Dead DS ethernet ports? Post here!

A HUGE problem at the Kansas City regional surfaced today - Dead ethernet ports on the Driver Station. A surprisingly large number of teams (including ours) had both ethernet ports on the DS spontaneously stop working. By surprisingly large, we were the 6th team (out of 61), and by the end of the day i was hearing numbers as high as 15. There was one loaner at the regional that was being passed around for teams when they were competing, but needless to say that sucks.

The Robettes want to give a big THANK YOU to team 525, who lent us their spare DS to compete and practice with.

Are the other regionals seeing this same problem?

190 blew out at least three Ethernet ports over 2 DS’s during robot testing pre-ship. We determined the culprit was a static shock to the switch on the enable-disable dongle. The switch has now been grounded to the case.

Had you grounded the DS according to the instructions prior to that? From what we could tell, most (if not all) of the teams with problems had their stations properly grounded. Additionally, with the extremely dry air in the stadium and the lack of much carpet, it doesn’t seem like a static shock to the dongle would be very likely, especially for so many teams.

One thought we just had was the practice arena - there is a little carpet around the edges here, and we all know those wheels when run on carpet can generate a ton of static electricity. Thinking hypothetically here (as we have no way to test this that i can think of), what if the robot generated some static and that somehow got transmitted down the ethernet cable (when you’re unplugging it it hits the frame, for example)? Shocks like that are why you wear a grounding bracelet when working on circuit boards…

Static Electricity is more of a problem in dry air, not less. Dry air acts as an insulator allowing a greater charge to buildup without bleeding off into the air.

The first one fried before the grounding modification, the second one that friend was grounded per the team update instructions. We instituted a precautionary protocol after the first one fried to always ground oneself before touching the DS in any way shape or form, but all it took was one slip to fry it again (even with the grounding modification).

Team 2228 lost 3 Ethernet ports on DS stations pre-season. Both DS stations had the grounding modification prior to going dead. The damage occurred in Tethered mode testing with the robot up on blocks I believe. Conveyor systems may be generating ESD. We did look at the box that had two ports fail and found that it appears that for each port the HALO isolation transformer has failed. There did not appear to be any ESD suppression on the design such as Transient Voltage Suppression diodes. Davidcom chip on the other side of the isolation transformer appeared to still be functional. We are working to assure we retain a common ground on both sides of the Ethernet cable with the addition of a wire with 1Mohm resistor tied to grounds on each side. We are also concerned that if damage is occurring on the DS side of the cable, are we damaging the other end? So far this does not appear to be a problem for us, and don’t see any indications on Chief Delphi. If the damage is limited to the isolation transformer and the DS unit is not warrantee replaced, the replacement part is readily available for $3. Surface mount repair on the 16 lead part and you might still have a unit useful for non-competition. Tough situation and we continue to encourage FIRST’s efforts to identify the cause and share way to avoid failures with teams.

Our RJ45 jack became loose and the jack physically broke off of the board right before the match started. The jack got caught inside of the DS so we were unable to switch the ethernet cable to the second port.We were dead the whole match. Luckily, the Human player right by us missed a majority of the ball, so we actually won the match. We borrowed a DS from another team for our final match.
As a word of caution, be careful when plugging in the cables.

Both of our ethernet ports also broke during the first day of competition. We had the robot tethered and were testing our conveyors and shooter which might have created some static (though it seems far fetched to me that this charge would be transfered through the tether cable without affecting the cRIO:confused: ).
If anybody has details on a fix for this, I’d love to get this working for an alternate controls system or something for non-official competition. I’m very surprised that this driver station is so sensitive to static shock and even more surprised that we can break the ethernet ports without damaging the rest of the unit.

That is some great information. The officials at the competition had a box open and i took a long hard look inside, and know exactly the part you’re referring to - it was my best guess as to the source of the problem, although the internet there wasn’t working at the time, so i couldn’t pull up a data sheet… If they don’t replace it under warranty, i’ll definitely be opening it us and taking a close look at that chip. It’s been quite a while since i’ve done any surface mount work though…

We’re not entirely sure if our Ethernet 1 port is completely gone or not. We had Comms drop on us twice in a row and it was due to that port. The straight thing is that it will connect and then with a slight jiggle lose the connection.

Our ETH2 port became loose and just barely attached to the board during week 5 of build season. It still worked though.

We contacted FIRST and sent them a picture. They sent for a replacement right away and asked us to send the broken one back for forensics. We’ve had no issues with the new one. (the old one actually hasn’t gone back yet due to logistical mishaps, but it should soon)

We had a bit of an oddity…

When in the pits, tethered to our robot, the both Ethernet ports on drivers station would work. However, every time that we went to a match, the drivers station’s Ethernet ports would not work with the arena. We used the spare and that worked with our robot in the arena.

…Then we would go back to the pit and it would work again with the tether! AGHHHHH!

We had 2 Texas Instruments employees look at our problem and none of them could figure anything out.

We just lost our second ethernet port like 190, we have since grounded the switch. I think I read a post about how to repair the DS after the ports fried, does anyone have a link. The rough part about this whole thing is we really don’t want to buy a new DS when hopefully FIRST will be fixing these issues for next year. We lost our first port at BAE on the practice field.

2228 lost 3 DS Ethernet ports pre-season when the DS ground update was in place. We have since implemented the grounding of the disable switch. We added a point to our electrical panel that connects to the CRio chassis, and a point to the Operator Panel that connects to the DS chassis. We made a discharge cable with 1 MegaOhm resistor that we always clip/connect to the CRio chassis and DS chassis prior to plugging in the Ethernet cable. We leave the discharge cable connected at all times during tethered. Since implementing this hardware and discipline we have not experience a single DS ethernet port failure. Humidity levels are now higher than earlier in the season, so this may not be definitive but it does look promising.

We have been noticing that on ours almost from week one, I did find that one of our ethernet cables was bad (flex it and it would work), replacing that helped but did not solve all of it.

We are also having problems with both of our ethernet ports, what happens is you plug the ethernet into either port and power on the DS, and you get no communitation ( the LED doesn’t light up ).

A fix for us is once powered on and booted up, switch the cable to the other port while the DS is still on, and then power cycle, this seems to work for us every time.

Like I said this works for us, but i guess its worth a try.

This happened to us at NYC. The problem was we did not have the driver station at the latest release. Worked tethered but not on the field. Problem resolved before matches began.

At DC Regional we lost port 2 on one DS and port 1 on our spare DS. Link light would continue to flash in appropriate manner on port 2 of first DS, but DS could not be pinged from PC when connected to PC via port 2. Second DS failure link light would stay on even when cable was removed. Using an ethernet switch allowed us to continue to program the robot with a single port failure. Wire DS to switch through good port, PC to switch, and cRIO to switch.

Before the first DS failed completely, we had been having difficulty staying connected in debug mode.

Several other teams lost one port. One team lost both. DS does not appear to be robustly designed. I have heard of no cRIO network port failures or PC network port failures, only the DS seems to be susceptible.