|
From: David S. <da...@bl...> - 2005-09-23 22:55:16
|
Hi Fredy, Well spotted. Your patch does indeed fix the use of D flip flops in the Ring Counter. However, I'm not entirely confident that the D flip flop is being simulated properly in all cases (this is largely a problem with ktechlab's internals, in that when two or more logic states change simulataneously, ktechlab makes no guarantees as to which order the logic callbacks will be made). I'll continue investigating tommorow, after some well-needed sleep :) Regards, David Saxton On Sunday 11 Sep 2005 19:13, Fredy Yanardi wrote: > Hi David, > > Today I discovered a bug in D Flip Flop. It only remembers current D > input, if clock and input are changed simultaneously, then the flip > flop will produce wrong Q output. Simple and obvious case is when you > create a Ring Counter. > > I don't know whether this has been fixed or not in the latest svn, > because I try to update my source but it gives me "svn: PROPFIND of > '/ktechlab': could not connect to server > (https://svn.fadedminds.com)". If this hasn't been fixed I've made a > small patch for this. I haven't tested other flip flop types whether > they have same problem or not. > > Regards, > > Fredy Yanardi |