoordinated. As an FRU, it's a throw-away item anyways, if it
shows signs of misbehavior.

These pins had c-clips on the edges of the hybrids. IBM used the latest
methods they had, in advance of using the same devices in the 360.


--

Reply in group, but if emailing remove the last word.

[dave]

On 10/28/2013 04:57 PM, legg wrote:
On Sun, 27 Oct 2013 11:38:13 -0400, "Tom Del Rosso"
wrote:


legg wrote:

Pins are brazed within the ceramic form. Internal bonds are
stitch-bond gold or Al wire to pad, high temperature solder between
substrate and sputter/print patterned ceramic.

Ceramic carrier is simply reflowed to printed wiring.

See the video. IBM welded the pins to the board without solder, but the
pins are soldered to the hybrid 2mm away.

Didn't watch the video. This machine doesn't even have sound.

You wouldn't solder pins to the ceramic without mech support. 'Modern'
commercial hybrids count on c-clip attachment for inline side attach -
older ones were brazed.

I suppose you could attach an IC to the board any way you pleased.
Spot welding would be good, if you could do it right every time and
had the budget.

The combination you suggest was used isn't impossible, or bad, it's
just uncoordinated. As an FRU, it's a throw-away item anyways, if it
shows signs of misbehavior.

RL


"Ceramic" is scary. Are you sure there's no Beryllium, etc.?

[[email protected]]

Geez. It's obvious she couldn't get good thermal contact because the solder all over the surface was oxidzed or whatever. Her antimony bearing or whatever solder never made thermal contact.

Even hear of a heat gun ?