View Single Post
Old Aug 10, 2013, 05:01 PM
aadamson is offline
Find More Posts by aadamson
Registered User
United States, GA, Atlanta
Joined Dec 2010
2,059 Posts

I don't have anything to release just yet, but we've found some issues with the 0.3x code. The resolution on the IMU's gyro should be 8192.0 (floating) and it's coded in 0.3 as 8000.0 (floating), there are also some other timer initialization issues that need to be corrected.... Deadband calculations, etc.

I have no idea of the impact good or band on these, but just recently since ala42 started to look at the code did he bring these to my attention. I've seen his prelim code and it's cleaned up a number of these issues, it also corrects an issue with turning on optimizations under eclipse that forces the code to be compiled with no optimizations as it is in the git repo (or it won't run). Those seem to be corrected as well.

I suspect in the next few days, I may have some code for someone to test... We want to go very slowly here as the deadband changes could have an impact on the FET's... Either good or bad ... Either way I wanted to let everyone know that some of the above may be related to some of the performance issues people are having.

Also we found a dT overrun issue that can occur that we hopefully have fixed. It was assumed that the code would always run at a specific interval, but if that didn't happen, all the timing could become wrong on the dT calculations.... We've fixed that now to use an actual time calculated dT.

This isn't really the optimization that the code needs, but it may get incrementally better code in your hands before we start the restructuring.

More as I know...
aadamson is offline Find More Posts by aadamson
Last edited by aadamson; Aug 10, 2013 at 05:18 PM.