"No Comms"

[rrossbach]

Originally Posted by pitzoid

Update:

Sent a bunch of test data to FIRST and DEKA yesterday confirming the DS Network chip susceptability to EMP. Today they were able to repeat my results, so it seems confirmed. We can't fix the DS easily, but we're working on a bunch of "workarounds". FIRST is looking into ways to eliminate static generation, which if they can will fix the source of the issue and I've been testing some FMS network mods today that enable the DS to do a minimal association with the the FMS network. I.e. now when the DS network chip resets on an EMP event, I can have it talking with both FMS and the robot in under 2 seconds. If the DS completely reboots its a few more seconds but it still comes back fast. Not the ideal situation, but it should help. Hopefully we won't have any static issues this weekend.

Oh, and the added extra benefit of the network mods? Looks like we dropped about 20 seconds off the initial DS/robot/fms linkup before a match starts

Real engineering: see a problem, analyze the data, if you can't "fix it", find the work around

Great news! Hopefully even if there are any issues in future competitions it won't impact teams & their alliance partners as much as in the past. Thanks!!!

Ron

[DRH2o]

Would it be helpful to put a small piece of wire rope dragging the floor to discharge the static. They have done this on carts at stores to solve static problems there.

[pitzoid]

Originally Posted by DRH2o

Would it be helpful to put a small piece of wire rope dragging the floor to discharge the static. They have done this on carts at stores to solve static problems there.

Might help teams to have a ground strap on the bots, they were seeing static issues at Boston today and wetted the carpet which seems to help. Boston in on an Ice Rink, I understand NH was too, so it seems Ice Rinks and Lunacy are challendges. Probably due to the ice pulling all the humidity close to ground level out. Humidifiers would help, just need to keep the humidity at floor level up.

[ezygmont708]

Originally Posted by squirrel

I was wondering if this would be a problem. Also does the orientation of the gateway seem to make any difference? we had some problems during testing before ship, and ended up putting the thing on a top surface of the robot.



Is that "new" update the one that has been on the FIRST site since Feb 10th or so?

http://usfirst.org/community/frc/content.aspx?id=10934

we found it the day before crating up the robot, it kind of messed up our plans for autnomous control.

A few things about the wireless communication system. There are a few things that will limit your tx/rx between the driver station and the wireless gaming adapter...

1) Gaming Adapter should, at no time, be blocked by a piece of metal on any side. I have seen teams that have mounted the adapter on a sheet of metal more than 1/8" thick.

2) Try to keep the adapter away from motors. Motors generate a magnetic field, which at times can interfere with 802.11x communications.

3) Ensure that you are using the correct cable type, and that each cable is properly seated and secured. Ethernet cables are definately more durable than a PWM, but they are suseptible to Electro-Magnetic interference. I have actually ordered some shielded Twisted pair for our next regional.

Being a network technician, I did a simple site survey at the trenton regional, while the fields were not in use. I found a plethora of 2.4Ghz communication. Much of it was the field communication system, but at trenton there was a fairly strong wireless signal active (Trenton Devil's Press Box). Another fairly important fact.... While bluetooth signals are not very strong they can cause interference with b/g/n networks.

Just my 2 cents...

[Dave Flowerday]

Originally Posted by ezygmont708

I have actually ordered some shielded Twisted pair for our next regional.

Using STP is against the rules. See R57:

Originally Posted by R57

The signal output from the wireless bridge must be connected directly to the cRIO Mobile Device Controller using one of the Ethernet cables provided in the 2009 Kit Of Parts.

Originally Posted by ezygmont708

I found a plethora of 2.4Ghz communication. Much of it was the field communication system, but at trenton there was a fairly strong wireless signal active (Trenton Devil's Press Box). Another fairly important fact.... While bluetooth signals are not very strong they can cause interference with b/g/n networks.

The control system this year uses 5GHz 802.11n exclusively, so you won't find any field communications in the 2.4 band. Additionally, Bluetooth interference should be a non-issue since it also uses the 2.4GHz band.

