Just an update

I'm continuing to use this test for an address match

         movfw   sspbuf       ;test received address
         xorlw   b'00001000'  ;against device address
         bz      header       ;match

rather than

         btfsc   pir1,sspif   ;test for address match
         goto    header       ;yes

and so far, after 2 weeks and much data sent from master (F452)
to slave (F88), no problems. I've tried quite a few configurations
of the F88, most of which don't suit this application so I'd never
use them in practise, and there is still unreliability with SSPIF

If anyone is able to reproduce the problem I'd feel better, even
though I'll stick with the XOR

I have an open ticket at customer support. The response from
Microchip so far has been just

========================
There's nothing on this in the errata document for the device.  Here's a
link to it:

http://ww1.microchip.com/downloads/en/DeviceDoc/80171j.pdf

What is the address that the 88 isn't responding to?  It looks like it's
properly ACKing the address"
========================

but that was 12 days ago and they've not responded to further
correspondence. I gave full details of the situation (including that
the F88 shows this with ANY address) and problem and asked
specifically if the F88 is undergoing a revision, but not a peep
from them

-- 
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist