|
|
 | | Re: "Servo" Driver for NQC
|
| The ON/OFF bang-bang control you are trying has some problems that proportional control does not. While it is true that if the update rate is high enough, or the controlled plant damped enough, that switching a drive ON and OFF can simulate (...) (24 years ago, 30-Jan-02, to lugnet.robotics)
| | | | | Re: Incorrect result
|
| (...) Yes, this is what happens when one is used to just making everything long and not considering overflows. Even worse was that I was doing the math on the computer's calculator, but not looking at the intermediate result, just typing in the (...) (24 years ago, 30-Jan-02, to lugnet.robotics.rcx.legos)
| | | | | Re: BricxCC release 3.3.3.4
|
| hi Guido, I just a very little time also the comming week, but my first impression is very good !! Thanks very much. And you said you weren't a fast programmer ;-) (...) It sound great indeed. (...) I couldn't get it to work, but maybe I did (...) (24 years ago, 30-Jan-02, to lugnet.robotics.rcx.nqc)
| | | | | Re: Incorrect result
|
| (...) 256 * 128 = 32768, one higher as the biggest positive integer value. The cast to long would help, but your cast is too late. I have not tested, but I think rc += ( (long)b->magnitude * (long)TA_costable[...] ) >> 7; will help. Try output in (...) (24 years ago, 30-Jan-02, to lugnet.robotics.rcx.legos)
| | | | | RE: Incorrect result
|
| <snipped code> (...) Robert, this is a classic problem with ranges of singed numbers in C rx = (long)(a->magnitude * TA_costable[a->direction]) >> 7; I don't have an RCX in frot of me, but I think you need to cast a->magnitude to long before the (...) (24 years ago, 30-Jan-02, to lugnet.robotics.rcx.legos)
|
|
