In message <199709230155.LAA28985@balrog.supp.cpr.itg.telecom.com.au>Simon Burg
e writes
>On Mon, 22 Sep 1997 17:20:34 -0700  Phil Tait wrote:
>
>> I see that "jonathon" fixed the character-doubling on OUTPUT problem by
>> adding DELAY(2) after SCC_WRITE DATA() back in 1995 (see excerpt from
>> source-changes at end of note).
>> 
>> I'm seeing a character-doubling problem on INPUT (reading an instrument
>> at 1200baud).  I'm looking for informed opinions as to whether it's
>> appropriate to assume symmetry, and add DELAY(2) after SCC_READ_DATA();
>> or maybe someone has a better idea for a fix.


the data book i used mentioned a problem on the output side; i didn't
see anything about a similar settle time required on the input side.
But the book was for an Intel second-sourced part, not a Zilog chip.


>You didn't say what version of NetBSD you're using.  This problem was
>fixed since the release of NetBSD 1.2.  If you're running a fairly
>current kernel, you shouldn't be seeing this.  The relevent bit to look
>at in scc.c is a few "DELAY(2)"s.  Let me know if you need more info on
>this.

Phil,

just out of curiosity, are you running NetBSD or OpenBSD on this
3MAXPLUS? There were some timing-related changes which OpenBSD may not
have picked up yet.