OpenTherm control behaviour

[kimber.kimber]

Sorry, yes I did get that, just thought it would be another thing to try.

TBH it's difficult to tell! We have installed cast rads, so they stay hot for quite a while. This morning while the boiler was firing, the room that had its temperature near its set point, had a cold rad, and the inlet pipe was also cold. I didn't get chance to check the others.

In the mean time reducing the set point has seemed to have stopped the boiler firing.

[HenGus]

That's loads of information on the valves too! Will definitely have a look at that. Is there a way to force them to do their calibration again? I've literally had a whole new heating system, boiler, tank, rads and pipe work, so it shouldn't be sticking valves. I fitted all new Honeywell Valencia valves at the point of instal.

Take them off the valve body; turn the adapter fully anti-clockwise and replace the HR92. You will then see CYCL as the valve motors open and closed.

[kimber.kimber]

Well with the set point dropped, the boiler hasn't been firing at all. So looks like that may have got it. I probably need to adjust the low set point to a temperature just below the rooms natural resting temp. We've only been in the house since a week before Christmas, so we're still learning its intricacies!

Even with a little bit of fiddling, it's still going to be better that what we're taking out...

This for heating... (no thermostat fitted at all)
IMG_6908.jpg

And this for hot water... (gas and on 24/7)
IMG_6909.jpg

Think the gas man might be round to check I've not been fiddling the meter when we're all up and running!!

[kimber.kimber]

In case anyone else finds this and wants to read, here's an article by Honeywell about TPI...

http://www.honeywelluk.com/documents/All/pdf/TPI%20Functionality%20explained.pdf

[HenGus]

In case anyone else finds this and wants to read, here's an article by Honeywell about TPI...

http://www.honeywelluk.com/documents...0explained.pdf

I bet you can't find a description of Honeywell Adaptive Fuzzy Logic.

Other than:

A fuzzy control system is a control system based on fuzzy logic—a mathematical system that analyzes analog input values in terms of logical variables that take on continuous values between 0 and 1, in contrast to classical or digital logic, which operates on discrete values of either 1 or 0 (true or false, respectively)

[kimber.kimber]

I bet you can't find a description of Honeywell Adaptive Fuzzy Logic.

Other than:

A fuzzy control system is a control system based on fuzzy logic—a mathematical system that analyzes analog input values in terms of logical variables that take on continuous values between 0 and 1, in contrast to classical or digital logic, which operates on discrete values of either 1 or 0 (true or false, respectively)

Clear as mud... clear as mud!

[StephenC]

It's probably related to the pin calibration of your TRV valve bodies, and the "minimum on time" setting which you get with a BDR91 but not with the OpenTherm bridge. I have seen similar symptoms (all radiators cold but the occasional cycling on of the boiler) on rare occasions even on a BDR91.

What I believe happens is this:

The HR92 has a certain range of pin travel when it opens and closes the valve. When you fit it to the valve it does a calibration where it pushes the pin down hard until a certain amount of force is reached and this is calibrated as the fully closed position. (0% on the reported valve position in option 10 in the menu) It then opens the valve a certain number of turns of the gear inside the HR92 (a greater number of turns in full stroke 1) or until it hits the limit stop. This gets calibrated as 100%.

This percentage is a proportional indication of how far the pin is pushed down or released with 0% being pushed right down. The HR92 sends this valve position percentage to the controller as a "heat demand" percentage which the controller then uses to determine whether there is sufficient demand to warrant firing the boiler, and if so by how much. (TPI duty cycle using a BDR91 or flow temperature using OpenTherm)

Here's the underlying issue - the HR92 doesn't really know for sure at what point of pin travel the valve actually starts letting water flow. In the fully closed position the rubber washer in the valve will be compressed. As the HR92 unwinds and lets the pin move out quite a lot of travel will happen where the washer is decompressing but still sealing the valve, therefore no water will flow.

From my observations it seems to be typical (at least on MY valves) that the pin will reach about a 30-40% reported opening (based on total calibrated pin travel) before any water starts to flow at all, and typically will be fully flowing by around 80%. Because of this the controller seems to be programmed to ignore heat demands of less than about 40% and won't fire the boiler until at least one zone exceeds this.

However what happens if due to valve pin dimensions, washer thickness, valve stickiness etc your valve doesn't let water start flowing until "60% open" ? It could very well be that the heat demand is sufficient to fire the boiler intermittently while the valve is not actually letting any water flow. And the system has no real way of knowing this. As far as it is concerned the HR92 is open enough that there must be some flow through that radiator, but there isn't.

If you use a BDR91 there is an additional factor - the minimum on time setting. By default its 1 minute (and can't be set any lower) and for the default 10 minute cycles it means that even if the controller thinks a valve is open enough to warrant firing the boiler, it won't do so until the required output from the boiler is at least 10%, and would thus exceed the minimum on time. So this would tend to avoid any brief firings where it might think a valve was slightly flowing when it actually wasn't.

