I have all my lights controlled by Fibaro Dimmer 2's located behind MK grid switches. Upstairs the walls are stud and plasterboard, so the back boxes are plastic 'dry line' boxes (as we call them in the UK). Downstairs several of the walls are solid brick or block and so the switch back boxes are metal. I always thought it would be better if they were plastic but couldn't source a plastic box that would go in a solid block wall, but then another forum user @BorrisTheCat pointed out there was a plastic box available and I thought I'd switch them out.
I picked one of the ground floor switches where the signal was fairly poor - Direct 8dB LWR RSSI and set about carefully removing the metal box. I've replaced with plastic, re powered the module and sent a few signals. Not a good result - the LWR RSSI was 7dB at first and has now dropped to 4dB. I guess the metal backbox was actually aiding the signal by reflecting it out of the switch, whereas with the plastic, perhaps more signal is being absorbed by the block wall.
Needless to say I'll hang fire before attempting to improve the signal in any more locations (I had another 4 I was going to swap out)
Certainly runs counter to the prevailing wisdom... IIRC the LWR RSSI being reported (for Z-Wave) is the minimum value in each hop of a multi-hop path; is it possible that with the previous installation, the RSSI value reported was that of an intermediate repeater (and not representative of the signal strength of the device in the outlet box)? Maybe inside the plastic box, the device now has further effective range, and the RSSI value now is that of a different repeater in the route... or maybe the device itself is now connecting directly (eliminating a repeater hop).
Oh great, just what I needed to hear after swapping out 4 or 5 of my metal boxes.
So in the US I am aware of 2 types of "plastic" electrical boxes. One is a lighter weight blue colored PVC,and is able to be flexed by hand. The other is a grey, very heavy duty, box, with almost zero flex, it's apparently some type of fiberglass?
Which type box was your experience with? And has the connect speed changed or just the RSSI?
I don't have any fibaro, but do have Qubino Flush 2 relays, and do my best to place the antenna wire as close to the switch faceplate as possible, even if having to rewire, so the antenna exiting from the relay is as close as possible to the faceplate. I don't know if it helps, but it makes me feel like it's helping
In his case he has block walls and probably has rebar in there as well. Generally you won't have signal issues with either blue pvc or the brown/gray fiberglass ones.
Where have you put the arial now? Also note that it's likely when you down powered the device others might have moved around. This happened to me and it took a few weeks for it to settle, once it did it was better of though.
I'm in the US. Our house has 90% metal boxes. We use plastic switch plates. We've had no issues with Z-Wave devices in these boxes. We have one Zigbee device in a metal ceiling box with a metal ceiling light that is 18" in diameter. We've had no issues but it is only 10' from the hub.
Boxes here are I'm guessing ABS or Polyprop. I've not used this type before as I wasn't aware you could get them. Our UK standard box for plasterboard walls (the 'dry line' box) is quite flexible. This one is more solid. Just the RSSI changed, the connection path (Direct) and speed of 100 is unchanged.
Internal solid walls in the UK from oldest to current tend to be single course brick, breeze block or lightweight, soft thermalite block. We wouldn't get rebar in house walls. This particular wall is breeze block.
One interesting thing I was told some time ago that may/will affect radio signals... Where I am in the North East UK is historically famous for it's iron and steel industry (now gone). Apparently the blocks used may have a high content of iron/slag particles in them, which was waste from the works. When you chase these walls to fit boxes, the removed material is filthy and black.
The module and aerial are positioned similar to before. Yes you're quite right, it's difficult to do a like for like compare. The three RCBOs I isolated to do the work will have down powered approximately 10 Z Wave modules.
What I do usually find after a power down is that nothing changes path immediately. The modules will use the same path as before power down and tend to make route changes once they receive digital commands from the hub. I usually open a dashboard and switch each device manually from there to update the 'unknown' entries in the Z Wave mesh table.
