More boiler woes

It could be the micro switch in the motorised valve has failed / becoming unreliable.

Normally, when the programmer has been set to turn on the CH, it sends power to the stat, if the stat is calling for heat, it then sends power to the motorised valve, the valve, when open, then sends power to the boiler and pump, that then turns on the pump and boiler. If the micro switch in the valve has failed on, then even with the programmer and stat off, the boiler can still fire.

Exercising the valve can make it work again, but it may have become unreliable, so may need it's head, or the complete valve replaced, if the valve side of it is getting stuck.

How old is the valve?

Toby...

Only if the valve microswitch is fed live mains from somewhere that isn't switched. If it's fed via the programmer and stat then it can only power the boiler when the programmer is "on" and stat is calling for heat.

Though looking at how I have configured my system the valve switches are fed with live mains rather than through the programmer and stat. I might change some jumper positions in the wiring box to stop that incorrect failure mode. All heavy power switching for the boiler and pumps is done via relays so the programmer and stat contacts only have to carry the tiny relay operating load not the 150W boiler and 50W or so pump load. Even so programmers and stats tend to be 3A inductive rated so switching directly shouldn't be a problem.

The message from "Toby" contains these words:

I can't remember.

There are 3 valves on the system - upstairs and downstairs heating and hot water. Over the years I have had to replace one valve complete and more than one motor but I can't remember which or when and can only trace one transaction, new motor in 2001. That motor could easily be the last purchase. The upstairs valve can't be a problem atm as that hasn't been operated at all over the summer but which of the other two could be at fault is moot.

Given that for this particular problem the power must be getting through regardless of the valve motors being on or off is there any other failure that could have the same results? Otherwise I am tempted to change both valve heads just in case always assuming I can find heads to fit. I am not sure who the manufacturers of the current valves are and may not be able to find out.

I replaced the micro switch in one of my valves recently, easy job if you have a soldering iron. Maplin stock the switches for a quid or so.

You clearly have an S-Plan-Plus system with a separate valve for each zone. Each valve motor will be fed from the programmer and appropriate stat. Inside each actuator is a micro-switch which closes when the valve is fully open. These micro-switches have a permanently live feed, and switch on the boiler and pump whenever one or more is/are closed.

If the system is now behaving normally - and switching off when all demands are satisfied - it indicates an inermittent (rather than permanent) failure of the micro-switch. However beware that it might *appear* to be working normally if there is a demand from the zone with the failed switch. The acid test comes when *all* demands are satisfied.

The message from "Dave Liquorice" contains these words:

My microswitches are wired to permanently live.

The downstairs CH valve is a Danfoss. Warranty expired Jan 07.

The DHW valve is a Tower. Might be relatively modern as it has no earth. (OTOH they might have been making them like that for donkeys years).

Currently working normally. I think I might leave the DHW off for a few days and then vv with the CH to see if I can isolate the fault but sods law says I will not get a reoccurrence until both are back on.

The message from Roger contains these words:

As I thought, sods law in full operation.

However when the fault reoccurred I managed to isolate it to the Danfoss by the simple expedient of removing the programmer and unwiring one side of the microswitch on each valve. Of course the wiring centre the Danfoss was wired into was underfloor so it took a bit of effort shifting furniture and lifting part of the floor.

Which is the way it should be to prevent any feedbacks and not to put heavy loads on the programmer and stat switches.