Unknown and short-lived MOS6522 Revision 1 (vs. Rev. 0)

[fhw72]

Some time ago someone pointed out that (contrary to my belief that there's only one revision of the VIA / 6522 (i.e. Rev. 0))
MOS also produced a 2nd version "R1" for a short period of time (around 1988) but resumed using the Rev. 0 layout:

One can distinguish the revision by looking at the second number (if present) on the MOS/CSG chips. For the 6522
it's either "10" (== Rev.0) or "11" (== Rev.1). This version can be found in 1541-II drives by example.

As the 6522 is plagued by the infamous shift-register bug we speculated if MOS/CSG fixed the bug but
withdrew that revision because of other incompatibilities or if it's even just a misprint!?
In order to find out finally I had to take a look at the silicon:

-

The result is that CSG just added a "(M) 1986 CBM" marker in the top left corner of the die AND
added a different (better?) ESD protection to the pins. The shift register control etc. doesn't seem to be different
at all (at least on the metal layer which I compared).

One can take a look at the bonding pad for the RESET# pin to get an idea of the change:

R0 v. R1:

-

[cjs]

fhw72 wrote:

As the 6522 is plagued by the infamous shift-register bug we speculated if MOS/CSG fixed the bug but
withdrew that revision because of other incompatibilities or if it's even just a misprint!?

You know, usually on the Internet we end it right there, in order to have wide open space for argument from ignorance.

Quote:

In order to find out finally I had to take a look at the silicon....

Nice work!

[kinzi]

Another MOS/CSG "special"

"Improve" something, do not fix the obvious, run the batch, trash it and continue using the buggy one.

Or the chip guys simply were feeling bored and possibly thought: "Hey, what do we want to improve today?"

@fhw72 Thanks for your efforts!