Re: VIC-II DRAM refresh

From: Gerrit Heitsch <>
Date: Sat, 15 Oct 2016 18:56:01 +0200
Message-ID: <>
On 10/14/2016 07:16 PM, Francesco Messineo wrote:
> On Fri, Oct 14, 2016 at 5:12 PM, MichaƂ Pleban <> wrote:
>> Hello!
>> Francesco Messineo wrote:
>>> isn't the 8520 a 6526A (or B?) but made with a different process? I
>>> could use the 8520 in place of the 6526 with no problems once.
>> No. The 8520 has a different TOD timer, if you use it in the C64 the
>> programs which rely on it will not work (or will display incorrect
>> time). Also the TI$ variable in BASIC relies in the TOD in one of the
>> CIAs so it will not work too.
> yes, it was 8521 then. I'm sure at one point there was the 85xx
> version of the 6526.

Not 'there was', all 6526 after the end of '86 are internally 8521. They 
just labeled them 6526 again, probably to avoid confusion. You can check 
that by software, a timer IRQ behaves slightly different. Also, a good 
hint is the '206A' and '216A' after the datecode. The '2' indicates 
HMOS-II process. NMOS would have a '1' there.

There is some labeling I haven't been able to understand yet though. I 
have a  few 6526 that come labeled '6526A-1', they're from '85. Anyone 
have an idea what the 'A-1' means?

> Even with TTL-compatible logic, they could invent any consistent
> numbering, like XXYY257
> instead of 7708 and XXYY258 instead of 7709...
> For example, their 74LS245 equivalent was following that golden rule,
> so at one point they decided to "go insane" on numbering.

No really 'go insane', more along the lines of just counting up. The 
leading '7' means HMOS-I, the next '7' means 'support logic' and then 
just start from 00. 7700 was the 82S100-clone.


       Message was sent through the cbm-hackers mailing list
Received on 2016-10-15 17:01:45

Archive generated by hypermail 2.2.0.