Problems With FCS and Samantha

[Team230]

Hi,

This year our team has frequently been having problems connecting to our robot via the Samantha box. We've diagnosed the problem to a point, but can't figure out what to do.

The Samantha is correctly connecting to the network, and our laptop is as well. We have confirmed this by being able to access the Samantha's information webpage by entering the IP from samostat. However, we have been unable to connect to the laptop over both RobotC and the Field Control System. We've tried to reset the system numerous times (nearing the 50s...), and have tried multiple laptops, NXT brings, and Samantha modules. We will occasionally be able to connect, and are able to use the system fine, but when we try to set everything up again the next day we experience the same problems. Our brick simply doesn't show up at all in the FCS Choose menu. We are using a dedicated router without internet access with an Open WEP password.

Does anyone have any ideas? We're really stumped, and this has taken up immense amounts of time. Thanks in advance.

[nathan_hui]

For those CDers who are completely confused as to what a Samantha is, this problem is FTC, not FRC.

If I understand your problem correctly, the brick isn't communicating with the Samantha module. I'm not too familiar with the setup (another mentor usually deals with this), but go ahead and replace the USB AB cable, which could be the problem. The other problem I'm guessing at is a misconfigured brick, so check the documentation for how to set up a brick to communicate with the Samantha.

[Team230]

Thanks for your input. We tried and fixed a bunch of things today, and now have a different set of problems. The problems seems to possibly have been caused by setting our network password to both the Primary and Secondary slots. The Samantha automatically connected to the Primary (as visible in samostat), and although the Primary and Secondary networks were identical, this seemed to be the problem, because when we overwrote the Primary network with a bogus address, FCS began to recognize the robot.

However, we now have a new problem. This happened to us at past competitions too, even though we've been able to connect successfully there. Randomly, the robot will suddenly cut power to all of the motors and servos. We've diagnosed the problem to determine that the NXT is still receiving signals from the Samantha, and the NXT still recognizes when joysticks are moved around or buttons are pressed (we have it beep). As well, the program is certainly working fine, because all of the controls are in the same loop as the beep, and there is no way that it could not be sending the command to run the motors. We've also ensured that power is not getting cut as all of the lights remain on and don't flicker when the motors stop. The problem can be resolved by restarting the robot and the NXT brick, and the robot will work again for another minute or so before refusing to run.

Do you have any idea why this new problem could by happening? Thank you very much!

[nathan_hui]

Could you write some very basic diagnostic test code (run this motor continuously, run that motor continuously, etc. but only one thing at a time) and see if you still have the problem?

You can't be sure that the NXT isn't sending the command to the motors until you explicitly tell it do only that and nothing else (by writing a program to do only that). If that doesn't work, replace the HT Motor/Servo Controllers. Also test the motors by connecting directly to a +12V power supply. If you happen to have a function generator, you can also test the servos to make sure they work.

When you debug a problem, look at the way the robot receives it's control commands and simply follow the same path, testing each part until you find the one that doesn't work. That's essentially what we're doing here: testing in sequence the network, software, motor controller, and motors.

[Team230]

We essentially switched everything around on the robot, and have come to the conclusion that the motor controllers are locking up when the power drops too low or surges. All of the symptoms remained except for when we added another battery that powered the motor controller with our drive motors separately. Once the motor controllers overload we have to restart the robot and Samantha (but not the NXT) for the robot to begin working again. We also ran the robot without the Samantha and just on a wire, and the problems still continued. We tried replacing this controller and that didn't solve the problem as well. Due to FTC rules, all 8 of our motors, our 3 servos, and our Samantha have to be on the same power supply, and all motors have constant moderate strain. The problem occurs whenever we flip our drive motors from forwards to backwards, and at any other point where we put too much strain on the robot.

After replacing the battery to a full one yesterday we were able to drive problem free for about 45 minutes before we packed up. However, we replaced the battery to a fresh one today, and the problems still continued. The power for the controllers and Samantha is wired in parallel, and all of our drive motors are one one controller and our lift mechanism is on another controller. Is there anything that you think we can do to resolve our power problem? Thanks so much!

[Andrew Lobos]

We've had what appears to be this same issue on both our 2012 and 2013 robots. In both cases, the JoystickDriver debug output for the external battery went to OFF even though the controller chain was powered and connected. We found that only the first controller becomes unresponsive - it would see the controllers if we made any other controller in the chain the primary one. A power cycle always fixed the issue, but it was just a matter of time before it happened again.

In both cases we ended up replacing the controller and not having the issue happen again on that robot. I'm not sure if the controllers are cheaply made, or if we just stress them more than they expected, but it seems like this is an issue with the design of it.

[Team230]

Hmm, good to know that this at least isn't an isolated issue. Do you think we should try to change the order of our controllers to see if the problem is indeed in our first one as well?

[Andrew Lobos]

What we did was:

- Waited for the controller chain to drop. At this point there is one battery icon in the top right of the NXT display and JoystickDriver's debug screen lists: "Ext Batt: OFF"

- Without cutting power, connect the NXT to the second controller in the chain. (assuming it's not a servo controller), our NXT saw the battery voltage and could talk to it.

- Again without cutting power, move the NXT back to the first controller. Our NXT was still not able to communicate.

Could you see if your robot acts the same?

If you wanted to change the order of the controllers, I would put the one you suspect to be bad on the end. That way if it drops, it shouldn't take anything with it.