Myson boiler short-cycling / nest thermostat Y-plan system

[Jon Parker]

Evening all,

Pointers appreciated.

Myson Apollo 50B boiler, serviced annually. Set to 'hi' which I think outputs water at 80 degrees.
Nest 3rd Generation system. Gravity fed. Header tank, yellow cylinder in airing cupboard etc.

When we moved into the house we had an old Randall 103 programmer and no room thermostat which has since been replaced by the Nest system including the installation of a cylinder stat and Honeywell 3 port mid-position valve.

Around about the same time the boiler has developed the annoying habit of short cycling (turning off and on at intervals of between 30 and 60 seconds whilst the system is still calling for heat) after about 30 minutes of continuous operation on either heating or hot water or both (i.e. when it's been running for a while). It never did this before installation of the Nest system. I cannot say for certain whether the short-cycling started at the exact time of the Nest installation.

In an effort to solve this problem I have:

Balanced the radiators
Replaced the gas valve (service showed a slight leak)
Replaced the pump overrun stat
Replaced the 'hi' stat
Replaced the 'lo' stat
Replaced the thermal cut-off device

All with no effect.

With the heating turned up to 30 degrees to force it on and the boiler working in it's on/off short cycling mode I have measured voltage across Neutral and pin 3 (heating call-for-heat) on the Heat Link and with the system calling for heat this is at 240v as expected. When the heating is turned down the system stops calling for heat and it drops to 120v and then 0v. This behaviour is also as expected, although I don't know the significance of the 120v reading, if indeed there is one.

Crucially, with the system calling for heat the voltage remains at 240v when the boiler is in the 'off' phase of the on/off short cycling. In layman's terms (if my logic is sound) the system is calling for heat but the boiler is not complying.

A plumber mate has suggested my circulation pump but I'm not convinced. The pump L and N are connected to the boiler, which in my mind means the boiler controls the pump. Boiler off = pump off, boiler firing = pump running.

I'm looking for some next steps. Thank you kindly in advance.

Jon

[Bob Minchin]

Try turning up the pump to max speed (usually a three position switch)
maybe the new three posn valve has increased flow resistance and the
heat cannot get away from the boiler fast enough and the over heat stat
is cutting in.
BTW 80 c flow temp is ridiculously high so you are probably very close
to the over heat stat switching point.
Also the efficiency will be down as the boiler will rarely be in
condensing mode.

[Graham.]

Are there any ball or gate valves in the primary circuit that have
been left partially shut?



--

Graham.

%Profound_observation%

[Roger Mills]

It seems to me that there are two possibilities. Either the boiler is
cycling on its own built-in stat which has too little hysteresis between
on and off temperature, or the Nest thermostat is telling it to turn on
and off.

If this problem started when the Nest and 3-port valve were installed,
you should first make sure that the valve is wired correctly. Can you
trace all the connections and reconcile them with a Y-Plan wiring
diagram. In particular, the programmer (Nest in this case) must have a
HW OFF output which makes the valve's grey wire live when hot water is
not required. Similarly, the cylinder stat must have 3 contacts - with
the HW Satisfied contact again connected to the valve's grey wire.

I don't know about the internal logic of the Nest, but many digital
thermostats cycle their boiler output on and off when approaching the
temperature set-point in order to avoid overshoot. There is usually
something buried in an advanced menu somewhere which enables you to
specify minimum on and off times in order to minimise short-cycling.

With a Y-Plan system, when the HW demand is switched off and/or
satisfied, the boiler and pump are switched on and off by a micro-switch
in the valve - which connects the white and orange wires together. So,
in your case, the CH demand from the Nest will go to the valve's white
wire. That being the case, if the Nest is causing the short-cycling, I
would expect to hear the valve whirring away every time the boiler goes
on and off, because the microswitch only operates when the valve moves
to a different position.

