FRC4188 disconnected for 50 seconds within a few seconds of Finals 2 at Houston Einstein.
VIDEO of F2 Einstein disconnect: http://bit.ly/2qMnlZb
We wanted to find the root cause so we set up some high speed videos, force sensors, and 3 axis accels and slammed the robot into our airship over 20 times.
Videos are below if you’d like to see a robot move a 600+ pound field element. Once grades get turned and other end-of-the-year teacher stuff this week, I will post our white paper here if you like data plots and still images of electrical components - also have a 3D printed cover to prevent this kind of DC.
Compilation of Impacts Trying to Disconnect: http://bit.ly/2r7nXM1
The trial that DC’ed - watch the wire carefully: http://bit.ly/2rrZbWk
After securing the battery cable, 14 more full speed impact trials were completed and there were no disconnects.
Compilation of Impact Trials after wire fix: http://bit.ly/2qJHBN9
Houston Champs was a great experience and FRC118 and FRC1678 proved they deserve their amazing reputations – I really appreciate how they managed the alliance and maximized everyone’s contributions.
Which wire should I be looking at specifically in the high speed? From looking at the battery and switch wires I’m not seeing anything that should shut down the robot although I could very well be missing the obvious given how tired I am rn.
Edit: This is cool that you guys were willing to do this with an Einstein finalist robot to figure out the issues your we’re having.
The black #6 wire smacks the red off button on the 120a circuit breaker just enough to cause a momentary disconnect, but not enough to shut the robot off permanently.
I saw that happen on several different robots over the course of the season.
Yea, something similar happened to us at WPI in 2016, in one of our qf matches we went over I believe the most and when we go to the other side we noticed the green eye of death was not on which we had wired directly into the vrm and when we got on the field we realise that the actual Anderson connector had hit the robot power off. I was in charge of batteries and just a simple mistake made by me to use a battery with a connector that was a little short and then not tieing it away rendered us useless for that whole match but luckily didn’t get us eliminated.
There’s a similar item on Andymark’s website. http://www.andymark.com/product-p/am-120ashield.htm
I’m sure it would be fine. I know we would have loved to have a bunch of these to use and to strap onto partners this year.
I am sure an inspector somewhere will not like it. Other inspectors will. :] Anyway I would look at it where it is installed. Out in the open would be fine. Obstructed by other structure maybe not so much. The important feature is the button needs to be easily reachable without putting your hand deep into the robot.
I’d like to see it in something other than black so that it stands out. Looks pretty open, but would need to see it on the actual bot.
The first time I watched the video, I couldn’t help but be drawn to the “pilot” in the airship protecting his forehead with his safety glasses right above what is going to be a known hard hit (at least his phone would probably have taken the brunt first of anything coming loose).
The second time through I saw the actual issue…and Hey, it isn’t the radio but a main breaker problem leading to a roboRIO cycle (you can see the RSL go out immediately and come back on just before the video ends). Hmmm. Like Mark, I saw that several times at my events as well, although it is really neat to see it happen this way and I’m glad that they followed up on root cause.
The edges look pretty sharp though. I would file the corners down. And then I would think you should be to pass inspections. Off-season or next year of course!
All that being said: This is a good example of the kind of things you need to do to find the real cause of a problem. Professionals do these kinds of things every day, and it is really nice for students to get a chance to witness this.
That’s a good idea. We will make a new version and print it bright yellow so it’s more visible. Also planning to make the gap in the front a bit larger to allow for hands/fingers to more easily reach the button in cases of emergency.
Having had my head and hands inside this robot as it came off the field, I can add that the battery in the robot at the time of fault had wires there were capable of much more movement than their other batteries. A different battery was installed and checked but the wires on that battery did not come near the main breaker.
