Between 2022 12 20 05:45 and 2022 12 29 14:31 the reported temperature changed from a stable accurate value of about 10 degc with negligible spread to nonsense values ranging between -20:80 degc with huge spread(changing sample to sample) AND INCIDENCE of nan. it was a real effect with temperature faults reported. it the reverted to normal behaviour. ??? candidate is a bad firmware update with say a units mistake eg float rather than int? ideas