How to prevent broken ethernet port on DS

[Bharat Nain]

After having significant amount of trouble with this years control system, we discovered that the ethernet ports on the Driver Station were obviously not designed for a FIRST Robotics environment. Personally, I am extremely disappointed and frustrated with the control system, especially the Driver Station. We have noticed an incredible number of teams having broken driver station ethernet ports. Our theory is it is because of static electricity but that is unconfirmed. It is certainly not a competition worthy equipment and it would be a great demonstration of Gracious Professionalism on behalf of FIRST to never give us such equipment again. Rambling aside, here is the solution.



1) Buy a decent SWITCH. We bought the Linksys 5-port Switch. Link (http://www.linksysbycisco.com/US/en/products/EZXS55W). Apparently, a HUB will not work.

2) Plug one ethernet cable from the Driver station (Any port) to any port on the SWITCH except the Uplink.

3) Plug the laptop and cRio using a Crossover/Ethernet into any open ports on the SWITCH except the Uplink.
This method was requested to be tested by a couple people including the NI guy (Joshua) at NJ. If anything will die, it can be your $10 - $50 SWITCH as opposed to your $162 + Shipping & Tax Driver Station.

Note: This will also work if you only have one working ethernet port on your Driver Station.

This system works for us and I hope it does the same for you. It is one less thing to worry about breaking at the competition.

Cheers and write to FIRST and providing your feedback if your Driver Station has/had problems.

Bharat Nain

[nathanww]

I might be mistaken, but isn't having anything sitting between the FMS and the DS forbidden at the competition?

[artdutra04]

I might be mistaken, but isn't having anything sitting between the FMS and the DS forbidden at the competition?The problems are arising on the practice field, where the robot is tethered to the driver station.

[EHaskins]

1) Buy a decent SWITCH. We bought the Linksys 5-port Switch. Link. Apparently, a HUB will not work.



A hub will work. I have tested it. Under most circumstances switches offer better performance, but with the amount of data being transmitted and the number of devices it will not cause any issues.

BTW, I've been using a switch I got a Goodwill for $0.50. ;)

[EricVanWyk]

Some very cheap hubs do not offer ESD protection.

[byteit101]

Someone on our team found out that most of the people with problems on the eth ports were using cat5, and people with no problem were using cat6, which has better shielding and grounding. Our team is ordering a switch and cat6's.

[rsisk]

Quote:
1) Buy a decent SWITCH. We bought the Linksys 5-port Switch. Link. Apparently, a HUB will not work.

2) Plug one ethernet cable from the Driver station (Any port) to any port on the SWITCH except the Uplink.

3) Plug the laptop and cRio using a Crossover/Ethernet into any open ports on the SWITCH except the Uplink.

Question: How do you power the Switch when you are on the practice field? Do you wire it into the cRIO with the router? Do you have some other source of power for it?

[Alexa Stott]

Question: How do you power the Switch when you are on the practice field? Do you wire it into the cRIO with the router? Do you have some other source of power for it?

Like the driver station, the switch will also require an external power supply. We brought along a power strip. If you have a decent inverter, you can use one of your 12v batteries and power it from that.

[charrisTTI]

Over the past 30 years, I have never seen a network card, hub, switch, or router fail due to static discharge. I have seen plenty of static discharges, which have done plenty of damage to other components, but the ethernet connections seem to always survive. This years DS being the exception.

[EricVanWyk]

I played with a switch and a 30kV ESD gun. I recommend using a switch to isolate your DS from your robot while in the pit.

BTW - 12kV to the finger is an experience not worth repeating, at least not with my own finger.

[Bharat Nain]

Over the past 30 years, I have never seen a network card, hub, switch, or router fail due to static discharge. I have seen plenty of static discharges, which have done plenty of damage to other components, but the ethernet connections seem to always survive. This years DS being the exception.

Just as a note, one of the ports on our linksys 5-port switch did die due to ESD towards the end of NJ. Luckily, we could just plug this into another port and it worked fine.

[Kevin Sevcik]

Someone on our team found out that most of the people with problems on the eth ports were using cat5, and people with no problem were using cat6, which has better shielding and grounding. Our team is ordering a switch and cat6's.
Garden variety Cat6 is highly unlikely to have better grounding or shielding than Cat5. I can categorically declare this because most common Cat6 and Cat5 cables are Unshielded Twisted Pair. You can get it in Shielded Twisted Pair, but it's much harder to find and isn't going to make a difference unless you're also using shielded connectors.

The primary difference between Cat6 and Cat5, is that Cat6 meets higher standards for crosstalk and various things, and usually has a higher gauge wire. These things may make a difference in ESD protection, but any ESD protection isn't due to increased shielding.

