During the last 45-30 seconds our matches, we noticed that our robot consistently began to have trouble turning and wouldn’t allow us to properly line up with the rope. Around the 20 second mark, the robot began acting normal again and we could easily line up. After looking at the battery voltage log we noticed the voltage started wavering towards the end of the match, yet it wouldn’t really explain why the robot would return back to normal about 10 seconds later. Any explanations?
Can you post details about your drivetrain setup, including # of motors, motor controllers, wheel types, and any drop center?
Several (all?) of the FRC legal motor controllers have overcurrent protection. They will cut off power to the motors if you draw too many amps for a certain period of time. Also, your PDP has self-resetting breakers which do the same.
If your drivetrain is underpowered or difficult to turn (pneumatic tires with no drop center at either ends of the robot, or some combination), then the motors may not be able to break static friction to turn. Instead they might stall, thus tripping the current limits.
At the end of the match your motors will produce less power due to internal heating and decreased battery voltage, making it more likely they stall instead of turning.
I suspect your motors return to normal at the 20 second mark just because the breakers/motor controllers reset after 10 seconds.
We use 4 CIMSs, each controlled by Talons, to power 4 of the COPs wheels, no drop center. We also use a CIM on our climber which works perfectly fine and climbs easily at the end of each match.
Are you using pneumatics, and do you have an on-board compressor?
It’s possible that your compressor plus drivetrain at the end of the match is enough to cause you to hit brownout (you can check this in your logs). Once the tanks are full and the compressor shuts off, the robot has enough power to drive again.
If this is a problem, you can add code to shut off the compressor manually, or while other functions are running. We shut off our compressor when shooting and climbing to ensure we have the maximum power available.
I appreciate the suggestion but we don’t have pneumatics
What are your track width and wheelbase length? If they are close to each other or length > width, perhaps your voltage dropped just below where your wheels could overcome scrub forces.
Can you check the driver station logs for packet loss?
If by “acting funny” you mean the entire robot starts stuttering, and the robot signal light flashes erratically, then it sounds like a brown out issue.
Given that it’s at the end of your match, that sounds like bad/not charged batteries, and that would make sense for browning out and when the issue occurs.
Please ensure for competition season you only use your new batteries from the current year. Older batteries become unreliable after 1 competition season (generally), just because of how abused they get during matches.
Your descriptions are representative of a large battery drain causing your crio to reboot. As you run your match your battery starts to deplete, once it drops below the threshold your crio reboots but it never totally shuts off so your logs don’t show a 0 volt period. Once the reboot is finalized it appears your battery is back to a “normal” voltage but if you continued to run your robot it likely would repeat within 30 seconds again. The big draw seems to be the final turn you’re attempting. If each wheel is independently powered by a cim, it sounds like you’re stalling at least one, likely two, and possibly all four of the cims, as they work against each other as you try to turn. When they stall the voltage goes below 7.2 and the brownout reboot occurs. A big warning here is the continual motor stalling which could be occurring can draw high current and start to fail battery cells if the battery is repeatedly exposed to these conditions. The cells are paired at 2 volts each and so a good example if you have any bad cells is if you fully charge your battery to 12 (or up to 13) volts and as soon as a load is applied the battery voltage drops by roughly 2 volt accumulations. If you find you have batteries with bad cells, it is recommended you dispose of them properly and not to continue to use them. Continual use of a battery with failed cells can result in unexpected catastrophic failure and serious injury to people around the battery.
We seemed to have fixed it by simply making the robot have a drop center. The voltage was remaining pretty constant for the entirety of the match.
If you’re running 6 wheel tank, you pretty much need to have a drop centre. Good on you guys for coming up with a solution.
Live and learn!
Something that our team had learned this season while at AZ North was that some older batteries will still appear fine in charging and on a battery beak (“Good” health) and still be horrible.
In one of our last qualifying matches, we were browning out close to the middle of the match. The battery was dropping below 4 volts (which the FTA and FTAA said they had never seen a robot do before without rebooting the roborio and other components). The battery had depleted to a point where we were barely able to reach our rope and then were unable to fully climb, costing us the climb, which would’ve turned that match from a loss to a win for our alliance.
Turns out, that battery was 5 years old and we had not realized it until the end of that match.
tl;dr: A 5 year old battery had appeared fine and in “good” health and had costed us a match because it had depleted insanely rapidly. (And also impressed the FTA and FTAA)
It is wise to ALWAYS label your batteries with the year you got them. I would not recommend re-using any batteries from previous years for competition. For practicing, no worries, but the abuse they take during competition just ruins the batteries.
As the said FTA, I can affirm. I had never seen the field monitor show a battery voltage at 3.75 before that match (and that is with the 1/2 second logging on the FMS…didn’t see what the DS showed in its logs). I had to go over to the player station to confirm I was seeing real data, and sure enough, it was. I was amazed that the robot was still able to move and not be rebooting at least something in the control system.
When we went to San Diego we had some problems that sound just like yours. What is your robot weight? We had a robot that was 120 on the dot. It drew a lot of power and made our robot very slow. We just removed our dual shooter that we weren’t using and we were quick and very nimble afterwards. Second. How many cameras do you guys have? We had two cameras on our robot that according to the tech crew there at the competition said might have been drawing too much power combined with other things. I am on mechanical team not programming/technical so the camera thing may not be right. Hope this helps.
We also had a weird issue at our Week 2 where we would seem to lose power on one side of the robot after a strong hit, and then the ability to drive would slowly come back. Turns out we has wired both CIMs on that side to 30A breakers instead of the 40A. After a little repurposing and relabeling of speed controllers (and adjusting a few PWM wires) we had moved the drivetrain fully onto 40A breakers and the ball lifter got downgraded to 30A. As we tested we also noticed we forgot to swap the Brake/Coast settings to match, but that got corrected as well. After that we’ve been a reliable driver.
It is for this reason that I recommend using a CBAIV from West Mountain Radio. This battery test will duplicate the discharge curves printed by the manufacturer and will allow you to overlay curves from other batteries or from the same battery over a period of years. the test takes about two hour per battery but it gives you all the data you need to know.
Al
Another thing you can do is to learn to read the manufacturers date codes on the batteries… :o