Room Thermostat

Sounds like an old fashioned sensor. Does it have a digital readout, or an old pointer and dial? If it's a pointer and dial, the pointer is probably mis-adjusted. Usually a small screwdriver will fix it. Pop the cover off, and adjust it.

There is usually a setting, too, for the temperature variance. The variance may be too large if the room gets really cold before the sensor is triggered.

LMGTFY

I suggest that his friend step into the digital age.

Google it? And forego the thrill of wild speculation? Where's the fun in that?

Google has never yet stifled wild speculation on this forum.

There's more than one way to build a watch.

...or a windmill.

They must be really bored, I found a brand new replacement thermostat online for $11.99 usd.

Great idea!!

Those bimetal sensors wear out and become sluggish in their responsiveness....no fix that I know of, just change it out...cheap and easy

Room thermostats start at £8.02GBP (usual disclaimer). It isn't worth worrying about - just replace it with a new one.

Thanks for responses - reply to everybody.

It was new about 2 years ago, never been right, but unlikely to be due to wear.

It's not just about the (his!) money, more curiosity. It replaced an older Honeywell, one reason for going same was the connections are in the same place. There's not much spare cable in the box, so if they're different on another make it'd be awkward to wire up.

Good point about adjusting the variance. I don't know if that's possible, it doesn't mention it in the spec, but then it wouldn't, would it. I'll give it a go.

I don't think the differential (or variance) seems particularly bad when turning the knob back and forth, I should have checked that, I'll do it mext time I visit.

The T6360 has a bimetallic diaphragm element. I had a Honeywell T403, installed in 1973 and still OK. This has a bimetallic coil, no accelerator, the contacts magnetic to give a snap action and prevent dithering. Differential fine. Currently I have an Eberle, put it in in 1998, cheapest going as I was short at the time. It has an accelerator and couldn't be any better, I'm well pleased with it. I've rested a mercury thermometer on it for a while as a test, and any variation is undetectable.

Thanks again for comments.

If this thermostat is located in a bad stagnant spot, it never, ever will work right, no matter what is done. The only remedy might be something like installing a fan to blow a selected parcel of air over the thermostat. See if that helps.

As to changing any electrical component, such as the accelerator resistor, I would be a bit leery of doing anything that might conceivably affect device safety or void any warranty, or affect homeowners insurance policy validity.

On the other hand, if you want to mess with it in a supervised setting only, place a smaller resistor value parallel with the original and see what it does.

Take it off after testing, then replace the entire thermostat.

It's not a problem of location etc, previous stat was OK. As I said in #10 it's far too much trouble to experiment with various resistors. I was curious what was a likely value for the resistant in these stats, but nobody has suggested a figure.

As #20, it might turn out to be a bellows leak. In any case it looks like a new stat is called for.

If you want accuracy and programmability try a Nest

<...nobody has suggested a figure...> because there isn't one. It's not about the watts dissipated, for the effect of those watts depends upon a range of factors, including the location of the thermostat and its construction; turning it upside down, for instance, will affect its behaviour. Thermostat manufacturers use a figure that just "seems to work" in the majority of installations - it's a simple fudge, and won't respond to predictive calculation.

From the postings so far, the resistor isn't the problem; it's something else.

Room temperature control is not the most demanding of control system applications! So changing the unit remains the top recommendation.

Please dispose of the original in accordance with the WEEE Directive, as it is not worth repairing and will just be a "bleedin' nuisance" to the next individual who picks it up and tries to use it.

There must be a typical figure for heat dissipation hence resistance, because manufacturers use a value. It might not be the same for all manufacturers, there will be a tolerance, and it won't be ideal for every installation, but there is a typical figure nonetheless. That's all I was asking.

I agree that now, with the help of the forum, it's turned out to be irrelevant. I've acknowledged (#22) that it looks like a replacement is needed.

In three-term-control terms, the thermostat is a high-proportional controller with a bit of derivative provided by the heating resistor. Increasing the heat output of the resistor, by reducing its value, is the same as increasing the derivative action. The purpose of derivative is to prevent temperature overshoot. However, too much derivative action will cause the temperature to bounce below the setpoint, and induce short-cycling on the heating equipment, which may not be the best thing for it in the long term.

Overcoming the effect of temperature bounce can be carried out by raising the setpoint on the thermostat. However, that is where the installation is at present.

Although changing the 330kΩ resistor for a 270kΩ ¹/₄W one or a 220kΩ ¹/₂W one is tempting, it might not actually solve the problem if the behaviour of other components within the unit is the actual cause of the problem. Plus, one's time and the heat dissipated within the premises while the unit is misbehaving are of value. So the recommendation of changing the thermostat for a new one is persistently persuasive.