With OpenTherm I don't think there is any minimum on time or minimum demand, so it would tend to try to fire the boiler with a very low flow temperature. (30 ??) This would bring the boiler on for the short time needed to get the flow up to 30, and probably leave the pump running a lot.

What can you do about this if it is firing a lot when radiators are not flowing ?

The first thing I would do is remove each HR92, unwind the black wheel and refit the HR92 to allow it to re-calibrate itself. Then give it a few days for the valves to fine tune their calibration.

If that doesn't help, next time you see the boiler intermittently firing when all radiators are cold I would nip around each HR92 and go into the settings, look at option 10 (valve position) and record what percentage it believes the valve is open. If you have a figure significantly higher than 40% and the boiler is heating up and the radiator is stone cold I would be suspicious of that valve body.

You could try enabling full stroke 1 mode in the HR92 settings, which increases the amount of torque to close the valve and also increases the number of turns of the motor between closed and open. The disadvantage of this mode is the batteries don't last as long and there is more tendency to get temperature overshoots, however on a couple of my valves I've been forced to use full stroke 1 mode.

The valve could also be sticky (causing hysteresis in the opening point depending on whether the valve is closing or opening) or it may not be 100% compatible with the HR92's due to for example a pin length that is a bit longer than average. Or if you are using an adaptor the adaptor might not be sitting right.

Thanks for the great ideas and info.

Let me digest this and try some of it out before I switch back to BDR91 Relay control next week. (Opentherm just isn't working for me with a Viessmann boiler as previously stated)

I'll try and report back outcomes just for others to gain info.

:-)

[kimber.kimber]

Ok, so this video shows the cycling I'm talking about. Shortly after this, the boiler turned off and then about 20 seconds later came back on, and started the same thing again. It flames for 2-5 seconds maximum.

It seems to be more of an issue when it's trying to maintain heat, as when it first fires up, it turns off fine.

[StephenC]

My suspicion is that it's only a small amount of water running through the boiler (e.g. Only 1 small radiator using the water, and others are not demanding hot water), and that the lowest modulation output of the boiler is too high to keep the flame on but maintain that temperature.

From the Manual, it looks like a 26KW boiler with 8.5KW as the lowest load (via modulation) it can handle without cycling.

http://idealboilers.com/installer/wp-content/uploads/sites/3/2013/10/Vogue-Combi-Installation-Guide.pdf

I.e. The load (cooler water in radiators on way back) being presented to the boiler is too low for it to handle (heat to designated Opentherm requested temp) without cycling on/off.

I have the same issue to be honest (35KW boiler that can modulate down to ~8.5KW), but needed a high KW boiler as it was to be a combi.

[DBMandrake]

My suspicion is that it's only a small amount of water running through the boiler (e.g. Only 1 small radiator using the water, and others are not demanding hot water), and that the lowest modulation output of the boiler is too high to keep the flame on but maintain that temperature.

From the Manual, it looks like a 26KW boiler with 8.5KW as the lowest load (via modulation) it can handle without cycling.

http://idealboilers.com/installer/wp-content/uploads/sites/3/2013/10/Vogue-Combi-Installation-Guide.pdf

I.e. The load (cooler water in radiators on way back) being presented to the boiler is too low for it to handle (heat to designated Opentherm requested temp) without cycling on/off.

I have the same issue to be honest (35KW boiler that can modulate down to ~8.5KW), but needed a high KW boiler as it was to be a combi.

You're very likely correct.

If the boiler is over sized or doesn't have enough modulation ratio then a small load from one or two radiators at a low flow temperature may be well below the minimum output of the boiler, in this case the boiler has no choice but to cycle the flame off (the pump will keep running) to let the temperature drop to the requested temperature and then turn the flame back on again.

It's not ideal (a greater modulation ratio would be preferable) but it's not necessarily a big problem. Remember that if you had a BDR91 instead of OpenTherm the boiler would be cycling on and off completely every few minutes under low loads anyway, and they are designed to be able to handle that as TPI controls are common now.

Personally, as long as it is roughly maintaining the desired flow temperature when it cycles on and off and it doesn't go into any kind of lockout (for example a flow temperature gradient limitation lockout) I wouldn't overthink the situation. It's probably working fine. Give it a chance to see how the system works overall.

[kimber.kimber]

My suspicion is that it's only a small amount of water running through the boiler (e.g. Only 1 small radiator using the water, and others are not demanding hot water), and that the lowest modulation output of the boiler is too high to keep the flame on but maintain that temperature.

From the Manual, it looks like a 26KW boiler with 8.5KW as the lowest load (via modulation) it can handle without cycling.

http://idealboilers.com/installer/wp-content/uploads/sites/3/2013/10/Vogue-Combi-Installation-Guide.pdf

I.e. The load (cooler water in radiators on way back) being presented to the boiler is too low for it to handle (heat to designated Opentherm requested temp) without cycling on/off.

I have the same issue to be honest (35KW boiler that can modulate down to ~8.5KW), but needed a high KW boiler as it was to be a combi.

