RoboRio Ethernet Connection Through Hub

[seg9585]

We are planning to use multiple ethernet-connected devices on our robot this year, but since the new Wireless Bridge only includes 2 ports, we purchased an ethernet hub to expand the number of ports we can use.

We are currently using one of the Bridge's ports to connect to the roboRio, and the other to interface with the hub (which we can plug our other devices or tether into).
However, in our current design we want to mount the Bridge in an area that will maximize reception/connection strength, while the roboRio and switch are sitting next to eachother in another area of the robot. Rather than running two ethernet cables from that area to the Bridge, we proposed running the roboRio to the hub, then hub to Bridge with a single cable. That way, the second port on the Switch is very accessible for tethering at competition, and we make the wiring extremely efficient.

However, I noticed the rule:

R58 The Wireless Bridge must be connected to the roboRIO Ethernet port (either directly or via a CAT5
Ethernet pigtail).

Does this exclude our proposed setup? Technically our bridge is still connected to the roboRio, albeit through a hub, via CAT5 ethernet.

[Mark McLeod]

No, that layout wouldn't get past inspection (although you can try an appeal to the Q&A), and it really isn't a good idea to add more devices to such a mission critical path anyway.
Only connect the roboRIO and radio directly together, no intermediate network devices, and use the radio port closest to the radio power connector (just a strong suggestion).

[seg9585]

Hmm, okay, we may look for a way to relocate the hub close to the desired Bridge location instead (rather than close to the roboRio)

[GreyingJay]

Originally Posted by Mark McLeod

use the radio port closest to the radio power connector (just a strong suggestion).

What is the reasoning behind this suggestion?

[Al Skierkiewicz]

Eric,
I am going agree with Mark here. Your first priority is being able to drive the robot. Everything else is secondary and useless if the robot is not driving.
One thing that many teams ignore is the minimum bend radius on the cables they are using. When you violate this simple concept, the conductors inside the cable begin to migrate and degrade cable performance. Conductors tend to pull out of connectors as well. On cheap cables I have seen wire shorts when the cable is bent at less than rated minimums.

[Foster]

Originally Posted by GreyingJay

What is the reasoning behind this suggestion?

When you zip tie the two cables together for support, they start out more or less next to each other. If you use the other port, it's farther away and you end up with a sharper bend to your first zip tie.

FIRST is trying to eliminate points of failure in the primary systems. A length of working CAT5 cable is unlikely to stop working in a match. Two lengths and a powered switch have more possible fail points.

[GreyingJay]

Originally Posted by Foster

When you zip tie the two cables together for support, they start out more or less next to each other. If you use the other port, it's farther away and you end up with a sharper bend to your first zip tie.

Ah! Yes, that makes sense.

I was trying to think of electrical reasons why to favour one port over the other

Edit: see post below - didn't realize there were differences in the ports on the OpenMesh radio.

[Mark McLeod]

Tying the cables together securely is one reason.

The two ports on the OpenMesh are also setup differently, for instance the Open Mesh should be configured through the 802 port while the POE port handles power input.
From Beta testing and other Beta reports the POE port seemed to be less prone to roboRIO connection issues like the one's in some of the current CD threads.

In general, I recommend following FIRST's instructions, since they might add adaptations to the firmware, get all the bug reports, do their own confirmation testing, as well as have the manufacturers ear.
But if you encounter repetitive issues with one port, try the other.
Approximately one percent (made up statistic) of the radios might suffer from manufacturing oddities that benefit from wiring that one percent differently than what may be the standard FRC practice.