On 09/24/2011 05:16 PM, Nicolas Welte wrote: > Gerrit Heitsch wrote: >> This also matches the appearance of the 8521 in late 1986. Could it be >> that the change in timer behaviour is due to a mask change done while >> transfering the 6526 from NMOS to HMOS-II? > > Until I read your test results, I thought the 8521 (as well as the > 8520!) shows the old 6526 behaviour. That's what I thought as well. Wasn't the case with this test program though. Furthermore, the 6526A with the 3089 datecode is also marked '216A' which some people claim means it should behave like the original 6526. Looking through my spare parts collection I found CIAs marked '6526A-1' with the datecodes '4585', '1986' and '1386'. All also feature a 'S' after the datecode. I tested them on my test board where all important chips are socketed. They were seen by the test program as '6526', meaning old timer behaviour. So the 'A' doesn't really say anything about how the timer behaves. The datecode seems to be the thing to look for. Or does anyone have a 6526 (A or non-A) from week 40 1986 and up that is identified as '6526' by the software? The results shown by the program are consistent, so far running it multiple times on the same system didn't result in any changes in the output for this system. Putting together a mix of old and new 6526 also resulted in a correct identification. Gerrit Message was sent through the cbm-hackers mailing listReceived on 2011-09-24 16:00:18
Archive generated by hypermail 2.2.0.