Re: Strange 8255 behavior

From: Gerrit Heitsch <gerrit_at_laosinh.s.bawue.de>
Date: Tue, 12 Jun 2018 18:54:18 +0200
Message-ID: <d1852830-a6ee-6dfe-b840-9f120cd8f7df@laosinh.s.bawue.de>
On 06/12/2018 05:44 PM, groepaz@gmx.net wrote:
> Am Dienstag, 12. Juni 2018, 17:05:20 CEST schrieb Mia Magnusson:
>> Den Sun, 10 Jun 2018 15:50:50 -0500 skrev Segher Boessenkool
>>
>> <segher@kernel.crashing.org>:
>>> On Sun, Jun 10, 2018 at 10:17:17PM +0200, Gerrit Heitsch wrote:
>>>> On 06/10/2018 09:46 PM, Segher Boessenkool wrote:
>>>>> On Sun, Jun 10, 2018 at 06:42:02PM +0200, Gerrit Heitsch wrote:
>>>>>> On 06/10/2018 05:36 PM, Segher Boessenkool wrote:
>>>>>>> http://siliconpr0n.org/map/mos/6526/mz_mit20x/
>>>>>>>
>>>>>>> The ports are the low half of the pins (PA on the left, PB on
>>>>>>> the right).
>>>>>>>
>>>>>>> (I have a .xcf if anyone is interested, marked quite a few
>>>>>>> signals, but I haven't done the port stuff very much.  It's
>>>>>>> about 400MB).
>>>>>>
>>>>>> That looks quite different from the one I posted. Looks like MOS
>>>>>> did quite a bit of redesign between the NMOS 6526 and the HMOS
>>>>>> 8521 (which still got labeled 6526). Might explain the little
>>>>>> differences in the way they behave.
>>>>>
>>>>> No, this is an actual 6526r4.  This is an 8521r1:
>>>>>
>>>>> http://oms.wmhost.com/misc/MOS_6526A_CIA.jpg
>>>>>
>>>>> (and this is an 8520r4, the CIA used in amigas; it has a different
>>>>> TOD clock, and as you can see it's different from 8521 in other
>>>>> ways, too. But clearly 8520 and 8521  are more related.  The
>>>>> lineage is almost certainly 6526 -> 8520 -> 8521:
>>>>>
>>>>> http://siliconpr0n.org/map/mos/8520/mz_mit20x/ ).
>>>>
>>>> Yes, but the 8521 is a drop in replacement for the 6526 (I have a
>>>> C64 Board with a 8521R0 on U2) and later revisions of that chip
>>>> have been labeled as '6526' again, probably to avoid confusing the
>>>> customers. You can tell them apart by the datecode or by the '206A'
>>>> or '216A' next to the datecode.
>>>
>>> _Almost_ drop-in replacement, yes.  But the die photos are really easy
>>> to tell apart (an 8521 does not say "6526" on the die, it says
>>> "8521").
>>
>> I can't remember which versions, but at least some version of the CIAs
>> can only be used as one of the two CIA's in a C64. Using it as the
>> other CIA causes keyboard problems.
> 
> no, both can be (and are commonly) used in either position
> 
>> And when using it as the CIA that
>> generally works causes interrupt failure in the common diagnostic cart
>> (the one that's not called dead test, not sure if it has a better name
>> than diagnostic).
> 
> that is correct - some of those diagnostic tests fail with the "new" CIA

Only if they don't take the 'new' behaviour into account.

  Gerrit
Received on 2018-06-12 19:00:09

Archive generated by hypermail 2.2.0.