When you talk about measuring voltages on Pin 3, is that in the box
which connects the Nest to the system? If so, that box is switching
mains (or not) to CH ON, and to one or other (but not both) of HW ON and
HW OFF. It is never *generating* 120v. However, if you are measuring
120v, this is almost certainly being fed back by the valve when it is
*not* receiving a CH demand. So, if that voltage is going from 240v to
120v in time with the boiler going on and off, it is pretty certainly
the Nest which is causing the short cycling - in which case you need to
see whether you can fine tune the way in which it operates.

Hope that helps!
--
Cheers,
Roger
____________
Please reply to Newsgroup. Whilst email address is valid, it is seldom
checked.

[tabb...@gmail.com]

It's the boiler's inbuilt stat that keeps stopping the flame. Likely cause is not enough primary cct water flow. But a misbehaving 3 way valve can also cause it. My first question would be how hot are the radiators?


NT

[Jon Parker]

On Thursday, February 2, 2017 at 9:59:38 PM UTC, Bob Minchin wrote:
> Try turning up the pump to max speed (usually a three position switch)
> maybe the new three posn valve has increased flow resistance and the
> heat cannot get away from the boiler fast enough and the over heat stat
> is cutting in.

Pump already at maximum speed.

> BTW 80 c flow temp is ridiculously high so you are probably very close
> to the over heat stat switching point.
> Also the efficiency will be down as the boiler will rarely be in
> condensing mode.

I'm now struggling to remember where I read that the 'hi' output temp was 80. I may have got that wrong.

Thanks for the suggestion.

JP

[Jon Parker]

On Thursday, February 2, 2017 at 11:56:44 PM UTC, Graham. wrote:
> Are there any ball or gate valves in the primary circuit that have
> been left partially shut?

Don't think so. There are valves either side of the circulation pump. I forget the name. I want to call them flange seals? Both of these are fully open as far as I can tell.

Cheers
Jon

[Jon Parker]

On Friday, February 3, 2017 at 3:41:28 AM UTC, tabb...@gmail.com wrote:
> It's the boiler's inbuilt stat that keeps stopping the flame. Likely cause is not enough primary cct water flow. But a misbehaving 3 way valve can also cause it. My first question would be how hot are the radiators?

Radiators get hot as usual. The 3-way valve is brand new and appears to be moving correctly when heating and hot water are turned on and off by the stat. When the boiler is doing it's short-cycling the valve doesn't move when the boiler shuts off.

This further lends weight (I think) to the boiler cutting out, rather than casting suspicion on the Nest.

Jon

[Jon Parker]

On Friday, February 3, 2017 at 12:53:04 AM UTC, Roger Mills wrote:
> It seems to me that there are two possibilities. Either the boiler is
> cycling on its own built-in stat which has too little hysteresis between
> on and off temperature, or the Nest thermostat is telling it to turn on
> and off.

Indeed. I'm fairly happy that the testing I've done rules out the Nest. When it calls for heat the terminal 3 (which is the HEATING ON terminal) goes to 240v and an audible click is heard as the relay closes, the valve moves and the pump spins up.

> If this problem started when the Nest and 3-port valve were installed,
> you should first make sure that the valve is wired correctly. Can you
> trace all the connections and reconcile them with a Y-Plan wiring
> diagram. In particular, the programmer (Nest in this case) must have a
> HW OFF output which makes the valve's grey wire live when hot water is
> not required. Similarly, the cylinder stat must have 3 contacts - with
> the HW Satisfied contact again connected to the valve's grey wire.

Yes, I can confirm all of that. My father-in-law and I installed the system ourselves, he an accomplished electrician of 35 years experience. I have opened up the junction box today and rechecked everything and it is all where it should be.

> I don't know about the internal logic of the Nest, but many digital
> thermostats cycle their boiler output on and off when approaching the
> temperature set-point in order to avoid overshoot. There is usually
> something buried in an advanced menu somewhere which enables you to
> specify minimum on and off times in order to minimise short-cycling.

I have been through every option, there is no such option.

> With a Y-Plan system, when the HW demand is switched off and/or
> satisfied, the boiler and pump are switched on and off by a micro-switch
> in the valve - which connects the white and orange wires together. So,
> in your case, the CH demand from the Nest will go to the valve's white
> wire. That being the case, if the Nest is causing the short-cycling, I
> would expect to hear the valve whirring away every time the boiler goes
> on and off, because the microswitch only operates when the valve moves
> to a different position.

