On 2011-01-10, Wild_Bill wrote:
The tool for sub-D terminals is available at those Shack stores, if the
store still stocks much in the way of parts.

The red color is the same for aviation terminal tools for the same size of
terminals, easy to remember if one is familiar with electronic
component/resistor color coding.. red (2) is for size 20 terminals.

Hmm ... note that the normal insertion/removal tools from AMP
for these (DB-25 sized) pins in the old days (pre TYCO) was two colored.
A red end contained the insertion tool -- a C-shaped tube of probably
bronze. The white end contained the extraction tool. It is a spring
steel (normal metal color), and while not fully cylindrical, it is almost
completely so. There is a very thin slot opposite the open end,
terminating in a small hole to stop stress cracks from forming. This
one is sprung open and slid over the wire before passing down into the
connector body to depress the locking tabs on the pin to allow it to be
removed. The red end has AMPs molded part number (91087-2), and the
white end has a stamped number M81969 -- probably a Mil-Spec inventory
number, along with a partial date /1-02.

Another which I use interchangeably with the above is quite
similar -- except for the lack of any part numbers, molded or stamped,
and slight difference in the shape of the plastic body (D cross-section
instead of square).

As mentioned in an earlier thread, there are two different terminal types
(and connector bodies) for sub-D connectors.. stamped-sheet brass formed
into the shape of mating M/F terminals, which have the dual flag tabs, and
the other type is the machined-from round stock into the mating M/F
terminals (much more expensive terminals and crimper).

And both can be used interchangeably in the connector bodies.

The formed brass sheet type terminals are common commercial-quality
terminals, found in many D-type computer cables (but also many other
applications).
These terminals are cheap to produce, and easily implemented with the proper
industrial assembly machines.

The machined type terminals are high-reliability types, which are far less
common in consumer goods.

Of course. Did you hear anyone suggesting that these be used in
the project in question?

Both of these types of terminals are extremely labor intensive and very
costly, compared to soldering.
Soldering is permanent and prone to less fabrication-related failures than
crimping, whether one is familiar with crimping or not (assuming that one is
capable of learning how to solder properly).

Interstingly, Ma Bell used to consider soldered joints to be
temporary, and wire-wrapped joints to be permanent. (And wire-wrapped
joints, like crimped joints, require special tooling.)

The vibration aspect keeps getting mentioned, and is definitely not much of
a significant issue with home shop machines.

Depends on the machine. My Nichols Horizontal mill, with 6"
conventional milling cutters makes a lot of noise and vibration under
certain circumstances. Granted, it is not a CNC machine, so if a joint
fails, the spindle motor stops -- nothing else. :-)

If the sub-D cable connector is installed with a connector shell, then the
soldered connection is already essentially immune to vibration issues, as
the shell will dampen cable vibration.

Depending on the frequency of the vibration. It can be tuned to
the wire length to cause a lot of vibration between the point of support
and the point of connection.

If various cables are so loosely installed that they're free to shake (or be
pulled on), then any failures are a result of poor installation..
free-hanging connections swinging in the wind are destined to fail anyway.

Of course.

The fact that military/aviation connectors are fabricated with methods to
guard against failures is obvious.
Home shop machines will not drop out of the sky if a connection fails.
Rarely will injury or deaths occur if communications are lost between a
control and a motion device (troops and intel analogy).

I would consider CNC machines (which is what this discussion is
about) to have some potential for injury or death -- especially if they
use servo motors. If one of the tach feedback wires parts, the motor
will suddenly speed up to maximum speed in one direction or the other.
This can be *very* fast compared to a stepper driven axis, and given the
leverage of ball screws, can apply a lot of force to a human pinned
between the table and some adjacent machine -- or between the spindle
and the workpiece.

Looking at the conditions that exist under the hood of a car, compared to a
stationary machine in a workshop, there are few similarities.
Yet, many connections in engine compartments aren't high-reliability, or
environmentally sealed connections.. and they generally hold up for many
years/decades.

Yes -- but the frequency range of the vibrations is well known,
and unsupported wire lengths which would be most subject to those
vibrations are normally avoided.

Many home shop machines will may even be modified/upgraded or sold within a
couple of years.

This can be an argument for crimped pin connectors -- it is
easier to move a connection to a different pin there.

Yes -- it can be done with soldering -- but every time you
unsolder and move the wire, you do more heat damage to the insulation,
ignoring the wicking of solder up into the wires, since you want to
ignore vibration failures anyway. And some connectors -- the ones with
white nylon bodies -- tend to be quite sensitive to the heat of repeated
unsoldering and resoldering of pins.

I understand the practicality and efficiency of doing something well, and
this is often the primary issue with most things I do.
Aerospace certified connections on a mill or other shop machine aren't
practical from any perspective.

Nobody was suggesting Aerospace certified connections -- e.g.
the machined pins. The stamped pins are fine for this application --
even better than in automotive use where spilled battery electrolyte, or
salt filled road splashing can get into the thin slots present in
formed pins and lead to corrosion -- not a problem in a typical home
machine shop.

[ ... ]

In most cases, soldered and insulated (shrinktubed) connections, properly
secured to a stationary surface, are just as reliable as any connector of
any cost.. practical, inexpensive, easy and effective.. wire markers
optional.
When the wiring requires changes, cut and repeat.. caution for any
tighter-than-a-frog's-ass types, this requires wasting a few inches of
solder and a small amount of shrinktubing.

When circuits need interconnection devices, and many don't, solder the wires
to the connector contacts (install connector shell or insulate well), and
secure the cables nearby.. the results are more reliable than many engine
compartment connections.

The project which started this discussion -- a rotary table
being CNC equipped for a mill -- is one which would not be installed on
the mill full time, so a connector does make sense here.

As for soldering the connections -- a lot of us are getting
older, and connectors are getting denser and denser. A three row
Subminiature-D connector (e.g. the typical VGA connector) requires good
vision and practiced skill to solder those pins without resulting in
solder blobs connecting adjacent pins, or soldering iron damage to the
insulation of the wires in question. And not all HSMs are experienced
electronic solderers. It is for these that the crimp pins are the
easier way to work. If you get a wire connected to the wrong pin in the
middle row of a well populated three-row connector -- with crimp pins,
you reach for the extraction/insertion tool, pop it out, and put it
where it belongs. With soldered pins -- you are likely to have to
unsolder several wires blocking access to the one needing change -- with
the increased chance of reconnecting *these* to the wrong pins.

And with printed circuit boards -- even the good glass-epoxy
ones -- each removal and resoldering increases the chances of floating
the pad away from the board -- or damaging nearby plated through
connections. If the board is an old phenolic PC board, the chances of
floating a pad free are vastly greater. They really were *not* made for
maintenance. :-)

Many times, the situation is matching a connector to an existing connector
(low voltage signal connections). If the existing connector doesn't need to
be disconnected regularly, then soldering wires to the device eliminates the
need for the connector. To disconnect, apply heat, remove.
This will work fine if the device is your own stuff, doesn't affect warranty
issues, resale value, or safety issues.

And is insufficiently dense to provide problems for the skill
level and vision of the person doing the work.

I've got good glasses, and a stereo zoom microscope for really
dense things. I've done repairs to boards with the high density surface
mount chips using these (and some 20 Ga wire wrapped around the tip of a
small soldering iron to make an even smaller tip at need). I still
prefer crimp connections for many things.

Enjoy,
DoN.

--
Remove oil spill source from e-mail
Email: | Voice (all times): (703) 938-4564
(too) near Washington D.C. | http://www.d-and-d.com/dnichols/DoN.html
--- Black Holes are where God is dividing by zero ---