Has my neighbour taken over my Drayton SCR?

[Jackrae]

As has already been suggested, the ID exists ('burned in') on the TX chip and these are probably like grains of sand - zillions of them

When you are pairing to a receiving unit, it will listen for all messages during the pairing period and then lock into memory the address of the TX unit it hears. Excuse the metaphors. From then on, the only messages it should respond to are those from the paired transmitter.

 

[DaPanda]

As has already been suggested, the ID exists ('burned in') on the TX chip and these are probably like grains of sand - zillions of them

When you are pairing to a receiving unit, it will listen for all messages during the pairing period and then lock into memory the address of the TX unit it hears. Excuse the metaphors.

There are two facts that are not consistent with this suggestion:

1) This started to happen recently and I hadn't matched the thermostat and receiver for years. So how come the SCR suddenly added the identity of this remote thermostat.

2) I re-matched my thermostat just after the remote thermostat had turned on and then turned off our boiler, i.e., at a time where it was not going to immediately transmit again.

(BTW it is unlikely for the id to reach zillion, it is most likely at most either 8-bit or 16-bit, i.e., 256 or 65536 possibilities so a collision is definitely possible.)

 

[bernardgreen]

A commonly used MicrChip device PIC16F15356

Address Range Standard Device Information
Microchip Unique Identifier (9 Words)
8100h-8108h
MUI0
MUI1
MUI2
MUI3
MUI4
MUI5
MUI6
MUI7
MUI8

72 bits, = 4722366482869645213696 different unique IDs

 

[DaPanda]

A commonly used MicrChip device PIC16F15356

72 bits, = 4722366482869645213696 different unique IDs

If that's what Drayton uses than I stand corrected. That would rule out collision possibilities though that was the most logical possibility based on the information you provided and my current reality.

Edited: A further thought, even if the id is 72-bit, it is possible that not all the 72 bits are transmitted and that would still make a collision possible.

There are no chips visible on the side of the back side of the PCB visible when taken off the wall plate. There was only partial visibility of the PCB when removing it as the rotating controller is clipped on top and also would need to be taken off which I didn't do. If I eventually do it, I'll let you know what chip markings are visible.

 

[DaPanda]

I started asking a couple of neighbours if they changed thermostat recently but just realised that if the range is typical as per the manual: '30m typically. The range may be affected by the composition / density and number of walls between the Digistat RF and SCR.', 30m is approximately four or five terraced houses. 30m in all directions - mostly two, along the road - and with all the flats in each, I'd be very lucky to find the neighbour controlling my boiler. So finding the offending thermostat is probably not a realistic option.

 

[DaPanda]

For those following this thread, earlier today I got an email response from Drayton and their diagnosis is that I likely have 'a failed SCR'. They are kindly offering me a replacement. If their diagnosis is correct, the replacement will fix things. If, however, my earlier guess is what's happening, the replacement SCR won't make a difference. So it will be interesting to see what happens. I hope Drayton's diagnosis is the correct one.

 

[Dan Robinson]

There's no documentation about doing this. I do not know if the matching process described by Drayton (in the documentation I linked to in response #7) erases the SCR's memory or not.

The first stage of the binding doesn't suggest an erasure so I would guess one binding action is transmitting an id.

As the receiver is pretty basic, it will only have one memory slot unlike a BDR with 4. I would summise that these products are no better than they were when we were fitting them 15 years ago before moving to Honeywell with a slightly better system.

Other than this, it pains me to agree with Bernie on the relative insecurity of the Drayton system.

 

[Dan Robinson]

But, fundamentally, you have a guffed SCR.

 

[DaPanda]

As the receiver is pretty basic, it will only have one memory slot unlike a BDR with 4. I would summise that these products are no better than they were when we were fitting them 15 years ago before moving to Honeywell with a slightly better system.

I'll give a go with the the new SCR when I receive it, but if the issue persists I'll likely replace it with a Honeywell system (they are much cheaper on eBay than the Drayton and appear to have a good reputation). If I have to do that I hope I can find some good instructions on how to change the wiring from the SCR to the BDR. If it comes to that I'll check my current wiring and then possibly ask for some help in a more appropriate thread.

 

[Dan Robinson]

Honeywell have a much better, but far from perfect system.

The binding process has an erasure procedure (for 95% of applications) and so time stamps the signal.

You'd have to be a very clever bunny with a lot of time on your hands to do anything mischievous with it.

 

[bernardgreen]

it pains me to agree with Bernie on the relative insecurity of the Drayton system.

You'd have to be a very clever bunny with a lot of time on your hands to do anything mischievous with it.

It is wireless, mischievous activity does not require a very clever bunny. Cloning a thermostat's transmissions is child's play. All the necessary bits can be bought over the counter in the High Street. Basic programming knowledge would suffice. Probably also available as ready to use items from dubious Internet sources.

 

[Dan Robinson]

Of course

 

[DaPanda]

Update...

I received on Saturday a new SCR sent by Drayton/Schneider Electric. I installed it when I came back in the night. I paired our thermostat (I had taken its battery out as there was no point fighting with the ghost thermostat) and set it to a low temperature. Sunday morning our flat was cold so I hoped it was fixed but also considered the possibility that our neighbour with the ghost thermostat may have gone away for the day and lowered their thermostat. As Sunday was cold I raised our thermostat and heated our flat. In the afternoon and evening I noticed on a few occasions that the SCR green light was off (and the boiler off) but the flame symbol on the thermostat was on. Basically the problem remains but this time I had set our thermostat higher than theirs. I've put back our thermostat to 10ºC and our boiler is currently working, obviously controlled by these neighbours who must be back home.

So contrarily to what Drayton/Schneider Electric customer care suggested, changing SCR unit to a recent one has not fixed the issue.

This does seem to confirm my diagnosis that the ghost thermostat and mine are viewed as one and the same by the SCR unit. They must send the same id (even though as discussed earlier they may have been programmed with longer different ids, the portion they send must be identical).

So one option is to get another Digistat thermostat and hope that its id would be different from the surrounding Drayton thermostats. It feels like Russian roulette. If Drayton wants to send me one, I'll try that solution, but I'm loathe to invest further in a system that seems broken. Another option is to replace the current installation with a competing system such as the Honeywell, which is the option I'll go for if I have to invest in it. However this will unfortunately most likely have to wait after the xmas holidays, and we may have a large gas bill if our neighbours with the ghost thermostat are staying in and setting their thermostat high.

 

[D_Hailsham]

As the new SCR hasn't solved the problem, you should go back to Drayton. It's their problem, after all, not yours. Trade Description and Sale of Goods Act etc (not fit for purpose and not of merchantable quality).

 

[bernardgreen]

Does your boiler have a Frost Stat ? If it does could that be switching the boiler on.