It doesn't. During short-cycling when the boiler cuts out the valve stays put.

> When you talk about measuring voltages on Pin 3, is that in the box
> which connects the Nest to the system?

It's the Central Heating call for heat terminal.

> If so, that box is switching
> mains (or not) to CH ON, and to one or other (but not both) of HW ON and
> HW OFF. It is never *generating* 120v. However, if you are measuring
> 120v, this is almost certainly being fed back by the valve when it is
> *not* receiving a CH demand. So, if that voltage is going from 240v to
> 120v in time with the boiler going on and off, it is pretty certainly
> the Nest which is causing the short cycling - in which case you need to
> see whether you can fine tune the way in which it operates.
>
> Hope that helps!

I think it has. I'm convinced the Nest is not at fault.

I have the option of trying another circulation pump on a sale/return basis. I think this might have to be the next step.

Jon

[RJH]

On 03/02/2017 08:04, Jon Parker wrote:
> On Friday, February 3, 2017 at 3:41:28 AM UTC, tabb...@gmail.com wrote:
>> It's the boiler's inbuilt stat that keeps stopping the flame. Likely cause is not enough primary cct water flow. But a misbehaving 3 way valve can also cause it. My first question would be how hot are the radiators?
>
> Radiators get hot as usual. The 3-way valve is brand new and appears to be moving correctly when heating and hot water are turned on and off by the stat. When the boiler is doing it's short-cycling the valve doesn't move when the boiler shuts off.
>

How hot? 80C is too warm I think, depending on where you measure.

> This further lends weight (I think) to the boiler cutting out, rather than casting suspicion on the Nest.
>

I can't see how it could be the thermostat. Presumably it's just on/off,
with no fancy weather/humidity sensors causing the boiler to modulate or
cycle on/off?

--
Cheers, Rob

[Jeff Layman]

Many years ago my 5-years old Worcester-Bosch started cycling. It would
work for a couple of minutes, then turned itself off. After a few
minutes it would fire up, but then go out. The very competent engineer
we had used for years spent several hours over three visits trying to
fix it. He checked the wiring, replaced the FFD, and then the main
board. It still cycled. He even contacted W-B who couldn't suggest
anything else to try. He decided to replace the pressure sensor in case
that was faulty. As soon as he removed the silicone-rubber tube on the
sensor, all became clear. On the non-visible side there was a 5 mm
hairline crack. When the boiler was cold, it was closed, but when the
boiler heated up, and the metal tube it was connected to on the pressure
sensor expanded slightly, it was enough to open the crack and the drop
in pressure triggered the sensor to turn off. Cutting 10 mm off the
cracked piece of tubing cured the problem!

--

Jeff

[Roger Mills]

On 03/02/2017 08:10, Jon Parker wrote:
> On Friday, February 3, 2017 at 12:53:04 AM UTC, Roger Mills wrote:

>
>> When you talk about measuring voltages on Pin 3, is that in the box
>> which connects the Nest to the system?
>
> It's the Central Heating call for heat terminal.
>

I'm still not sure which terminal you mean. Do you mean in a 10-way
junction box often used for central heating systems? If so, that's
probably being fed by the nest, and in turn by the 3-port valve. I can't
see how that can be 120v when there *is* a demand for heat - so I'm
still not sure that you can rule out the nest.

Can you get at the demand terminal on the boiler itself. If you can
measure the voltage at that when the boiler cycles, it will throw a bit
more light on the subject. If that stays at 240v throughout, the boiler
is cycling on its own stat. If that goes off when the boiler does, the
boiler is cycling on its external control - i.e. the Nest.

[tabb...@gmail.com]

On Friday, 3 February 2017 08:04:23 UTC, Jon Parker wrote:

We know it's the boiler from info in the OP. Hot rads eliminates the momo valve as the possible culprit. You either don't have much hysteresis in the boiler stat, or more likely with the cold weather the boiler exchanger is just cooling rapidly, causing faster cycling. Old bimetal stat boilers didn't have much hysteresis or time delay built in.


