Announcement

Collapse
No announcement yet.

Is anyone else worried that this new platform is too difficult/unstable

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • #61
    I'm new to FTC and robots in general. However , I've got 30 years experience in working with electronics as a profession, I also have 30+ years of remote control car racing and bashing. I was and still am excited to be helping my daughter's robotics team this year, I was kind of happy to know that things had switched over to the Android/Java platform... I went out and bought Modern Robotics electronics, to be able to build my own robot, partly just for fun and partly to help my local club by allowing me to gain experience with the electronics, motors, sensors, and such...

    My observations after having a mostly functional test robot setup...

    The USB cables are too stiff. The physical USB connectors don't seem to lock in place very well. The ferrite bead on the cable just exacerbates the problems with the cables. USB cables in general seem to be a poor choice for the electronics connections between all the devices, the Mini and Micro connectors, just aren't really designed to be used in a environment that vibrates constantly and gets jostled periodically by moderate to heavy impacts the robot will take in competition. I feel that maybe a lighter weight USB cable, one that flexes more easily would be better then the ones that come with the Modern Robotics hardware. Even so, I'm not sure it's going to be very easy to overcome all the inherent problems resulting in trying to use the Mini/Micro USB cables as the primary means of connections between the hardware.

    Having tried to secure cables as best I can, I am still seeing intermittent shutdowns that I don't believe are directly related to poor connections in the USB chain. I've seen discussions about modifying the hardware, adding a capacitor and such, I'm not ready to do that yet, given it would probably void the warranty of these things, and I'm still not sure there isn't some defect that Modern Robotics should be responsible to fix.

    Competitions are fast approaching... It will be very interesting to see how may people have problems that are related to the USB communications platform we have to work with. Normally, on the internet I feel that those complaining, are just a few isolated people with problems being vocal. However, in this case, I wonder if the complaints we see here, are just the tip of the iceberg.

    Meanwhile... I've got to find some other USB cables, because I'm quite confident that at least 50% of the problems I've been having are related to the cables , or made worse by them at the very least. Off to do some googling and/or ordering of new cables.

    Comment


    • #62
      Our team has had ups and downs with the new hardware

      1) Not detecting motor controllers. We had a lot of issues detecting one motor controller but not the other. A solid "Fix" seems to be to:
      -Power off the robot and unplug the battery
      -Unplug the USB cable to the phone
      -Count to 10, really, to ten. If you don't wait long enough, the issue will still persist
      -Plug the battery into the core unit, power on, wait a few seconds, then plug in the phone
      This is the only sure fire way we have found to always detect all of the controllers plugged into the core unit.

      2) Saturday, we were testing some new programming for motor control to help climb the mounting, but after 10-20seconds of driving, the robot would continue driving from the last inputs for a few seconds, stop, then the phones displayed a communication error. We reverted back to standard code that worked for us last week, but we continued to have the same problem. We tried the above steps to try to resolve it, but no go. We unplugged all the cables, plugged them back in, but same issue kept coming back. I am not sure what to do at this point. We meet next on Tuesday, I can post the specific error message then.

      Comment


      • #63
        Originally posted by sethmo View Post
        Our team has had ups and downs with the new hardware

        1) Not detecting motor controllers. We had a lot of issues detecting one motor controller but not the other. A solid "Fix" seems to be to:
        -Power off the robot and unplug the battery
        -Unplug the USB cable to the phone
        -Count to 10, really, to ten. If you don't wait long enough, the issue will still persist
        -Plug the battery into the core unit, power on, wait a few seconds, then plug in the phone
        This is the only sure fire way we have found to always detect all of the controllers plugged into the core unit.
        I'm sure this way works.
        However, if you want to minimize the number of connect cycles on your connectors you should try the following and verify it that is also works:

        1)
        -Power off the robot (no need to unplug the battery)
        -Close the Robot app and Swipe if off to remove it from memory (press and hold lower menu button to show apps)
        -Sleep the phone by tapping the power button to blank the screen. (This seems to have the same effect as removing the USB cable)
        -Count to 10, really, to ten. If you don't wait long enough, the issue will still persist
        -Wake the phone (tap the power button)
        -Power on
        [/QUOTE]

