Is the 1wire input not CRC checked?

LarrylLix

Senior Member
Doing further testing and removing the +5v wire on a probe to the terminals I must have glitched the input to the 1wire input.
 
The probe AIP2  that I removed locked onto  an input reading of 85.0 C. The probe read room temperature, before the disturbance, and I understand the WC locks onto the last known reading received, but this seems to indicate it read erroneous data from the 1wire line but didn't reject it.
 
Does this mean the CRC of the 1wire inputs is not CRC checked so bad date could be read from probes with noise in the data?
 
Larry,
 
DS18B20 default register value is 85C, if the measurement was unsuccessful.  Please look page 7 Figure 7 in DS18B20 datasheet.
WC always do CRC for each reading.  When you removed +5V, the conversion will fail, so that register contains its default value 85C.
 
Basically, when you removed +5V, device through parasite mode can still communicate with WC8 board.  However, temperature conversion requires to have 5V constantly on at pin3, or WC8 can not talk to any other sensor during conversion time.  Because we want to speed up the measurement on all sensors, while one sensor is doing temperature conversion, we talk to other sensors, so that they can not work on parasite mode.
 
Hope this helps.
 
YES!
 
Thanks for the education!
 
 Sounds right on the money!
 
Time to modify code, adding trapping decision code.
 
Yes, different sensor default value is different. You may check them out by disconnecting the power pin to them, only let 1-wire bus commands talk to the sensors, not guaranteed always work for all 1-wire sensors in that way, but it is good to read back default value so that you can detect power wire broken.
 
DS1822 power on default value is also 85C.
If power pin is not connected, that is probably the value read in by WC8.
 
Back
Top