NT

[tabb...@gmail.com]

Hot rads means you've got enough circulation. If you still suspect the pump, open it to see the impellor, it might be blocked. No point in a new one if it spins ok.


NT

[Roger Mills]

It's most likely the boiler stat - but you can't be sure of that until
you check whether the boiler demand stays on when the boiler goes off.

My Honeywell programmable stat (not in the Nest 'smart' stat category)
keeps the boiler demand on continuously when the difference between
target and actual temperatures is high, but then goes into
'proportional' mode as the target temperature is approached. This cycles
the boiler - using on/off times which can be configured - *before* the
target temperature is reached in order to prevent overshoot.

The Nest may do something similar - I don't know. It would be
interesting to know how close the room temperature is to the target
temperature when this cycling occurs.

[Jon Parker]

On Friday, February 3, 2017 at 8:21:58 AM UTC, RJH wrote:
> How hot? 80C is too warm I think, depending on where you measure.

I haven't measured, just recalling a section of the boiler manual which gives maximum output as 82 degrees.

> I can't see how it could be the thermostat. Presumably it's just on/off,
> with no fancy weather/humidity sensors causing the boiler to modulate or
> cycle on/off?

It does have a few fancy features but these are all off and I'm just using it as on/off.

Jon

[Jon Parker]

On Friday, February 3, 2017 at 9:28:17 AM UTC, Roger Mills wrote:
> I'm still not sure which terminal you mean. Do you mean in a 10-way
> junction box often used for central heating systems? If so, that's
> probably being fed by the nest, and in turn by the 3-port valve. I can't
> see how that can be 120v when there *is* a demand for heat - so I'm
> still not sure that you can rule out the nest.

No, I don't. On the Nest heat-link (which is the brain of the system and makes the electrical connections to all the other components via the junction box) Terminal 3 is the CH ON terminal.

> Can you get at the demand terminal on the boiler itself. If you can
> measure the voltage at that when the boiler cycles, it will throw a bit
> more light on the subject. If that stays at 240v throughout, the boiler
> is cycling on its own stat. If that goes off when the boiler does, the
> boiler is cycling on its external control - i.e. the Nest.

Yes I can do that, I'll report back once it's done.

Jon

[Jon Parker]

On Friday, February 3, 2017 at 9:35:42 AM UTC, tabb...@gmail.com wrote:
> We know it's the boiler from info in the OP. Hot rads eliminates the momo valve as the possible culprit. You either don't have much hysteresis in the boiler stat, or more likely with the cold weather the boiler exchanger is just cooling rapidly, causing faster cycling. Old bimetal stat boilers didn't have much hysteresis or time delay built in.

Okay, but why did it not short-cycle prior to the installation of the Nest and 3 port valve then?

Jon

[Jon Parker]

On Friday, February 3, 2017 at 11:25:35 AM UTC, Roger Mills wrote:
> The Nest may do something similar - I don't know. It would be
> interesting to know how close the room temperature is to the target
> temperature when this cycling occurs.

Sometimes it's as much as 1.5 - 2 degrees.

Also, in testing I have turned the stat up very high, like 25-27 to eliminate the possibility of any effect similar to the one you describe.

Jon

[tabb...@gmail.com]

On Friday, 3 February 2017 14:08:36 UTC, Jon Parker wrote:

> On Friday, February 3, 2017 at 9:35:42 AM UTC, tabby wrote:
> > We know it's the boiler from info in the OP. Hot rads eliminates the momo valve as the possible culprit. You either don't have much hysteresis in the boiler stat, or more likely with the cold weather the boiler exchanger is just cooling rapidly, causing faster cycling. Old bimetal stat boilers didn't have much hysteresis or time delay built in.
>
> Okay, but why did it not short-cycle prior to the installation of the Nest and 3 port valve then?
>
> Jon

I don't know. Possibles are colder weather, a 3 way that lets some water into the HW exchanger when the CH is on, more crud in the pump or a drifting boiler stat. It's not something I'd lose sleep over.