        Comment


        • #64
          90% of the time with our hardware, closing the app with just the back arrow key, unplugging the hardware from the phone and plugging it back in fixes this "Not Found" issue. We have whatever version of the hardware that shipped in early October. Probably bad on the connector. I hope we can have a USB to X and an X to PDM solution that were that X can change between the OTG and the standard USB protocols as while as being less susceptible to disconnects, until we get a new Android-powered device to run the Robot.

          I wonder what happens while the phone is sleeping for those ten seconds, maybe if we can emulate that behavior we can shave off 10 seconds of this process of "re-finding" the hardware.

          Comment


          • #65
            EMERGENCY_STOP; disconnects...

            Originally posted by dmssargent View Post
            90% of the time with our hardware, closing the app with just the back arrow key, unplugging the hardware from the phone and plugging it back in fixes this "Not Found" issue. We have whatever version of the hardware that shipped in early October. Probably bad on the connector. I hope we can have a USB to X and an X to PDM solution that were that X can change between the OTG and the standard USB protocols as while as being less susceptible to disconnects, until we get a new Android-powered device to run the Robot.

            I wonder what happens while the phone is sleeping for those ten seconds, maybe if we can emulate that behavior we can shave off 10 seconds of this process of "re-finding" the hardware.
            In our case, we have found that we don't even need to touch the cable; just restarting the Robot (can be done remotely from the Drive Station) suffices.

            Which says to me that we still don't understand the underlying failure that causes EMERGENCY_STOP; some explanations lie in not having strain-relieved the system's cables well enough, but I'm not buying that theory so much, at least not in our case.

            We changed to a different PDM this weekend and suddenly all seemed more reliable; but we had used that same PDM the previous week and still had had the problem.

            Martin Haeberli
            Mentor, Team 7593 TigerBots

            Comment


            • #66
              Originally posted by mhaeberli View Post
              In our case, we have found that we don't even need to touch the cable; just restarting the Robot (can be done remotely from the Drive Station) suffices.

              Which says to me that we still don't understand the underlying failure that causes EMERGENCY_STOP; some explanations lie in not having strain-relieved the system's cables well enough, but I'm not buying that theory so much, at least not in our case.

              We changed to a different PDM this weekend and suddenly all seemed more reliable; but we had used that same PDM the previous week and still had had the problem.

              Martin Haeberli
              Mentor, Team 7593 TigerBots
              Unfortunately, we still need to unplug the robot most of the time to get the robot controller to see the robot electronics.

              Comment


              • #67
                I'm seeing a lot of the same things that have been said in this thread. Our motor controllers or DIM doesn't get recognized, so we go through steps similar to what sethmo wrote (Try "Restart robot" command, then try pulling USB, then try backing out of RC app, then try switching power unit on/off, then try reinstalling cables). Most of the time, those work.

                What we're seeing these days is even more frustrating. The RC will be communicating just fine with the DS, but for whatever reason it'll just stop communicating. The ping shows, so we know Wi-Fi Direct is working. The RC starts the "init/running" sequence upon a "Restart robot," the DS lights up so we can pick the op, and then suddenly the lights go off and we get the "Robot stopped" message. We try all of the usual tricks in the above paragraph, plus we do power downs and power ups of the phones, plus we try downloading our op programs again. Eventually, after about 20 minutes, for no apparent reason, the RC starts talking to the DS phone again.

                Our only thought is that this has happened twice, and both times it was when our FRC team was in our space and maybe there there a lot more bandwidth usage of our space's Wi-Fi system from all of the connected devices than our space could handle. Which makes me worry about the capabilities at the actual competition... although technically everybody's supposed to be on airplane mode at the competition, I guess. But count us among the teams that are having a lot of unexplained connectivity issues all the time.

                Jeff Stillman
                Mentor 8133

                Comment

                Working...
                X