Battery Smoke

I think it is appropriate to discuss this now so teams know for next year. There were three battery incidents in Atlanta that I know of. Team A & Team B both designed battery trays such that the battery mounts on it’s side. In both cases, a battery was inserted into the robot with the terminals down against the the frame. The result was that the frame (aided by the battery weight) pierced through the insulation (yes both terminals on both batteries were insulated) and the frame provide a nice low resistance short to the battery. Team A’s battery swelled in all directions and distorted the case while providing significant heat. (It was a practice battery) Team B’s battery provided so much heat that the positive battery terminal (which is soldered inside the case) detached from the internal terminals and pulled out of the battery. Both teams had designed the tray so that a battery when mounted with the terminals up had no problems. Just a simple mistake in the heat of battle.
Team C was having some trouble with one of the battery connectors and decided to bend the terminals inside the Anderson connector to make things easier. (don’t do this at home) In the process, one (or both) of the terminals was released and backed out of the block producing a high resistance connection. Eventually it appeared that terminal pulled back far enough that it contaced the opposite terminal and shorted the battery.
Pictures will be added when they become available.

Wow - pretty scary!

This confirms what I told teams during inspections about the importance of carefully insulating the battery lugs and terminals. I must admit that I hadn’t verified that the battery couldn’t be inserted in robots with the terminals against the chassis or battery holder frame. Our (Team 980) robots have been designed such that the battery is horizontal, but couldn’t be installed with the terminals against the holder frame or chassis.

Next season, FIRST should address this in the robot rules (as much as I don’t like adding rules, this is a very important safety issue). A very conspicuous warning message should be put on the battery box in the KOP as well.

As an inspector, I always checked for well insulated terminals, and kept an eye out for terminals that were too close to things. Never really saw any that were in danger of a short. But an interesting story yes.

Neither of these robots would have raised an eyebrow for an inspector. The batteries were used in accordance with the rules and well insulated. In the heat of the moment things happen unless design prevents it.

As an inspector, I always checked for well insulated terminals on all of the batteries a team brought, and kept an eye out for terminals that were too close to things. Never really saw any that were in danger of a short. But an interesting story yes.

Team #1726 uses an extruded fiberglass frame. Since the fiberglass does not conduct electricity, it is the perfect material to build a chasis with, and provides a nice place in which to place a battery.

Another thing to mention, inspectors can’t always catch everything. While they see the robot at the beginning of the weekend, much can happen between Thursday and Saturday. I saw many exposed electrical connections when doing rechecks on Saturday in Curie, but also saw a lot on Thursday doing the initial inspections.

We have kicked students off of our team before for intentionally shorting batteries out (during build season, they never had the chance to make it to competition). One thing I did notice (which is a rather large pet peeve of mine), is the number of people in Atlanta (and any competition in general), that carry the battery around by the Anderson connector. Now, if you have properly crimped and soldered lugs (NOT screw on terminals), this isn’t as serious … but very few teams actually crimp lugs onto the battery cables. Depending on your insulation job, this could cause the terminals to become exposed, and short out your frame. Sadly, for most teams (including ours) the battery is one of the last things to get mounted on the robot, and usually ends up in a bad place.

Please, enlighten others on your team about proper mounting, wiring, and carrying methods for the batteries. While there has never been an incident (that I’ve heard of) where one of these batteries have exploded … I’m not going to say it won’t happen. Nothing worse than a robot covered in lead and acid to ruin your weekend (although, I suppose an arena of people and multiple robots covered in lead and acid could be worse).

This is also a big problem for me. Not only does this promote using one hand to carry the battery (always a bad idea), it also causes significant battery damage. By holding and carrying the battery by the wires, there is undue stress placed on the joint inside the battery. This joint will fail unexpectedly causing an open circuit when powering the robot, but may check OK with a meter without a load.

As Al said, unfortunately all three of these cases were not something that an inspection would have caught. A number of the inspectors, including myself, were circulating the fields in Atlanta to check for occurrances like this, especially after the first battery issue, however there’s only so much that can be seen with the number of teams moving around throughout the day. Particularly for things like the Anderson connector modification, what might seem a logical thing to do at the time can result in a very dangerous situation out on the field.

I would also like to add that teams should be very weary of loose/uninsulated wires within their robots, as well as chassis grounding. There were a number of occurrances in Atlanta this year that I am aware of, including 3 that I found, in which teams’ robot frames were grounded to their electrical system. This is not overly common but it can easily go unnoticed. It can also happen from a loose wire or a shifted component, just like any other failure. All of the teams with this issue in Atlanta were not grounded at their respective regionals. Teams should be careful to check for grounding occasionally.

Finally, please mount electronic components on a nonconductive surface.

<R58> All wiring and electrical devices must be electrically isolated from the ROBOT frame; the
ROBOT frame must not be used to carry electrical current (this isolated ground arrangement
is necessary due to polarity reversals that occur under certain operating conditions such as
during motor direction reversals).

Every grounding problem I have seen has come from a rogue bolt contacting a terminal or crimp somewhere, that would normally go unnoticed. Remember that this rule isn’t meant to make your electronics inconvenient to place, only keep things safe at the end of the day.

our battery this year was mounted vertically but we have a demo for a sponsor this week, and so we’re breaking out the old bots. my number one priority right now will be making sure that those old bots have the leads up on their batteries (i know that one of them has a flat-lying battery). thanks!

amazingly enough, even as a rookie team, our electrical system was(or at least started out being) quite organized. it was the only thing that was fully designed in inventor before being used. unfortunately it was changed on ship day, along with our bot being changed from a ramp bot to an arm bot. yes, I said on ship day. :ahh: we will never ever do that again. but designing the electrical system up front was definitely helpful, and made it easy to make changes later.

