DIY Doctor UK DIY Forums

The boiler is a baxi system boiler 60/100, the heating is an unvented sealed system, with open vented hot water cylinder and tank.

External controls are;
Siemens RWB9 programmer (working), Boss therm room stat(eliminated),
Boss therm cylinder stat (working)
Newly replaced Honeywell 3 port valve, old was Boss therm with same wiring.

The problems with this system have apparantly been from day one, boiler is around 7/8 years old.

To start with the old 3 port valve had been placed in manually open position, probably to disguise the fact when in auto it didn't work, Honeywell replacement now fitted, which has cured that part of the problem.

Then 2 non working radiators have been cleansed of sludge and the system cleansed and is now hotter than ever.

Really everything works fine until the programmer is put into CH on mode,
after that nothing will turn it off, except the fuse spur.
So my thinking is to try and find what is holding the boiler on, but i'm running out of ideas.

To clarify, if boiler is then switched back on, with programmer off, boiler behaves correctly and nothing happens, with HW on, boiler fires, diverts correctly cylinder gets warm, cylinder stat turns boiler off or programmer turns boiler off, only when CH is on, and then switched back off, the boiler keeps going!

Any ideas, questions, points i've missed, please post.

Have you checked the wiring of the three port valve?

In a Y plan, when heat is called for by the timer and room stat, power is supplied to the valve motor. Once the valve is open, it operates a micro switch which in turn supplies power to the pump and boiler.

If the pump and boiler refuse to shutdown, this possibly points to a problem with the valve.

For sure this has to be an electrical wiring problem, could be that there is a backfeed to the boiler some how. Is there a wiring centre? This would look like a junction box where all the wires would feed back to. If so look at the Honeywell web site Y plan wiring and make sure that what you have matches this diagrame. Remember that every thing will be 230v which can kill, so if you don't know electrics call in a electrician.

Thanks for your reply, the system is a bodge but as far as I can see matches up to the diagram I have from baxi,



The wiring centre is boss therm and is in the airing cupborard, so terminal block numbers don't match up but wires seem to. The fuse spur and boiler terminal block are located under the house so don't wire directly into the wiring centre, and the pump inside the boiler itself.

Something must be keeping the switch live open, but only after CH has been switched?

Been back today with an electrician and multimeter,

Found that the orange wire from the honeywell 3 port valve, (the switch live), is not completely losing voltage when CH is switched off,

The readings on this wire go

Hw on = 240v / Hw off = 0v (as expected)
CH on = 240v / CH off = 100v which is keeping the boiler firing

So for the time being have set Hw to run 10 mins longer than heating which ought to save switching off at spur.

Anyone having any ideas what is causing this?

Assuming you are not unlucky enough to have a short within the programmer, I would suggest drawing out a very detailed plan of the wiring especially that within the programmer. The next stage is to check out the actual wiring wire by wire with the plan. Label each wire and check that its other end is the one you expect. With bodgers, you can get some very strange connections. Again, if possible, check the programmer gives the expected results as you move it thru its sequences.
How has the room stat been eliminated and what has happened to its wiring? Without a room stat, how do you expect the CH to turn off when the room temperature rises?
This is a job that must be done with a torch battery and a bulb ( or multimeter) and definately not with mains voltage. Disconnect the mains cables from the boiler having first removed the fuse and put it in your pocket.
If you change any wiring, test the system before making any further changes. Think carefully what you are doing before you short circuit any components. Remember your personal safety is paramount. If in doubt get a qualified electrician to sort out the mess.

Cheers guys, feel I'm getting close now.

Just got a fault finding e-mail from honeywell, it states that a.c. voltage may appear on specified wire or terminals due to back feed from Three port mid position valve.

What are the possibilities of this back feed passing through the cylinder stat & making the orange wire live?

I'm going back tomorrow night, so will wire the stat out to test this.
If it is found that voltage is passing back through the stat, what can be done to rectify the problem?

Your thoughts please....

It is really difficult to offer advice at a distance.

If I was there, I would check the following:-

1. When the fault is occurring and the heating being pumped, I would remove the programmer from the backplate. The heat should immediately go off. IF it does not, it points to faulty boiler circuitry.

2. Whilst the programmer is removed, check if there is a cable link to 'pin 1 on the backplate. If not check pin 2. If there is a cable to either, it must be to pin 1. (It could explain your problem.)

3. Ensure the white wire from the 3pv is connected to the room stat and the grey to the cylinder stat. (Are these wires transposed?)

4. Check that the blue wire from the 3pv is at ground potential.


It seems strange that you have 100 volts in the orange wire. From what information I have, I can't see what could be dragging the voltage down to 100v and then have the current to keep the pump active. Are you sure it's not an inducted voltage, or a back feed from the switched live from the boiler?

