Thermostat replacement

Before we start to worry about that old Honeywell thermostat — which is likely to be far more reliable than anything you could buy today, by the way — we need to be sure what it is we are trying to do.

The first, and probably most important question, is how is the space comfort in the house? Does the temperature hold fairly constant, or does it go up and down more than you would like? If the space temperature is comfortable, changing the thermostat isn't particularly reasonable — if it aint't broke, don't fix it!

Now on cycles. You mention that the system "cycles more than it should" and has "long run times". These are actually opposite conditions. If the thermostat is what is controlling the cycles and run times, there are ways — even with that old one — to get shorter run times. There are also ways to get fewer cycles. Unhappily, you can't do both. Your house needs a certain amount of heat to stay warm. You can do this either with shorter bursts of heat more often, or longer bursts of heat less often. It's a compromise act — the idea is to get the thermostat set so that the temperature of the space holds reasonably constant, whatever combination of "how often" and "how long" that takes.

Now if it is the boiler that is cycling — the boiler turns on and off usually on pressure — while the thermostat is still calling for heat, that is a very different problem and changing the thermostat isn't going to make a difference at all.

So we sort of need to know if the boiler is turning on and off — but your comment on pressure suggests that it isn't. And, incidentally, NOT seeing pressure while the boiler is running is a good thing, provided the radiators are warm. You don't need much pressure to get nice warm heat from steam!

Maybe you need to take the cover off and adjust the anticipator setting. Is this a mercury thermostat? IIRC they have a variable resistor heating element near the mercury glass bulb that heats up the bimetallic coil. If that is set to a slower heating setting the burner should run longer.

That looks like a snap switch type with no mercury and no anticipator (or at least not an adjustable one). Typical adjustable anticipator heater would draw too much current for a "millivolt" thermostat. This one was designed to be powered by a thermopile only. Date code may be 48th week of 1994.

I think that varnished adjustment screw is just for factory calibration of the switching temperature.

Do you have a millivolt boiler or is there a 24V transformer?

You definitely need a more sophisticated thermostat.

Jamie Hall had selected a thermostat that is compatible with a millivolt gas valve system. You should post a photo of the gas valve that operates the burner. If you have a millivolt gas valve operating your steam boiler, you have a power failure proof heating system. You do not need electricity from the grid to operate that boiler.

That said, you must replace that thermostat with one that is compatible with 750 millivolt systems. If you use a 24v. thermostat, there is a possibility that the thermostat contacts will have too much resistance to reliably make the connection on every heating cycle. That could leave you with no heat at times until you cycle the thermostat to reestablish the contacts (off then on)    

If, however, you have a different system, and that is the old thermostat from a previous boiler, then that is the wrong thermostat for your system, and that may be the problem with your system cycling problems.

Some systems have a low water level control that periodically shuts the flame off to do an intermittent test. level check. Here's a link to some info on it. If you have something like this there is are two lights on the top. The green one comes on when the test is performed. But these devices only interrupt the flame for about 2 minutes (see the info)

https://s3.amazonaws.com/s3.supplyhouse.com/product_files/Hydrolevel-45-410-2090-Product-Overview.pdf

pretty sure new peerless's usually come with a cyclegard type lwco that stops the burner every x minutes to check the water level. they can be wired 2 ways, one to start the timer when the burner starts firing, and the other is for the timer to continuously run with no information about when the burner is firing. it sounds like yours is wired to do the latter so it might get to its 10 or 15 or 20 minute count 2 minutes after the burner stared.

or it could be a millivolt thermostat with a millivolt anticipator going crazy at 24v ac as @EdTheHeaterMan said. or both.

@EdTheHeaterMan For education purposes, can you explain how the installed thermostat creates the long delay time fro turning off the flame to allowing it to reignite?

I think it is a TS819A (or TS819 with no dual needle/thermometer?) because those didn't have the mercury switch. It probably has a fixed anticipator somewhere that isn't easily seen, just like in the TS822A which was a mercury switch model.

I still wonder why it ran for 2 minutes, switched off for an hour, when obviously it wasn't enough time for the room radiators to even get steam, although if the pipes in the basement are uninsulated that heat could've made it's way into the thermostat room.

I get that this is not to be used with 24VAC and 750mV or less only and 24VAC will heat the anticipator too fast, causing early burner shutdown. It seems like it would normally behave more like he earlier described with lots of cycles close to each other where it didn't shutdown for very long between the short burner on runs.

Maybe since it is only rated for 100mA it is simply drawing so much current that the contacts have become damaged and is acting flaky.

https://www.ebay.com/itm/186426380376

Here's a TS822A with mercury switch where I think the anticipator is visible and attached to the white wire and looks burnt up

https://www.ebay.com/itm/188095301630

is that anticipator free or does it have an anticipator that works on millivolt or 24vac if the 24vac is under 100 ma? the common fan center they use for isolation of newer boilers has a relay that draws about 375 ma so that would cause the anticipator to end the cycle very early if that is what it has.

Hard to find info. on the snap switch TS819 but the TS822A mercury switch one was also for millivolt only use and it shows a non adjustable anticipator resistance in this diagram. I'm "assuming" the TS819 had one as well

https://cdn.designair.com/Resources/HNYWLLE09863_1-5.pdf

Except why would the anticipator resistor take a long time to cool down unless the contacts of the snap switch were damaged by too high current for too long and they were allowing a trickle current through but not enough to trip the isolation relay or whatever other circuitry the signal goes through to turn the burner on or keep it on?

The previous behavior observed was short on time followed by short off time and this was repeating enough that the house was fairly comfortable even though the radiators in the thermostat room were not getting steam due to distance from the boiler and smaller venting. Now all of a sudden though it was very short on time followed by a very long off time. Why would all of a sudden an even shorter on time than ever heat the bimetal coil for a longer time than ever? It has to be damaged contacts in the snap switch.

It is hard to tell exactly what is going on without a more concise detail of the on times and off times consecutively over the day.

What makes you think from what I have written in this thread that I don't understand that the anticipator is probably getting more current than it was rated for? Am I not writing clearly? It was an interesting puzzle is all and got my interest and I'm interested in old thermostats.

You don't think the behavior changed. OK well it sounds to me like it did from what I read that @geemalar has written:

"My system doesn’t appear to have long run times and cycles more than I believe it should."

"I’m just wondering if the thermostat is cycling the system too often. The system will run, I’ll get little heat from the 2 closest radiators to the thermostat and then shut down. 5-10 minutes go by and the system cycles again. "

And then most recently there was this which certainly sounds different to me:

"I happened to be in my basement a half hour ago. …..the boiler had not run for at least an hour or more. The boiler came on, ran for 2 minutes, then shut off. There’s absolutely no heat from the 2 radiators close to the thermostat to affect its operation."

I believe that is anticipator free