|
Re: Real time versus normal timing
We had an interesting day. Early on, one of our limit switches failed. This exposed a flaw in our interlock logic where our catapult could and did fire. Parts flew.
The problem was traced to the switch. We had the exact same failure later on, on the other side of the robot. Luckily, no one was hurt.
On the positive side, we had cleaned enough cruft from our code that our CPU usage had dropped to a steady 30%. We achieved a large portion of this by moving the CAN calling portion of the drive code to another loop (thread) apart from all of our driver filter, gyro, and PID. With this change, we also stopped getting any CAN errors.
|
Those poor suckers on the West coast get three hours LESS sleep than we do afterward. Anyone else glad to see that crate go every year, regardless of what's in it? - Rich Kressly [more] 
Hybrid Mode
