A Dallas team reported unbearable vexnet reception.
When I looked at their robot,  the Cortex/Vexnet key was mounted directly under a motor, which gives out strong EM fields and noise from the DC motor brushes.    
  Once they used a 7-18" USB extension, their problems went away.

Analogy: put your TV antenna (or cortex vexnet key) on the roof, not in the basement.
Quote 0 0
Thanks for the advice.  We had one round with a dropped connection that fortunately reconnected.  The cortex is located within 2 inches of 3 motors.  I will suggest that our team accommodate a short usb extension.

We had another round with NO connection after the first 15 seconds.  This second issue seemed to be a loose usb connection on the cortex as it was solved by taping the vex key in place. (old wifi keys and tether also cut out.)  We are switching to a different cortex until that problem is resolved.
Quote 0 0
Are USB extension cables even legal?
Our hub didn't let us use ours (mind you it was about three feet long).
Quote 0 0
"In the Rules"
1 18" (or less) USB extension cable (between Cortex and VEXnet key) Optional5 monoprice 5431

Quote 0 0
We also had problems with interference. As we were driving the wheels would jiggle and we did notice a lights indicating lack of communication. Oddly, when the VEX joystick was tapped by one of our players, not too lightly I might add, connection problems were resolved. We have checked motor wires for interference but all connections look good and wires, while close, do not have any uninsulated parts touching.

Any ideas?
Quote 0 0
EM interference does not require uninsulated parts touching, but sparking between electric wires is a source of EM interference.
The BEST DC motors have commutator brushes, so they generate lots of EM, although much of it is shielded by the metal can body.

Dallas BEST Game day had many reported issue of dropping Vexnet links, among teams with low practice time.
Dropped links often "disappeared" after taping in the Cortex key, but that may be chance as well.
There were no dropouts in finals, and my impression was that all the finals teams had 1.5 weeks of driving practice, or more, so they had ample time to experience, fix, etc their setup.
  If I could find a robot vexnet that was reliably unreliable, I've love to test the "game mode" setup (9v backup battery plus dongle in Joystick).

Quote 0 0
Another design that I observed here in a local Florence, AL competition had a neat outside skin of metal.  It looked cool, but as an old retired electrical engineer, when I saw it I wondered if the communication was reliable (due to the effective Faraday Cage that completely enclosed their Cortex controller).  Sure enough, during a critical phase of the contest, that robot lost communications with the joystick.  Later I found that they had did most of their testing and training without the outer metal skin, only adding that for the competition.

I guess the lesson learned here is to not build a shield around your antenna, and to always test and train with the final form of the robot.
Quote 0 0
Our Vexnet problems did not occur at our local hub but showed up at South's BEST in Auburn.  See this post where WiFi interference was ruled the culprit.  I'm hoping this is addressed for the upcoming season.

Quote 0 0
KevinH, about your comment:

"I never heard it mentioned that interference might be in placement of the VEX and key to the motors as suggested in the Texas post."

It is true that massive amounts of high-frequency communications (frequencies such as Wi-Fi on the same 2.4 GHz band) could interfere with the Vexnet wireless communications from robot to joystick.  But you have little control over that problem, except to make your objections known (as you did).

The thing you can control is the environment for your Vexnet keys.  The "cleaner" your path is between devices, the less likely you will be to lose communications.

It is my understanding that the Vexnet keys contain the transmit & receive antennas for the 2.4 Giga-Hertz wireless communications for the Cortex controllers and joysticks.  Any metal structures on your robot that block the Vexnet key could cause potential disconnects or add to the probability of communications loss.  The larger, thicker, and more massive the metal shielding structure, the more likely there will be loss of communication (as with any radio-type communications).  In competitions the joystick will normally be held while standing, so to avoid obstructions the best location for your Vexnet key becomes the top (or near the top) of your robot.

To maximize your signal strength, don't place a metal box or other metal parts around your Vexnet key, and locate the 7.2-volt battery (a relatively massive lump of metal and metallic compounds) as far from the Vexnet key as possible, and below it. 

During competitions, to maximize signal strength, an effort should be made to keep the joystick Vexnet key pointed toward your robot, and avoid blocking with the joystick battery compartment (avoid having the top of your joystick facing toward your robot).
Quote 0 0