If you disconnect the orange: what stays live? The orange or where you took it from?

Had no joy this evening, went back and wired out the cylinder stat, and found the same problem, :cry:

In answer to marrtin

1. Removing the programmer from the backplate DOES cut the boiler out and make it stop,

2. There are no cable links wired to the programmer.

3. The white wire from the 3pv is connected to the room stat and the grey wire is connected to the "satisfied" connection on the cylinder stat.

4. Have not checked the blue wire on 3pv for ground potential, what problems could this cause?

Very strange I know, the 100v reading on the orange wire has been taken with orange wire disconnected from wiring centre, both the HW off on the programmer and cylinder stat are seperately able to reduce this to 0v, but CH off on programmer or roomstat, which have been told doesn't work anyway, isn't able to.

Thanks for your advice, hope theres more...

3 Plumbers have either walked away or contributed this systems problems in the past. But I am in this to the end now and will keep posting updates.

I would be grateful for clarification on the following
Which model of Baxi eg Solo etc
Siemens RWB9 programmer (working), The only RWB9 I can find is by Landis & Staefa.
Boss therm room stat(eliminated), What has replaced this, what are the colors of the wires. what has happened to the old wires?
Boss therm cylinder stat (working) . Please give details of model number. Which colored wires are connected to which terminal
Newly replaced Honeywell 3 port valve, old was Boss therm with same wiring. Please give reference number of Honeywell.

Wiring centre. Please list the color of wire attached to each terminal as follows. Identify any internal links eg 5 linked to 7.
Terminal six: white with blue tracer from cable from pump.
Please indicate whether or not you are sure of the source of each cable
Usual precautions on electrical safety.

A negative fault will often give unexpected voltages in unexpected places, but having thought about this more, it may be irrelevant.

The mv contains a ballast resistor between the grey, white and orange wires so 100v may very well be a normal result.

In a Y plan system, there should be 3 control wires from the programmer to the wiring centre. One each for dhw on and ch on. The remaining wire is from pin 1 on the programmer backplate and feeds 230v to the grey mv wire when the dhw is timed OFF.

If this link is omitted, the valve may 'stick' in the ch on mode which would only clear if the dhw stayed on longer than the heating.

If the dhw and ch are timed off, and the only +ve from pin 1 is not there, how is the power passing through the programmer to keep the boiler live?????

I suggest you have a closer look at programmer wiring.

Thankyou to you all who answered, I think the solution has finally been found :D

It hadn't occured to me to check this as the voltage was present on the orange wire of 3pv when removed from wiring centre, but it seems it was the frost stat that was keeping the boiler fired!

I quickly went back last night after looking through the manual and found if the switch on the facia of the boiler is in hot water only mode (its a baxi system boiler), this bypasses the frost stat which is integral to the appliance, and everything worked and turned off as it should.

No calls yet anyway, they said they would let me know either way....

I know exactly what is causing this problem. I am on my tenth valve from Honeywell and I have just written to them to ask how they expect this valve to work. Briefly, when the white wire is energised the valve motors to the mid position (Htg and HW on) and stops when a cam operates a microswitch which cuts off the 230v and supplies 60v to the motor, which is just enough to hold position against the spring return. It does this by a internal printed circuit board which has a diode and two resistors. The orange wire, which is supposed to power the boiler and pump, is supplied via the two resistors, which lower the voltage to between 80v and 100v.
If the HW cylinder thermostat is calling for heat, this also powers the orange cable but at 230v. If the valve is already partly open, then that 230v from the HW thermostat backfeeds to the motor and holds it in position, and the cam keeps the microswitch closed, thus the heating is on whether to room thermostat or programmer is calling for heat or not.

Wow ! I have just read this again and it seems really complicated, which it is. I was a lecturer in control theory at a tech college and I can't see how Honeywell expect this to work - but they sell thousands and I don't know how widespread these problems are. Anyone out there with a solution ? I stripped down a valve and made a drawing of the internal wiring. How do I post this on this forum ?

That is an excellent description arni, and after long hours staring at wiring diagrams and reading through honeywell fault finding pages, is something like what I was thinking could be happening.

If you could somehow post that diagram I would like to see it.

What I don't understand now is, why switching the mode on the front of the system boiler has solved the problem, I've just been back to balance radiators and everythings fine! Maybe I'll never know.

This forum does not always allow links to external websites, although I have seen some posted, and have even managed to have some published myself, other messages have links edited out, and whilst other posts have been deleted entirely.

I understand you can post a link to an image already posted elsewhere on the net.

I took an old valve apart myself to try and see why the orange wire was held at 100v. I would like to see your diagram. I searched the web for internal information, but to no avail. I see what you mean about the backfeed from T1 of the cylinder stat, but the scenario does seem to work in all circumstances I have seen.