While we are on the topic…
A good way to insure the connection of the terminals to the battery post is the use of star washers. By placing a star washer (internal or external star lockwasher) between the battery terminal and wire terminal, you will prevent the terminals from rotating around the provided bolt. An additional lockwasher under the nut will prevent the hardware from loosening as well. Any time the terminals rotate, you have the ability for the hardware to loosen and the resistance of the connection to climb. In addition, the lockwashers will pierce any surface contamination on the terminals giving you a lower resistance connection.

As the head Electrician on my team(1647), i make sure that every single person on the team knows how to and IS carrying the batteries right. Now- the issue of the battery being mounted on its side. If you do that aways make sure that the terminals are UP and pointed out in the clear. thats the best way- i found at least- to mount a battery. keep in mind though that not the entire battery needs to be in the open just the couple of inches on the end that have the terminals.

These tips are wonderful about handling the battery but this year, the controls mentor saw all of the trouble with last year’s battery terminals and decided to get lugs and extra bent Anderson pins for our batteries. To secure the lugs, we used lock washers and a nut that had a star washer at the end of it for some added safety. To test how strong this combination was, I had our chief engineer pull on the battery to test the battery and he was for the first time satisfied with how the batteries turned out. If anyone is interested in the lugs we used or the extra bent Anderson pins I will try to find out the part numbers.


While we’re on the topic of electronic safety features I wanted to bring another thing to light. Too many times I’ve seen robot controllers placed in unsafe locations, either because they were installed late in the build season, or for convenience of access. I’ve seen a number of RCs lose radio, tether, and programming ports due to unfortunate placement choices. This is not something that you can fix on your own (I’ve tried :rolleyes: ), and unfortunately “but it got hit in competition!” doesn’t cut it when dealing with your warranty.

It’s happened to me in the past and it can certainly happen to you. This is something that can easily be avoided however, so I urge teams to keep this most sensitive of components in mind early on in design decisions. More recently I have seen teams that build extensions to the tether/programming ports for easy access, while setting the RC a bit further back for protection. I think this is an excellent idea, however make sure you keep the indicator LEDs visible at all times, and make sure it isn’t too hard to get to the RC to make changes or to secure wires (which should be done frequently since they have a habit of coming loose). It’s very easy to overlook your controller as a point of failure, but it hurts a lot if that’s what ends up breaking on the field.

in pre season this year a few of us stupid people in the team decided to stick a threaded metal rod on the two terminals of the batery. The next day it led to a short that melted the battery, the wires, when it cooled, were then stuck inside of the casing.

You have to be extremely careful when dealing with components like this. We tend to take these things for granted but these batteries are nothing to shake a stick at. They are capable of extremely serious current output and can be very dangerous to yourself and everyone else. A good rule of thumb is to treat the battery as if it could attack at any moment: don’t make any sudden movements, and certainly don’t encourage it (like prodding it with a threaded rod for instance :ahh: )!

We use a bolt with nylon lock nuts to secure our terminals to the battery. I torque these bolts down as tight as I can (unlike what you do with automotive batteries). We have not had a single problem with the bolts coming loose or with the terminals rotating. I always use new terminals when I am preparing a battery, so I know I am getting a good low resistance connection to the battery. I am also very peculiar with how students carry batteries and I have seen engineers on our team who also carry the batteries incorrectly by the cables. This year I wrote a message on all of our batteries on the proper way to carry them.

On another subject. In Atlanta our robot developed a problem with our gear tooth sensors grounding to the frame causing us to look through the robot to pass inspection. It took me about 5 minutes to figure out, but what I want to know is why the screw holes on the gear tooth sensors are grounded? We have about a 1" nylon spacer to extend our gear tooth sensors out to the gear teeth, but the bolts that hold them to the robot go through the nylon spacers and then bolt directly into the frame, which caused it to be grounded. I don’t believe it was like this at Lone Star or at the Bayou regionals.

Your method is one of the better ones I have seen. Unfortunaltely, the softer metals that make up the battery terminals and the copper wire terminals will deform with time under that kind of stress. So check them on a regular basis just to be sure.
The gear tooth sensors are infamous for causing the kind of circuit to frame short you describe. Many designers use this method of grounded mounting holes to help reduce noise and insure that part of the circuit is at the same potential as the frame. In our competition, FIRST has made a decision that the frame of robots remain neutral. I agree with this decision simply based on the electrical designs I have seen over the years. As this thread has pointed out, the battery is capable of significant (welding) currents. As such, I would not want a robot to be electrically destroyed by the failure of another team to insulate sensors or custom circuits. There are nylon shoulder washers used to insulate power transistors that can be effectively used to insulate the sensor packages from robot frame.

A slightly late reply, but I’m going to do it anyways.

We’ve put extensions on our tether and program ports in previous years, but the IFI people don’t seem to get a kick out of it. In the 2 years we went to the Championship (2004 and 2005), the IFI people strongly advised against doing this, as it can cause interference with the radio.

Just to add to what Adam said, at our second regional this year (San Diego), one of the teams in our alliance had a poorly covered controller. A bad tangle with another robot, caused the opposing robot’s arm (or appendage, or something) to grab all 8 (or so) PWM cables and bend every single pin that they were using on the PWM side of the controller. They carefully bent the pins back, and by some miracle, didn’t break any. But, everyone should know that the controller is the single most expensive part possible on the robot, so treat it accordingly.