NT

[Jon Parker]

I have revisited the radiators and found there does seem to be a small drop in performance, and the radiators at the end of the circuit are suffering particular poor performance.

Jon

[tabb...@gmail.com]

On Friday, 3 February 2017 20:50:06 UTC, Jon Parker wrote:
> I have revisited the radiators and found there does seem to be a small drop in performance, and the radiators at the end of the circuit are suffering particular poor performance.
>
> Jon

Lack of flow then. I'd look at the pump.


NT

[Bob Minchin]

Jon Parker wrote:
> I have revisited the radiators and found there does seem to be a small drop in performance, and the radiators at the end of the circuit are suffering particular poor performance.
>
> Jon
>

You said the pump was at max already so I just wonder - assuming your
system has a bypass valve or it may have one rad without a TRV giving a
similar function, is passing too much through and not leaving enough
flow for the rest of the system.

I tend to close the bypass valve to start with and get everything else
working. then you should get the effect that the pump overrun time
extends to a long time as the boiler/pump is failing to get rid of the
excess heat. Open the bypass valve a bit at a time until the overrun
time is reduced a few minutes.

Slightly perversely, balancing the rads means constricting the flow a
bit at a time though the hottest radiators to increase the flow through
the cooler ones.
If you have an obvious hotter one then it is easy otherwise can be a bit
hit and miss until you start getting some improvements.

[Jon Parker]

On Saturday, February 4, 2017 at 11:29:27 AM UTC, Bob Minchin wrote:
> You said the pump was at max already so I just wonder - assuming your
> system has a bypass valve or it may have one rad without a TRV giving a
> similar function, is passing too much through and not leaving enough
> flow for the rest of the system.

Towel rail in teh upstairs bathroom has 2 LSVs on, these are both left fully open to act as a bypass. There is also a bypass next to the pump by way of a gate valve. This is in hte almost closed position, but I have experimented with different positions without any impact.

I've had my plumber / gas engineer mate round tonight. He's put a new pump in which is almost silent, which is a bonus. Unfortunately this had no effect on the short cycling problem.

So, we then wired the pump to a mains plug temporarily to force it to run, irrespective of what the boiler was doing. The result was that the pump ran continuously - obviously - and crucially so did the boiler.

We then closed off all radiators and opened them one at a time to fully vent everything. There was a bit of gurgling now and then and then all of a sudden there was silence. Each radiator (including the conservatory at the end of the circuit) got piping hot almost instantly.

We then opened up every radiator fully and again, all piping hot, very quickly.

I can only conclude that the pump overrun function of the boiler is now working properly, and perhaps my attention will turn to the PCB inside the boiler.

However, a contradictory observation was that when I measured the voltage at the boiler ON terminal (i.e. the one that fires the boiler up when it receives power) when the boiler turned itself off during a short-cycle the volatge dropped to zero. I assumed that the boiler was cutting out on it's safety thermostat and therefore the voltage would have remained at 240v.

With the now silent pump and satisfactory radiator performance if the pump is forced to run on I think I am going to plump for rigging up a run-on relay to the pump, which will have the effect of keeping it powered for 2-3 minutes (I will experiment with times to achieve continuous operation) after the boiler turns off.

This means that the boiler can continue to short-cycle if it wants to but the water will still be circulated around the system, and the constant on-off-on-off of my old pump will no longer keep me awake. This is important to me as I work shifts and sometimes sleep during the daytime.

Thoughts?
Jon

[Jon Parker]

On Friday, February 3, 2017 at 10:20:19 PM UTC, tabb...@gmail.com wrote:
> On Friday, 3 February 2017 20:50:06 UTC, Jon Parker wrote:
> Lack of flow then. I'd look at the pump.

See my reply in another sub-thread about my new pump and forcing it to run on by wiring it to a mains plug (as a test!).

Only things not investigated now are the valve (only 3 months old) and the cylinder stat (also 3 months old).

