Hysteresis on the Honeywell old-style bulb thermostat

at

to

es

to

g

The voice of wisdom speaks. Thank you Dave. If you take the thermostat off the wall and tilt it left and right you can see the heavy mercury drop flop back and forth causing the hysteresis in the temperature by-metal spring. An elegant design from old school physics that does not require microproces sors and such. I would add more reliable as well.

There is a lot of information both accurate and not-so-much in the replies, so, in summary:

a) correct that the mercury bulb is designed as slow on/less slow off. Th is allows for the over/undershoot. b) the anticipator matches the nichrome resistor load to the power availab le from the gas valve. Keep in mind that back in the day, many hydronic sys tems ran on gravity, no circulator, no mains power. So, power was supplied by a 'millivolt pile' heated by the standing pilot. The anticipator in this case would be at the lowest setting. As gravity systems were slow start-hi gh overshoot systems, the stat would want to respond directly to ambient te mps as overshoot was built in. c) systems with circulators were more capable of holding a steady temp, ra n the gas valve at higher voltages, and used limit switches to manage syste m response. The trick noted of using the MA setting on a VOM is the way to go here.

Now, if one lives as we do in a 4,200 s.f. 3 story center hall colonial bui lt in 1890 with an hydronic heating system installed in 1928, such a stat w ould not be suitable unless several were used in a zoned set-up. Our syste m uses a gas-fired, modulating, condensing boiler with sensor for supply an d return water temps, external air temp, and response time (a function of t he change in return water temp over time). We use a smart-stat that also le arns, so that it will reach temps 'as timed', rather than starting to heat at a set time. That, too, has a outside temp function. This can get very ef ficient such that even this big house is quite reasonable to heat.

Peter Wieck Melrose Park, PA

[snip]

The mercury switch provides some hysteresis.

--
Mark Lloyd 
http://notstupid.us/ 

"In fact, when you get right down to it, almost every explanation Man 
came up with for *anything* until about 1926 was stupid." [Dave Barry]

Yep, I understand that the "Spring controls the position of bulb depending on temperature". That was, I thought, my intention to describe. When the spring bends far enough to tip the bulb, the weight of the mercury in the bulb swings the bulb a bit farther , requiring the temperature to cause a greater swing in the other direction to make it switch back. That's where the hysteresis comes from. That was the question from the OP, which is what I was trying to answer.

cheers, Dave M

ssinzig wrote: "arrow, makes whatever you are adjusting 'LONGER'. No ambiguity there at all, except for what that 'LONGER' adjustment is specifically doing (I "

Should be pretty obvious what is happening for a "longer" period of time: the boiler or furace is running longer! Moving that slider toward Shorter results in shorter, more frequent run times of the boiler or furnace.

Assuming setpoint = 68.0F Longer means swing from 66 to 70, but house gets hot and cold. Shorter means swing from 67.8 to

68.2, but heat cycles on and off constantly.

Just right(anticipator setting measured with meter) and you should stay between

67.5 - 68.5F.

Same concept with digital "Firing length"

3-4 position menu setting.

That's the main reason it is so important to level the t-stat.

I forgot to add that, in more direct response to the OP's question, that the hysteresis (temperature difference required to switch the HVAC unit from off to on, and back to off) is created by the temperature characteristics of the bimetal spring, the weight of the mercury ball (both unchangeable by the user), and modified by the anticipator resistor.

More Cheers, Dave M

Okay, that seems reasonable. Although I would argue that it is not 'pretty obvious' what is happening for a 'longer' period of time, without already being familiar with its operation or consulting an operating manual.

Is the scale (1.2 to .10) measured in seconds, minutes, or hours? Is it a multipler, ie. 0.8 x burn duration? Is it the time the burner stays on? Is it a time delay before the burner comes on? Is it a time delay before the burner turns off?

Everything is always 'pretty obvious' when you are already familiar with how it works.

S.

built in 1890 with an hydronic heating system i"

That changes things. In a hot water system there is little to no overhead i n starting and stopping the system. Also the water has thermal mass so it e vens it all out.