[pitzoid]

Originally Posted by Dave Flowerday

The control system this year uses 5GHz 802.11n exclusively, so you won't find any field communications in the 2.4 band.

Actually, the field does use 2.4Ghz, just not for the robots.

[ezygmont708]

Originally Posted by Dave Flowerday

Using STP is against the rules. See R57:



The control system this year uses 5GHz 802.11n exclusively, so you won't find any field communications in the 2.4 band. Additionally, Bluetooth interference should be a non-issue since it also uses the 2.4GHz band.

I was not aware of the rule regarding shielded ethernet cable. Thank You for the notice in advance.

As for the wireless communication, I know that I was seeing massive comms across the 2.4 Ghz network, and I guess it was possible that it was the arena's wireless, but coverage across all 11 channels in the 2.4 Ghz range is abnormal, as one would normally try to keep it seperated out to three non-overlapping channels (Normally 1/6/11). As with rule R57 bluetooth is not permitted on the robot, that is very smart on FIRST's Part.

Last year, while reviewing issues on a wireless system, we noticed an issue with bluetooth clients that were crowding the network, and therefore providing interference across the spectrum. Microwaves, and HVAC motors have been shown to pose similar problems.

I still highly recommend moving the adapter way from any large metal plates, and keeping them away from motors (within reason (5-8"))...

[pitzoid]

The FRC arena has an intelligent Access Point (Cisco 1252), when it starts up it scans the entire spectrum and then tries to establish itself on the least interferring channel. You can force it to a channel, but right now on the road they are all in defualt, which means they'll do this. When we get to Atlanta, I'll have to force them to a certain channel for coordination with 5 arenas.

[Mentor007]

We experienced no comms during competition and tracked it down to one Fisher Price motor being driven by a Jaguar. The WGA was mounted high and well away from the motor however the Jaguar was only a few inches off the CRio near the ethernet port. We kept getting comms drop outs even when tethered. When we disabled the motor, problem went away. We then also ran the motor but at lower PWM levels and the problem went away. At high PWM values the dropouts came back. We swapped the Jaguar for another without any impact. We swapped the Jaguar for a Victor without any impact. We are implementing a rerouting of all wiring to keep all power lines away from the CRio. Putting EMI suppression cap on the FP motor to the extent legal. The problem did not crop up until elimination rounds, so perhaps brush wear or contamination had an impact. The above troubleshooting strategy seems to have pretty definitively pointed the FP motor's brush noise out as the source. Better electrical board layout may have helped to prevent the failure. Hope this can help someone else. Thanks to Eric of 1511 for his help.

[Al Skierkiewicz]

Please check the Q&A http://forums.usfirst.org/showthread.php?t=12228
Dragging wires and other attempts at fixes to esd issues are not to be used while First researches the problems. Until further notice, please standby.

[Darkcrosbone]

at our regionals there was a static electricity problem that would make the robots stop talking to the DS except when you had pneumatics

[pitzoid]

Originally Posted by Mentor007

Putting EMI suppression cap on the FP motor to the extent legal.

I've talked to FIRST for years about this. We went through a painful phase in robot combat where there would be all sorts of radio drop out issues due to noisy motors and it could be fixed with supression caps across the motor leads. I actually think this should be mandated for the KOP motors as most of them are extremely "inexpensive" motors that will create alot of brush noise EMP. Some of the better designed robots I've seen the last few years as the head mechanical Judge at the PHX regional were the ones that would seperate the control electronics (in this case it'd be the cRio, WGA and breakout boards) from the motor controllers and motors by some sort of sheilding screen or bulkhead. But this is a consideration with initial design.

I'm in the camp right now that bots should have a drag chain, I don't know what the hold up is on that, ultimately its FIRST's decision. I know they are contemplating putting drag chains on the trailers for this weekend, I'd prefer that they give teams the option to put them on their bots as we've proved that the KOP wheel from this year is very efficient at isolating static voltage buildup.

On another note: FMS ran well this past weekend, we had a couple equipment issues with Estop that were manufactured incorrectly. But most other issues seen were related to robots and ESD.

Hopefully, tomorrow we'll have some resolution in the team update.