CompuServe Thread

FRANCES problem

1 messages in this thread
#95876From: Phil SaundersMar 28, 1990 11:13 AM
I have been having some unusual problems with Lucas/Fr ances. The addressing problem I reported before has cleared up–I had a bad connection to a jumper I have been over the board twice with an ohm meter and all the connections appear to be proper. AFM now reports 1 Meg of 32 bit ram at 00400000. If I run modeload and addram with the memory test option, the memory tests out perfectly. (I have run the extensive test for 4 hours with no problems) CPR also allows me to examine and test the memory–it appears normal. When I try to add the memory to amigados, various things happen. 1) Dos commands (maDE resident or typed in)–no problems. 2) Zork text adventure–usually no problems, sometimes a crash. 3) WordPerfect–sometimes no problems, often crashes, sometimes while putting up a requestor 4) Maxiplan–always crashes. 5) Speed check program–usually no problems. I'm puzzled by this. The most successful test had Kickstart in 32 bit ram and let me run the speed program, Zork, Wordperfect. All appeared normal. The system crashed while loading maxiplan. I have fiddled with U-9 on Lucas. It makes a difference.–the 74als74 and the74ls74 were most successful. I should mention that Lucas runs perfectly until the Frances memory is added. Speed is 16MHz, mode is 36c4c4. I come up with four possible/plausible answers: 1) Overall ground/bus noise is too high (My Amiga is sensitive) 2) Conflict with Kickstart addresses ? 3) Trying to use 32 bit ram as chip memory ? (Unlikely) 4) Graphics screens raise level of DMA contention, pushing system over working margins and crashing it. CaN anyone suggest an answer or line of approach in debugging this board?