For most places, hydronic heating is much superior to forced air. The only problem is it does not filter the air, and AC cannot be easily added. If yo u ever want AC don't let anyone talk you out of that system, make them put in a separate air box for it. Then you will have not only balanced heat, bu t balanced AC, which is extremely difficult to achieve with a combined syst em in a multi story house. With separate systems you will be much more comf ortable.

But the bottom line here is that in a hot water system, the anticipator mak es little to no difference in performance or efficiency. You don't have to worry about it.

Here here , I second that. IKEA furniture being the only exception.

We

There are always a few (not so few around here) outlyers. Gravity systems h ave a *huge* overshoot such that the stat must be set properly! Keep in min d that most (very nearly all) of these systems were built without mains pow er, and, of course no circulators. So it is critical that the stat responds to ambient temps fairly closely. It takes *time* before the differential o n the supply side is enough to restart circulation once equalized. So, the under/over is established in part already. In such cases, the stat should r espond very nearly to ambient temperatures (and should be on an inside wall not facing a radiator).

When we moved into this house, I - with more than a little help - removed a 400,000 BTU steel jacket oil burner and replaced it with a Weil Mclain 230 ,000 BTU gas fired Ultra that also makes our hot water. The system was orig inally gravity, with 4" ID risers in a 2-pipe configuration. During the ins tall, I added two circulators on the heat side, one for the hot water - the computer on board the boiler manages the dance for all three. But what it means is that all 38 radiators are at the same temperature unless otherwise managed. This allowed me to install thermostatic valves in strategic locat ions (17 in total) such that we have lots of flexibility. In 8 winters, we have never had a lick of trouble.

But a smart stat is essential to keeping all this in balance.

Peter Wieck Melrose Park, PA

Yabbut.....IKEA uses *pictures*!!

That is an exception!

Peter Wieck Melrose Park, PA

" In such cases, the stat should respond very nearly to ambient temperatures (and should be on an inside wall not facing a radiator)."

Peter Wieck Melrose Park, PA

++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

And don't forget to consider what is inside the wall. Such as hot and cold water pipes, return air ducts, wiring,...

Except if it is a thermopile system. If so, there is no anticipator.

It's the current draw for the solenoid valve.

That depends on the furnace. MY blower comes on as soon as the burner ignites, it's a high efficiency furnace, so the blower must start to cool the secondary heat exchanger.

ssinzig wrote: "- show quoted text - Okay, that seems reasonable. Although I would argue that it is not 'pretty obvious' what is happening for a 'longer' period of time, without already being familiar with its operation or consulting an operating manual.

Is the scale (1.2 to .10) measured in seconds, minutes, or hours?" No time bearing whatsoever. It's VOLTAGE. "Is it a multipler, ie. 0.8 x burn duration? " No. "Is it the time the burner stays on? " Yes!! "Is it a time delay before the burner comes on? " Sort of. You're starting to get it. "Is it a time delay before the burner turns off? " Getting warmer!(pardon the pun). It's function is determined by voltage(the numbers on the anticipator slider are in Volts).

A heat anticipator generates 'false heat' - it tricks a traditional bulb-stat into firing the boiler in a more energy efficient manner, by generating a small amount of local heat within the wall unit itself.

The correct amount of false heat both prevents the thermostat from calling for heat too soon after it drops below set point(what you the user set it for), and prevents thermostat from running so long that it overshoots that set point by significant amount.

At its extremes, a heat anticipator could cause too frequent and short boiler runs, or not so frequent boiler runs between which the house gets too cold, and then gets too hot before the boiler turns off. Set properly, the anticipator will turn the unit off just before reaching the setpoint, so that remaining hot water(or air) in the system can be pushed through the system without a significant overshoot.

Succinctly, you want neither a scenario where the boiler turns on and off every three minutes, nor where it turns on for one half hour, and is then off for nearly one half hour. The anticipator, when properly calibrated, prevents both situations, and keeps the temp. within a tolerable range.

"Everything is always 'pretty obvious' when you are already familiar with how it works. "

S. "

That's why we're here to share knowledge. ;)

How high efficiency? Ours are 96% 2 stage one. There is built-in delay for blower to come one. Off delay is adjustable by dip switches. Inducer blower purges vent already before ingnition comes on.

Sounds funny, then your system will blow cool air before warm air start blowing out.