Richard Atkinson wrote: > Whoops, it was in Knowledge Base and there's a full explanation from > Nicolas about it being due to faulty CIAs not the 8500 or anything else. > Incidentally, Nicolas, is this the 8521 you're referring to? I have a 64C > in which one of the CIAs is an 8521 and the other is a 6526A! So perhaps > Commodore were aware of the bug in the 8521 and used up remaining supplies > somewhere where it wouldn't matter. There must be far more versions of the CIA than the type numbers tell us. The chips with the different timing that cause the problems are normal MOS/CSG 6526 (1,2,3 MHz) with a date code of '87 and later. They were introduced with the C64E and the 128DCR and were not replaced till the end. I bet those are HMOS-II versions, even if they're still labeled 6526. In 1986, Commodore built some C64 and 128 machines with the strange chip combination you described above. One is a 8521, the other is a normal 6526 (pre 1987). There must be a bug in the 8521 for sure, but this chip has at least the same IRQ timing for the timers as the old 6526 chips! I don't remember which of the two CIA slots has the 8521, do you? Judging from the type number, the 8521 was Commodore's first attempt to make a 6526 compatible HMOS-II CIA. The 8520 was introduced much earlier with the Amiga, but we all know it has the TOD clock replaced with a 24bit binary event counter. And then there is the good old NMOS 6526. I found that especially some versions from 1984 that are marked 6526R4 die with a stupid failure: they draw too much current from one or more internal data lines and mess up reading from the colour RAM. I just observed this in the second machine. Too bad, the chips are otherwise still okay, but one had to put them behind a wall of bus drivers to keep the rest of the system running. Nicolas - 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.