Quote:
Originally Posted by Chris is me
Some have suggested that the system is more vulnerable to brownouts than in the past. (I'm not particularly informed on this, I don't know if people have tested this etc)
|
While we haven't done any intentional testing of the brownout, we have definitely found a difference with the RoboRIO in practical (non-competition) situations. In previous years, we didn't always realize when it was time to change batteries, because there would be different symptoms. We would lose acceleration, or we might drag a bit as we tried to do a kick or other high-current function, but the degradation was usually gradual, until it caused one of the digital components (most commonly our raspberry pi vision processor, but occasionally the D-Link radio, and in one case apparently the cRIO) to reboot. In each case, the failure took several to many seconds for recovery. The roboRIO has a defined brownout sequence, which cuts power to the motors first, then proceeding to other specific busses until the cRIO itself is eventually shut down. In practice, this means that our robot runs just fine at extended practices and demos until the drive system enters a "jitter" phase in which it is energized, draws the voltage down, shuts off, allows the voltage back up, is re-energized, and shuts off again, creating a cycle that executes about 7-10 times per second (based on my non-calibrated biological clock, but with a lot of acoustics experience). In Recycle Rush, this appears to have rarely occurred during a match unless a team was working unusually quickly, or using particularly old batteries. I suspect that in the near future, especially if we get more defensive games, teams will have to monitor battery voltage and total current draw, and implement software limits that prevent or at least limit the impact of these brownouts. The good part of this change is that at least we are much less likely to spend most of a match waiting on a reboot!