Jon

[tabb...@gmail.com]

A minute overrun is enough.


NT

[Roger Mills]

On 04/02/2017 20:13, Jon Parker wrote:

> However, a contradictory observation was that when I measured the voltage at the boiler ON terminal (i.e. the one that fires the boiler up when it receives power) when the boiler turned itself off during a short-cycle the volatge dropped to zero. I assumed that the boiler was cutting out on it's safety thermostat and therefore the voltage would have remained at 240v.
>

> Thoughts?
> Jon

Yes! If the boiler is cutting out as a result of the external demand
being turned off, it's not a boiler problem per-se.

What turns the external demand on and off? You've guessed it - the NEST.
That's where your problem lies. [Don't be misled by the fact (if it
does) that it displays a flame symbol throughout to indicate that the
demand is on. My Honeywell wireless stat does that all the time it's in
proportional mode - even when it's cycled the boiler off.]

I'd be surprised if the Nest isn't configurable for minimum cycle times
even if you haven't found how to do it.

[Jon Parker]

Apologies for not trimming. Using my phone to read and post.

I agree it feels like the nest but how can it be? Even if I wack it up to 30 it still cycles! That can't be right! And the CH call for heat wire stays at 240v!

In my previous post I mistyped "now working correctly" instead of "not working correctly" when regfering to the boiler's pump overrun function.

All my reasoning is telling me now there's an issue with the pump overrun function which I intend to solve with an overrun relay.

Hopefully that will be an end to my woes!

Jon

[tabb...@gmail.com]

On Saturday, 4 February 2017 23:44:52 UTC, Jon Parker wrote:
> Apologies for not trimming. Using my phone to read and post.
>
> I agree it feels like the nest but how can it be? Even if I wack it up to 30 it still cycles! That can't be right! And the CH call for heat wire stays at 240v!

then plainly it's not the nest stat cycling.

> In my previous post I mistyped "now working correctly" instead of "not working correctly" when regfering to the boiler's pump overrun function.
>
> All my reasoning is telling me now there's an issue with the pump overrun function which I intend to solve with an overrun relay.
>
> Hopefully that will be an end to my woes!
>
> Jon

NT

[Jon Parker]

This relay looks about right...

https://uk.rs-online.com/mobile/p/time-delay-relays/8966822/

Jon

[Roger Mills]

On 04/02/2017 23:44, Jon Parker wrote:
> Apologies for not trimming. Using my phone to read and post.
>
> I agree it feels like the nest but how can it be? Even if I wack it up to 30 it still cycles! That can't be right! And the CH call for heat wire stays at 240v!
>

I'm confused! In your previous post you wrote "when I measured the

voltage at the boiler ON terminal (i.e. the one that fires the boiler up
when it receives power) when the boiler turned itself off during a

short-cycle the voltage dropped to zero."

Now you're saying that the "CH call for heat" wire stays at 240v when
the boiler cycles.

I had assumed these two commodities to be one and the same. If they're
*not* please explain fully exactly what you mean by each term, and
exactly where you measured the voltages.

[Roger Mills]

That looks as if it should allow you to control pump over-run. I use a
less elaborate timer designed for controlling a bathroom fan to provide
pump over-run in my system.

But I struggle to understand how over-running the pump for longer is
going to solve your problem. Surely, if the boiler cools down faster
after it stops firing, its own thermostat will re-fire it *sooner* -
thereby *shortening* the cycle time?

How does the pump behave at the moment? When the boiler is
short-cycling, does the pump stop when the burner stops? If there's an
external demand, the pump should still run regardless of whether the
boiler stat tells the burner to fire. If the pump stops, it suggests
either that the external demand has been turned off (which would down to
the Nest) or that there's something seriously wrong with the boiler's
pump control logic.

[Jon Parker]

On Thursday, February 2, 2017 at 9:11:28 PM UTC, Jon Parker wrote:
> Evening all

Problem solved.

Took all of the wiring apart and put it back together again.

Everything now works. Not sure what's different but it works. No more short-cycling.

Thanks for all of your contributions.

Jon