Quick update, things appear to be working! It's now been several days with no overshoots, so what did I do?
- I removed and reinserted the insertion probe into the water cylinder to ensure it was fully seated.
- I moved the EvoTouch controller into our Living Room so that it is almost directly below the airing cupboard where the water cylinder is located.
At this stage, I'm going with moving the controller having made the difference, why? When I looked back through the temperature logs, I noticed the overshoots didn't start appearing until I moved the controller into our hallway onto the wall amount. At the same time as wall mounting the controller I also tidied up the temporary wiring in our airing cupboard, and also wall mounted the BDR91's, and CS92A controller so lots of changes at once!
I'm going to leave the controller as-is for another week or so. I'll report back again then with my findings. Oddly, which is why I'm still slightly dubious, I'd already tried moving the controller onto our landing so that it was a lot closer to the airing cupboard before but this appeared to make no difference. Maybe the signal is travelling better through the floor.
This fits in with what I've observed - unreliable reception of the sensor by the Evotouch.
Unfortunately in my case they are only about 5 metres apart with one solid wall and one wooden frame glass door in between. If it can't work reliably across this distance I don't know how it could ever work reliably in larger houses...
I also always get a 5/5 report for the signal test mode...
I've just bought a spare controller (with a broken screen, so I'm just going to pull it apart). My intention is to see how feasible it is to add an external 868MHz antenna. Or even a trailing wire of the right length.
Incidentally, although I bought it on eBay I've only just noticed that it was sold my the owner of these forums!
I'll report back if I have any success with a suitable mod.
My theory is that one mod improving the comms on the controller is a lot easier than lots of mods with CS92s, HR92s, etc. Everything goes back to the controller.
Last edited by paulockenden; 14th September 2020 at 04:51 PM.
Has anyone actually considered that the issue might be with the controller rather than the CS92A?
Seeming replacing the CS92A doesn't solve the issue has anyone tried an alternate panel with a problematic CS92A.
Mine has never showed a comms fault, and happily is still working perfectly...
As distance does not appear to be a reliable predictor, could the problem be interference or crosstalk?
In my case I think it is indeed the controller to blame. I have had the CS92A replaced and also have moved it several times, never has it been completely reliable.
However when I've compared the response of the controller (watching the temperature reading change during a hot water reheat and also looking at the fault logs) against my little Raspberry Pi / NanoCUL based monitoring system located near the controller, it is receiving the CS92A messages which the controller isn't, which also points the finger of blame more at the controller and less at the CS92A.
Whether the controller just has a poorer receiver or antenna than my NanoCUL or something else is at play I don't know. I like Paul's idea of extending the antenna in the Evotouch - mine is long out of warranty so its something I'd be willing to try if it's a fairly low risk procedure...