External Probe Sensor with odd DS18B20 fault values

[Alastair Comrie]

Hi Support, 

Since the recent server recovery we have noticed a number of external probe sensors are given odd values, which we assume are either fault values from the DS18B20, or switching briefly to the onboard sensor. The number of affected sensors may be due to more visual coverage of all our sensors, but the frequency appears to at least have increased. (We don't think the server issues are related, merely that it has brought this to our attention when checking up on all our many sensors.)

The odd temp readings are usually accompanied by a noticeable change in the battery voltage and often affects the reported signal strength also. In most cases the temperature swing is impossible i.e. going from 3.5°C to 70°C and back again within a fridge chamber in the space of 2 minutes.

A good example to look at is 7E5698715688 as it has occurred multiple times to both their sensors this week.  

Here are some screenshot examples of this (Drive link): 

Sensor 1: https://drive.google.com/file/d/1aC-vJpPHrBxLSS4x-QjqpAFuqJagB11O/view?usp=drive_link

Sensor 2: https://drive.google.com/file/d/1JwehOZpbmBuP-CZfl4g2epGIKRT32Wu2/view?usp=drive_link

Other examples of Tag Managers/Sensors it has occurred with:

EA2FF6E5895E - Drug Room Fridge @ Jul-28 6:45AM (almost identical to example 2 image)

6C870559FEE0 - Single sensor - Occurred @ Jul-17 2:00PM

D5180342C167 - Single sensor - Occurred @ Jul-18 12:14AM

We were wondering if there is some kind of filtering of DS18B20 possible fault values which is no longer being applied which might explain the uptick in these faults since the server was recovered? We do not recall seeing these patterns as often before the event, but obviously can't be directly affected by the issue.

Regards,

Alastair

[Brian Powell]

I've also seen some anomalous values from some Pro sensors since the server recovery.

The gaps around July 25 might have been antenna interference issues, but I don't have an explanation for the spikes/drops.

[Zhiheng Cao]

Please use the edit feature in web interface (scroll down to see the edit button) to remove these invalid data which was caused by the server outage last week. 

--
You received this message because you are subscribed to the Google Groups "Wireless Sensor Tags" group.
To unsubscribe from this group and stop receiving emails from it, send an email to wireless-sensor-...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/wireless-sensor-tags/826b6a3b-54ca-468e-9219-14eb5608e43cn%40googlegroups.com.