Victor-PWM Cable connections

[eugenebrooks]

It is indeed the case that we are having a lot of problems with intermittent connections on the victors, and presumably the spikes as well (our team is not using them this year). A new cable does not work, you take it out and inspect the crimps and it is perfect, you plug it back in and it may work, or it may not. Get another cable, it works, and then if the cable is moved or replugged, you find it does not work.

We are finding that the fit of the metal pins on the cable into the corresponding socket on the victor or spike is loose when this problem happens. In addition to all the recommendations above with regard to proper tie down of the cable to prevent movement, you can spread the metal pin a little in order to cause the socket to get a better grip and produce a little insertion force.

If you look at the metal pins, they are made of stamped metal and have the shape of a U. If you can find a suitable tool (a wire cutter that has a flat side on one of its yaws works, or a small end nipper with one jaw ground flat will work), you can spread the U just a few thousandths and as a result you will note that the socket will increase the previously non-existant insertion force. At this point the connection should not be intermittent. It is important that you don't overdo this, just a spread of a few thousandths will produce a little insertion force and then the gold plated contacts will do their job for you. You can measure the spread with a dial indicator to make sure that you are not overdoing it, you don't want the spread pin to force the socket contacts past their spring limits.

It would be nice if IFI would use "high insertion force" sockets when victors and spikes are made, or carefully screen the components that they use for this critical, and very problematic, connection. The high insertion force sockets and pins I buy from Mouser have a good grip for a single pin, so it would seem that a good grip could be obtained with the socket in the victors and spikes by using high insertion force parts.

With a little insertion force on the socket, you will find that the connection will have less tendency to walk out and become unplugged.

Have fun,
Eugene

Originally Posted by Donut

This is a year old thread, but the problem for us is arising this year.

We have had extremely bad unseating issues with some of our spike relays on this year's robot. Sometimes there's no connection, sometimes there's only a connection for forward or reverse, and sometimes they work fine. Do any teams have methods of attaching PWM cables to spikes better, or know of any "alligator clips" that will work with Spike Relays?

[eugenebrooks]

Originally Posted by RbtGal1351

Hello all,

I am wondering what fixes teams have come up with to ensure a good connection between the Victor Speed Controllers and its PWM Cable.

The wadded up extra length of the pwm cable can present an extra
bit of mass that wiggles the cable when the robot moves. We cut
our PWM cables to the required length, replacing the female end with
a Molex C-grid connector, using the "high force" sockets. We don't
use the C-grid connector on the male end, the pins seem to be too
short to make a good connection to the victor/spike.

The cable itself is zip tied to the electronics board right next to the
controller, or the victor, so that the cable comes up and over with a
inverted U shape, providing for stress relieve in the event of impact.
Attaching the electronics so that the force of typical impacts is
perpendicular to the connector, removing the tendency of the impact
to pop it out, is a good idea.

We check all of these connectors between every match to make sure
that none are walking off, and we keep track of any that might be
moving. If it moves a half millimeter when you push a little with your
fingernail, it is walking off and you need to do something about it.

There is no substitute for checking the connectors between every match,
and laying out the electronics to make it easy to do this without bumping
wires, but you can take measures to improve things:

* You can make a bracket that pins the connectors down when
it is screwed to the electronics board. With due attention to the
possibility of a short, the 3 dollar clip available at robot marketplace
would seem to work for a victor, but you need a single bracket to
hold a large number of connectors on the RC. I would want these
brackets to be made of plastic to reduce the chance of a short,
but the aluminum ones on our robot this year were insulated
with electrical tape. If you have th luxury of using every other
PWM output on the controller, you can make a "comb" style of
bracket that will hold the connectors better.

* Another strategy is to glue the female connectors to each other
as you install them on the RC, producing a single connector with
a large pin count. The resulting high pin count connector is almost
impossible to walk off, although you can get if off if you need to.
The drawback here is that if you need to replace one cable for some
reason you get to do some cutting, but it produces a connector that
does not walk off without using a bracket. We use this strategy for
the array of switches used to provide autonomous modes, or inhibit
motors for safety in the pits, that are connected with a single multi-
conductor cable.

[Rickertsen2]

Overall, PWM cables seem to be the #1 source of electrical system problems. .1" headers are NOT designed for this type of application. They are however the standard for RC control. I prey that they will change during the next IFI design cycle.

As far as fixes, our team hot glues all of our pwm cables in place. Hot glue holds well enough to keep them where they belong, but also breaks away cleanly if you ever need to change something. We also use electrical tape to make a single connector our of the RC side connections.