[Chris_Elston]

We are not having a problem with static, but rather the port itself needs to be mounted to the board more firmly. What I saw was there are only two plastic post that are through holes to the bottom side of the board. When you insert the cat5 cable repeatedly, you seem to cause stress on these anchor posts. When these anchor posts give or break loose, your in trouble because then the connections in the back tip up and break off the pads on the surface mount. You can attempt a trace repair. It's quite alot of work but I believe if that is the only thing wrong, it can be done. Since we had the cover off, we went ahead and put hot clue on the second working port along the sides and the posts on the bottom to give it extra strength. As well as the advice already given, don't take the cat5 cable in and out unless you have to. Take the other end out. You can see from my pictures just how bad the trace pads lift. I lit the board from the bottom side so you can see the traces and how tiny and fragile the trace wires are.

1. So my advice if you have the cover off, hot glue the working ethernet port for extra strength.

2. Don't take the cable in and out unless you have too.

3. Order a spare DS anyway, because you don't want a $162.00 part causing an end to your robotics season.

4. Configure your spare ready to go just in case something bad happens at the competition and get it firmware upgraded and ready to drop in and go.

[Mr. Lim]

We are not having a problem with static, but rather the port itself needs to be mounted to the board more firmly. What I saw was there are only two plastic post that are through holes to the bottom side of the board. When you insert the cat5 cable repeatedly, you seem to cause stress on these anchor posts. When these anchor posts give or break loose, your in trouble because then the connections in the back tip up and break off the pads on the surface mount.

We've also run into physical port damage through daily wear-and-tear as well. We've had traces lifted as well as the plastic posts wear and release from the board, leaving the connector to essentially float on the leads alone. Eventually the leads bend or lift the traces off the board.

Beyond connecting and disconnecting cables from the DS once or twice a day, there hasn't been any unusual or unreasonable stress on the connectors.

Curious to know if others are also encountering physical wear and tear issues with the DS ethernet ports? Some early indicators are the port having more freedom of movement in the up/down direction - or when you insert a cable, the entire port seems to flex backwards and upwards a bit.

[Travis Hoffman]

I wonder if anyone would take issue with combining something like this (http://www.trianglecables.com/cat5e-001.html) and something like this (http://www.deepsurplus.com/Shielded-Ethernet-CAT5e-RJ45-Inline-Adapter-Coupler) such that you'd NEVER have to remove the cable from your DS at a competition....

[Chris_Elston]

I wonder if anyone would take issue with combining something like this (http://www.trianglecables.com/cat5e-001.html) and something like this (http://www.deepsurplus.com/Shielded-Ethernet-CAT5e-RJ45-Inline-Adapter-Coupler) such that you'd NEVER have to remove the cable from your DS at a competition....

Nice FIX, Travis. I think the more you can prevent the cable coming in and out the better. Great idea.

[pitzoid]

For people having issues with blowing out and breaking RJ45 ports when tethered to the bot, this might help along with some of the other things being talked about here

Surge Protection (http://www.tripplite.com/EN/products/model.cfm?txtModelID=151)

We were considering putting them on the field when the DS issues started, but they don't help the EMP events, the DSes still reset.... But it will help with tethered bots.....

[Peter Randall]

We configured a short extension for our ethernet ports after blowing one in Trenton in week one. We were asked to disconnect them from our DS when we were in competition last weekend in Pittsburgh. The rules clearly say nothing between DS and Field System. We blew another port and had to rely on a borrowed DS to compete. Not happy with the DS.

PGR

[Chris_Elston]

For people having issues with blowing out and breaking RJ45 ports when tethered to the bot, this might help along with some of the other things being talked about here

Surge Protection (http://www.tripplite.com/EN/products/model.cfm?txtModelID=151)

We were considering putting them on the field when the DS issues started, but they don't help the EMP events, the DSes still reset.... But it will help with tethered bots.....


Just placed an order for one at Newegg.com we'll try it out at Boilermaker and see how it works. Thanks for the tips!

[Travis Hoffman]

I've just been reminded that my earlier suggestion comes in a unified product called a "port saver (http://www.l-com.com/item.aspx?id=2541)". I should have recalled this, as I purchased a very similar crossover adapter from this site a while back.

[pitzoid]

Just placed an order for one at Newegg.com we'll try it out at Boilermaker and see how it works. Thanks for the tips!

Don't take it out to the official field, but I'm sure it'll help in the pits and on the practice field.

We ran a bunch of testing on the fields having to do with if the ethernet connection from the SCCE (field end FMS cabinet) to the DS was causing issues and it wasn't that (besides the physical breakage you guys are seeing). They don't allow you to have anything between the DS and the SCCE because they don't want you injecting anything besides the DS to cRio comm string in the system.

I don't like the DSes either (had nothing to do with its development....), we've been asked to work on a solution for it though.

[Travis Hoffman]

I have it on good authority that FIRST has been petitioned to permit passive "port saver" devices. One hopes a response either way would be forthcoming prior to next week's regionals.

Requiring the port saver to be removed before coming to the field would yield no benefits over the current "plug the DS into a switch in our pit" methods (other than being cheaper than a $20 switch). Presuming you tether for a function check before each match (if you don't, you should!), you would still have to initiate one engage/disengage cycle at DS Ethernet Port 1 in the pit and one at the alliance station for each match.

