Once again I have a problem with the hot water sensor on my system.
The CS92A is definitely the most unreliable part of an Evohome system.
Not sure what the cause is yet as I'm at work but I switched the heating on remotely for my other half who is going home early (another story - the total connect app frequently signs us out on our phones and she doesn't know the password to re-enter so it was easier for me to just do it) and not long after that the hot water temperature is "NaN", and my graphs show no updates from the sensor for over 3 hours. The rest of the system is working fine but I know the hot water will be luke warm when she gets home as the system will refuse to heat it when when the temperature reading goes to "NaN".
I can predict what the situation will probably be though from the past 10 or so times it has happened in the last year or two alone. There will be no reports of low battery on the controller, just the report of no communication from the sensor and an hourglass for the temperature. If I press the signal test button on the sensor it will give me a 4/5 or 5/5 result. If I remove the batteries and refit them it will miraculously start working again, and if I measure the voltage of the batteries they will be fine and well within spec for Alkaline batteries.
I've already replaced the CS92A once under warranty (the first one was much worse) and relocated it twice. I've tightened the battery terminals behind the PCB. I always keep fresh batteries in it and don't let them get too low as I know it's unusually sensitive to low voltage. I think next time I change the batteries I'll fit Lithium to pander to its unusual sensitivity to anything other than brand new 1.6v battery voltage.
I think most of us can agree it's fundamentally broken in design and has systemic problems as all the hot water sensor problem threads here attest, including a thread of my own where I analysed the problems in great detail a couple of years ago around the time I got a warranty replacement. Please Honeywell, if you do nothing else, fix the design of the CS92A. It really is unreliable!