Cheers for the quick reply. The boiler I have is an S32, so 32/10.5kw. It's actually the lounge rad that's on, and it's a monster (about 11k BTUs). Interestingly the total comfort ap was telling me that the lounge was at temperature, yet the valve was open 65%. There will only be 14 rads in total, but 32k was the perfect size to satisfy the heating and hot water demand. 32 and the open therm was the reason I went for the Ideal. Unfortunately big Edwardian rooms = big rads...

I think I really need to get all of the rads on, system balanced and see how it's behaving then there's probably about 13k btus missing at the moment, so that may we'll be having an effect.

[kimber.kimber]

You're very likely correct.

If the boiler is over sized or doesn't have enough modulation ratio then a small load from one or two radiators at a low flow temperature may be well below the minimum output of the boiler, in this case the boiler has no choice but to cycle the flame off (the pump will keep running) to let the temperature drop to the requested temperature and then turn the flame back on again.

It's not ideal (a greater modulation ratio would be preferable) but it's not necessarily a big problem. Remember that if you had a BDR91 instead of OpenTherm the boiler would be cycling on and off completely every few minutes under low loads anyway, and they are designed to be able to handle that as TPI controls are common now.

Personally, as long as it is roughly maintaining the desired flow temperature when it cycles on and off and it doesn't go into any kind of lockout (for example a flow temperature gradient limitation lockout) I wouldn't overthink the situation. It's probably working fine. Give it a chance to see how the system works overall.

I think you're right, tabout needing to let it settle in. Problem with being an Engineer, that I end up overthinking everything and analysing it! The wife will probably be quite happy when I stop crouching in the airing cupboard looking at the boiler!

I did read into modulation ratios when I spec'd the boiler, and the Ideal was advertising up to 7:1. However, now looking at the numbers in the installation manual, and the spec sheet on the web (they are both different!) I can't get it more than 5:1. I'm assuming that when it's quoting at 70 and 40 degs, is that the flow temperature?

Installation manual
IMG_6919.jpg

Product Flyer
IMG_6917.jpg

IMG_6918.jpg

[kimber.kimber]

Oh and thanks for the replies all. Great to know there's a good support community out there, who are willing to offer advice.

[StephenC]

Ok, so have monitored very crudely via the onboard HR92 config/display the Valve Positions for a few radiators.

What I am seeing is:

For a couple of radiators, 0.5 degrees C setpoint increase higher than reported room temp is giving a valve position of 98% and 100% (2 radiators). Both old radiators and circa 10 year old valves. 1 has a danfoss adaptor for the HR92, the other HR92 is straight on.

These radiators overshoot *consistently* as they always seem to get 81/82 degree ch flow temps through them via Opentherm which makes sense now based on valve position.

I have set the stroke to 1 (was 0) and also re-calibrated he HR92 on one of these radiators to see if there is any difference.

For another radiator, 0.5 degrees C setpoint increase higher than reported room temp is giving 64% and 1.0 degrees C giving 79%. This radiator is about 2 years old and the HR92 is directly fitted.

This radiators overshoots *reguarly* as as it always seems to get high temp degree ch flow temps through it via Opentherm.


After all, even 64% of 90 degrees C is 57.6 degrees, and 79% of 90 degreesC is 71.1 so still fairly high ch flow temps for such a small increase in temp request (note my boiler is not providing the Honeywell R8810 Opentherm bridge a max ch flow temp, therefore the issue I have is that the R8810 Opentherm Bridge makes an assumption of 90 degrees C from what we can tell)

So a newer/different valve and more appropriate valve open percentages does not seem to massively affect overshoots for me though.

Now this is just a small sample and not very practical to keep monitoring other than spot checks.

Of course if most radiators are within 0.5 degrees and valve slightly (or fully as some of mine seem to be) open, and then just one zone/HR92 calls for max heat, it's going to throw the other zones off by sending through significantly hotter water through than was requested, and expected. This will result in overshoots as I am seeing.

[HenGus]

I have HW priority and when the system reverts to CH, I am only seeing overshoots of 0.5C which is less than the case with the old non-condensing system boiler that my new boiler replaced. When I asked Honeywell for more information on OT, I was told that Evohome treats CH as one zone to inform the boiler. If higher than anticipated boiler flow temperatures are detected then Evohome will adapt. My advice, after 3 years of Evohome ownership, is 'let Evohome do its own thing'. It learns and it settles down. I have never changed any of the HR92 parameters and I get approximately a 2 year life out the batteries. In the early days, if I thought that a given zone was a little too hot or cold for comfort, I just adjusted the set temperature for that zone. There are pages and pages on the accuracy of HR92s. Accuracy depends in part on the location of the HR92 within the zone and local obstacles. If the boiler is pushing out a flow at 80C, it doesn't follow that one or more radiators will overshoot: in my experience, you will just hear more individual HR92 motor activity. Opentherm seems to reduce HR92 activity which may in part be due to the boiler change and reducing boiler temperature flows.