Karl Reichert wrote: > Hello, > > I read the TDMA-spec and I understood it in general, but thinking > about some details. What I don't understand is, how the so called > 'normal startup' is performed. > > If a new client comes into RTnet, his clock isn't synchronized to > master's one. So it means, it neither knows t_offs nor t_trans. > > So what it does is waiting for master's sync frame, which contains > T_sched and T_xmit. So slave can calculate master's jitter from this > values. > > The problem is, to calculate t_trans, slave must send calibration > request in its time slot. But how can it now when this is, if its > clock isn't synchronized yet? To synchronize its clock, slave needs > t_offs, but this is calculated with respect to t_trans. > > So my question is, how does slave knows when to send calibration > request frame? Or is it sending it with CSMA/CD discipline which of > course destroys any realtime for some time, cause it can cause > collisions on bus?
t_trans is set to 0 until the calibration has been performed. As indicated earlier, that will cause a slight but typically uncritical delay of the small calibration frames. The network designer has to take this into account by keeping an appropriate distance to the succeeding slot. Jan
signature.asc
Description: OpenPGP digital signature
------------------------------------------------------------------------- This SF.net email is sponsored by DB2 Express Download DB2 Express C - the FREE version of DB2 express and take control of your XML. No limits. Just data. Click to get it now. http://sourceforge.net/powerbar/db2/
_______________________________________________ RTnet-users mailing list RTnet-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/rtnet-users