Yes, I wouldn't disagree with any of that.

The T403 has an ordinary resistor inside, but only connected at the live-when-calling end. There is no neutral wire at the stat, but just for fun to see what would happen, many years ago I connected the other end to the earth terminal (obviously as good as neutral for a test). It was much too powerful, and switched off after a few minutes, well before the room was warm. Causing short cycling as you say. I've no idea now what the resistance was.

On the T6360 you can't see the resistor, but it would likely be easy enough to add another in parallel. But far more trouble than it's worth in the absence of firm data on what the power should be. Or even with that data!

<...connected the other end to the earth terminal...> There's only another 29 milliamps to go!

??? earth and neutral (elsewhere on the premises) are at the same potential.

BS7671 says, "Don't do it, Ethel".

Rules are made to be broken. What's the relevance of 29ma?

An installation to the latest edition of BS7671 will have an imbalance trip device ("earth leakage circuit breaker") upstream set to trip at 30mA to protect occupants from the effects of electric shock. Introducing a current deliberately onto the earth conductor of about 1mA reduces the headroom to the trip point to about 29mA, thereby deviating from the standard. That's why such a connection falls outside the standards, and to carry it out transfers the risk away from the premises' insurers to the deviant.

Yes but no such nicety as an ELCB was installed in 1975-ish. There probably still isn't. In the unlikely event of anything happening involving insurance I'd have disconnected it quick before the loss adjuster turned up.

That suggests a distinct lack of ambition...

Maybe, but let's not make a mountain out of a molehill .

When I did the little test I described I was perfectly sure there was no risk to me, my wife (at the time), my young (at the time) family, my house (at the time) or my insurance contract! Otherwise I wouldn't have done it. If there had been an ELCB installed, and it had tripped, so what? That's what it's for.

Anybody who wasn't confident it would be OK wouldn't do it.

...and anyone disconnecting the Earth conductor upstream of the thermostat, in pursuit of new installation work or existing maintenance, and touching it can expect a mild, though not life-threatening at this time, electric shock, which shouldn't be the case. That's why it isn't done, and it's an important habit to get out of. Although "only a milliamp" now, persistence with this practice could lead to serious consequences later.

"Don't do it, Ethel." _{- Anonymous poster #0}

<unsubscribes>

You do things your way, I'll do them mine. Good idea to unsubscribe.

<...Rules are made to be broken...>

"Rules are made for the obedience of fools and the guidance of the wise" _{- Anonymous poster #0}

"One may not make a sufficient number of nor stringency of laws to govern the ungodly" - James Stewart

"Our Constitution was made only for a moral and religious people. It is wholly inadequate to the government of any other." - John Adams

It appears that now that time is upon us, when morality, Christian principles of behavior, and good common sense has taken leave of us all.

Will we not turn again to first principles, and enjoy the warmth of the sunshine that does proceed from it?

As far as I can make out the anticipator is just a little heater used to inject a bit of extra heat into the thermostat element (while the heater is powered) to make it turn off early (to limit temperature overshoot due to heat stored in the heating element after the power has been turned off). I guess they use a resistor a a cheap way of making the little heater.

Correct.

Of course. There's never been any doubt about how it works, not in my mind at least.

A room stat is just a switch that's operated by an expansion coil or bellows.

Bazzer

Can you tell me where you found that please? I had a good look on the web and didn't see it. The Eberle data also says differential 0.5°C and if it worked as well as that it would be great!

www.honeywelluk.com/documents/Full-Specification/pdf/811.pdf

Thanks for that, interesting. It has a filled bellows actuator, not a bimetallic diaphragm as I thought. Maybe the problem is it has leaked, I had that problem once on a modulating gas valve actuator.

Any guess what For thermal actuator control under the Specifications table? It doesn't seem to mean the temperature element, as the same symbol appears at several places in the table. From the table you might wonder whether any model has an anticipator, except T4360C which might or might not! Mine does, at least according to the wiring diagram inside the cover.

Haven't got a clue what For thermal actuator control might mean.

The wiring diagrams on the last page show that many of the units (6 of the 11 varieties) have anticipators.

That's right, the table and the wiring diagrams don't seem to agree. Confusing, but not a problem in practice.

Sounds like the bellows has failed....probably defective.....or as someone suggested you have an air leak from the wall behind the t-stat, rare but a possibility...

...air leak from the wall behind the t-stat, rare but a possibility...

Thermostat exposure to air flow from a hollow wall is far from rare.

If it's a forced air heating system, this is what might well be taking place:

1. System OFF: Fan off, no air flow, thermostat senses falling room temperature.
2. System ON: Fan running, air flow from wall cavity over the thermostat cancels anticepator (accelerator) effect, room overheats.
3. System OFF: Back to step 1.

