On Fri, 23 Apr 1999, [ISO-8859-1] Marko Mäkelä wrote: > On Fri, 23 Apr 1999, William Levak wrote: > > > Four sprites on the same line cause the disk drive to hang up when > > loading. > > > > They can be the first four, the last four, or every other one, it doesn't > > make any difference. > > Actually 2 or 3 sprites should be enough, if they are not successive ones. > As you can see from the timing diagrams in the document > <URL:http://www.funet.fi/pub/cbm/documents/chipdata/VIC-Article.gz>, the > timing of a bad line without sprites will be like this: > > rrrrrXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX++0-1-2-3-4-5-6-7-rrrrr > > The diagram is for the 6569. On the 6567R56A, replace "++" with "++++". > On 6567R8 and later, replace it with "+++++". > > The BA line will be made inactive 3 cycles before the DMA takes place. > That'll happen after the 2nd "r" (refresh cycle) on the left. If sprite 0 > is active, the processor won't get any cycles after the character fetch on > PAL systems (the 40 cycles marked with X), or it'll get 1 or 2 cycles on > NTSC systems. Considering the 3-cycle rule, it won't make any difference > whether you have sprites 0..3 active, or just sprites 0,2,3 or 0,1,3: the > processor won't get any time unless the DMA will pause for longer than 3 > cycles. > > I think that activating e.g. the sprites 0,2,4 or maybe even 0,3 should > hang the protocol. > You're right, both those combinations cause the load to hang up. Bill - This message was sent through the cbm-hackers mailing list. To unsubscribe: echo unsubscribe | mail cbm-hackers-request@dot.tcm.hut.fi.
Archive generated by hypermail 2.1.1.