I feel the benefits of an "installed all the time" port saver far outweigh the costs. I would suggest FIRST allow any teams wishing to use these devices to do so on Thursday at Week 4 regionals and let FTA's and such determine if there are any adverse effects on comms (I doubt it). If no ill effects are observed, then I feel teams should be permitted to use these devices on Friday/Saturday.

Finally, I might still elect to use the 1 ft. Ethernet cable to female/female adapter solution instead of the port saver adapter installed right at the DS ETH1 port - the cabled solution would further minimize the risk of unintentional forces being transmitted to the reportedly weak circuit board connections of the DS ports relative to a rigid "port saver" that someone might still jiggle a bit in connecting/disconnecting cables.

Next question - would such devices also be legal for installation at the cRIO ethernet port 1? I'm sure that port is much more "ruggedized" than the DS ports, but it would still be nice to have the option of protecting our $750, forced-to-reuse-over-multiple-seasons investment.

[Chris_Elston]

Well Bob has a point though, use it in the pits even if it's not permitted at on the field, that just furthers the life and wear and tear in the pits and only a few in and out clicks at the field. ***what maybe 11-15 matches?*** So until the rules changes, we won't use the "port saver" device AKA 1 foot extension or the Triplit surge at the field, but we'll use it in the pits while we are tethered just to try and be safe and careful with the port as long as we can. It's a first time product, "they" who ever that is, did a good job designing it for the first time. Maybe next years revision will have some improvements or suggestions from everyone would be great.

[EricVanWyk]

"They" is KwikByte (http://kwikbyte.com/)

As Bob said, the rule is there for a good reason. However, that reason is not valid for these sort of 'port saver' parts - they do nothing to the traffic, and can not change it nor read it. Perhaps it is worth a Q&A to ask for an exception.

[Al Skierkiewicz]

At least temporarily the rule is
<R85> During competition MATCHES, the competition cable at the Alliance Station must connect
directly to the competition port on the Driver Station. The competition Ethernet cable must
connect directly to an Ethernet port on the Driver Station. No intermediate connectors,
cables, or “pigtails” are permitted. Only the Driver Station may connect to the competition
cables – no direct connection of team-provided portable computers, PDAs, or alternate
devices is permitted

Emphasis is mine. If the GDC were to ask me, I would respond that it is better to save the DS with a short pigtail with suitable strain relief than to see the rampant failure of the port connectors that many teams have thus far experienced. I would rather add this jumper than to have a team not compete due to a component failure that could have been avoided.

[adman]

Well Bob has a point though, use it in the pits even if it's not permitted at on the field, that just furthers the life and wear and tear in the pits and only a few in and out clicks at the field. ***what maybe 11-15 matches?*** So until the rules changes, we won't use the "port saver" device AKA 1 foot extension or the Triplit surge at the field, but we'll use it in the pits while we are tethered just to try and be safe and careful with the port as long as we can. It's a first time product, "they" who ever that is, did a good job designing it for the first time. Maybe next years revision will have some improvements or suggestions from everyone would be great.


Al, Chris

If the teams use the "pigtail" 1 foot extension with one of those 1 x 1 adhesive backed wire tie attachers on their driverstation it would be obvious to the casual observer that no malicious intent "other than staying in
a working condition" would be the purpose. I hope GDC lets go on this one
so the teams have a fix for such a high stress point wear area .

[Travis Hoffman]

"They" is KwikByte (http://kwikbyte.com/)

As Bob said, the rule is there for a good reason. However, that reason is not valid for these sort of 'port saver' parts - they do nothing to the traffic, and can not change it nor read it. Perhaps it is worth a Q&A to ask for an exception.

I submitted one.

[heydowns]

"They" is KwikByte (http://kwikbyte.com/)

As Bob said, the rule is there for a good reason. However, that reason is not valid for these sort of 'port saver' parts - they do nothing to the traffic, and can not change it nor read it. Perhaps it is worth a Q&A to ask for an exception.

The other reason often cited for this rule (with fairly good reason) is that intermediate connections make it harder for field personnel to diagnose any DS<>Robot link issues.
However, if an exception were allowed for passive devices/pigtails and those devices/pigtails were visible and easily removed, I don't think there would be a huge issue with this either.

[Al Skierkiewicz]

However, if an exception were allowed for passive devices/pigtails and those devices/pigtails were visible and easily removed, I don't think there would be a huge issue with this either.

Agreed,
I would think that the GDC might rule as they have in the past, this little addition could be legally made but any failure related to connection issues would lie solely with the team not the FMS.

[Andy Brockway]

Agreed,
I would think that the GDC might rule as they have in the past, this little addition could be legally made but any failure related to connection issues would lie solely with the team not the FMS.

Al called it! See update 18.

http://usfirst.org/community/frc/content.aspx?id=450

[Chris_Elston]

Al called it! See update 18.

http://usfirst.org/community/frc/content.aspx?id=450

Thanks Dave Lavery for monitoring CD and feeding information to the GDC. Happy to see this one. I known next year there will be improvements on the driver station.

[Al Skierkiewicz]

I must be psychedelic!

[Alan Anderson]

I must be psychedelic!

On Wildstang, I don't think there's any way you can avoid it.