He can't check that out now since he's not at the site, but it's the first thing I'd want to look for.

Classic!

Agreed, probably the bellows and replacement needed. It's unlikely to be location etc as the previous one worked. How he managed to break it I don't know, I'll ask him!

It's unlikely to be location etc as the previous one worked.

What may be likely is that during thermostat replacement, additional holes in the wall were made or existing ones enlarged.

It's worth checking out.

A couple of thoughts. Mounting location, air currents, anticipator resistor across line instead of load(wired wrong).

https://ensupport.getconnected.honeywell.com/servlet/fileField?entityId=kaN80000000TUV7EAO&field=File_1__Body__s

Calibration

Is the thermostat mounted on a hollow wall?

Make sure that cold air is not entering the thermostat from the wall cavity. That would cause the temperature overshoot you describe.

Clever and very possible!!

Thanks everybody for suggestiions.

From the useful links posted, it seems the accelerator isn't that important to give adequate control in normal circs.

Most likely the sensing element itself has failed, probably due to leakage of the bellows fluid. I'll check it tomorrow, but it looks like a new stat, as I suspected and several posters have said!

What type of fluid is in a bellows ?

Can barely read your image. I'm no expert, but the principle is to find a fluid that is liquid at the lower end of the temperature range considered, but whose vapour pressure rises rapidly with temperature. It would be one fluid, maybe diethyl ether for a roomstat, 10 - 30°C, something else for a boiler control stat, 60 - 100°C. But you're not limited to atmospheric pressure, my guess is operating pressure is typically a few bar, but somebody might correct me. Also as the bellows expands it is restrained by the flexing of the bellows walls and because of a bias spring, usually adjustable to enable change to the temperature setting.

So if you were designing one from scratch there'd be a number of variables to consider, but I'm sure the makers have got it sussed by now!

I suppose you realize also that air expands upon warming, yes?

I see no reason why a bellows over an air chamber (sealed) could not be employed for a thermostat.

Yes, I have heard a rumour to that effect .

I see no reason why a bellows over an air chamber (sealed) could not be employed for a thermostat. I suppose it could in principle, but I believe in practice a fluid whose critical P and T are somewhere near the thermostat operating range is employed. This gives a bigger change in volume for a given temperature rise. As I said, I don't claim to be an expert, anybody who wants to comment is welcome.

Critical temp and press, or just the normal boiling temp? IMHO gases near critical are usually (there may be some exceptions) at some pretty outstandingly high pressures.

https://www.engineeringtoolbox.com/gas-critical-temperature-pressure-d_161.html

I do not see even one with Pc near 1 bar.

I had meant to think this through a bit more, and you've prompted me.

You're right, critical temp and pressure likely to be much too high. Checking, a range of organic compounds have critical pressure around the 50bar mark. So boiling point more appropriate, but not necessarily normal BP, at atmospheric pressure.

As an example, designing one for around 150°C using water (only because I have vapour pressure vs temp data to hand)

At 150°C, VP = 3.75barg, at 155°C, VP = 4.5barg (roughly)

So as a starting point for design, I would fill the bellows 1/2 full with (cold) water, then pump in air to pressure 3.75barg. When installed, on temp rise to 150°C not much happens, but going further the bellows expands (moving a pointer or switch contacts), but restrained by a spring, spring rate selected to give the right force change over the bellows movement. The other end of the spring could be adjustable to vary the temperature response.

Does that make any sense? Of course more often the temperature is lower and the volatile fluid isn't water. I think earlier engine thermostats, to open at ~ 85°C used ether or something similar. I put the air in only so the set-up pressure could be provided, but maybe that's not necessary, and it could be filled completely with the fluid.

Also in some cases eg a boiler control-stat, the bellows is not exposed to the hot environment, but connected to a filled phial via a capillary.

A lot of people have thermostats fitted in the wrong place either too high, too low or in a place where there is a cold spot ie a entrance hall where the front door makes for a draft which would be calling for heat when the rest of the house is baking hot, one thing don't rely on the temp numbers on the stat there not always accurate.

Bazzer

The modern thermostats are very accurate with a hysteresis that is close to perfect for me personally.

We also have them radio controlled, that is the temperature sensing part can be placed close to where you are sitting and the rad has a receiver, that winds a small motor in and out to control the water through the rad.

Why fix such an old fashioned bit of kit as new ones today are available here at least for under $30!! With no guarantee that the fix will work as required.....

Mine are now 12 years old and still work well, though some really new ones I have seen are apparently even better! But as far as I can tell, not better enough to make a change worth having!!