Tim,
     
     I haven't heard anything, but my own use of the 'Count remain' and 
     'count per deg C' bytes has shown one oddity:-
     
     Although the Dallas block diagram shows the main temp counter 
     increasing when count remain = 0, checking the raw data out shows that 
     you can have 0 as a count remaining. Also you can have the value of 
     'Count per deg C' as 'count remaining'
     
     ie
     if Count per degree C  = 4a hex
     
     Count remaining can be 0 to 4a hex INCLUSIVE
     
     therefore using the equation 
     
     (Count per degree C - Count remain) / Count per degree C
     
     can equal 0.000 to 1.000 inclusive.
     
     If you code assumes it is always going to be 0.999 or less, it'll get 
     upset once in a while (probably showing a jump of a degree sometimes)!
     
                Hope this helps
     
                Steve Lawther
     
     


______________________________ Reply Separator _________________________________
Subject:      DS 1820 Silicon Problems ?
Author:  MIME:paul@SYMPAC.COM.AU at INTERNET-HUSKY
Date:    11/02/98 13:56


I have heard somewhere that the Dallas DS 1820 has a problems in high 
resolution temp mode (using counters for higher resolution)  We 
occasionally seem to get incorrect temp readings but with correct DOW CRC.
     
Has anyone heard of the problem or know of a solution ?
     
Tim Modra: tim@modra.com.au
     
Modra Technology: http://www.modra.com.au 
Custom Sensors and Embedded Control