Search the forums

Loading

How accurate is your thermostat?

Nuke's picture

My wife is a teacher. She's got a PhD in Inorganic Chemistry and teaches at a state college here in Georgia. As such, she has access to lab equipment. I asked her to bring home a cheap thermometer. She brought home an alcohol-based thermometer and I draped it over the thermostat.


The alcohol-based thermometers are relatively inexpensive compared to mercury-based thermometers, and accurate to a half-degree Celcius (0.5º C). My thermometer in my master bedroom reads 72ºF, which is about 22.2ºC. The alcohol-based thermometer reads 23.5ºC, or about 74ºF.


Ok, so a two-degree temperature difference may not seem like much, but to the eyes of a wife that writes the checks for the bills every month a two-degree difference in where one sets the thermostat can change her mood.


Now, I plan on measuring the temps in the other three bedrooms on this second floor. This includes a bedroom converted to a small theater room (complete with a 145-pound CRT projector) and another converted bedroom that is my computer room (complete with a PC with four hard drives in it as well as a 19" CRT monitor).


I am doing all of this to put into numbers for my wife that sometimes understands these other rooms are warmer in the Spring-Summer months, but she doesn't adjust her thinking of tolerability within scientific evidence. lol

(post #112082, reply #1 of 24)

I have read the instruction and many of the high end programable thermostats will allow you to put in a offset. It is burried in the setup program and not visable from the day to day temp programing.

. William the Geezer, the sequel to Billy the Kid - Shoe

(post #112082, reply #2 of 24)

It's only a number that is related to your (her) comfort. As long as the difference is the same at different temps, no big deal. Set it to where she is comfortable.Many manual stats can have the thermometer adjusted.

(post #112082, reply #3 of 24)

Nuke, I don't know what thermostat type you've got, a couple of years ago ours died -- an old White with a mercury bubble -- and I installed a Hunter programmable one from the blue borg.

Ran me about fifty bucks, as I recall -- so far I'm making money on it, as we seemed to forget all to often to turn the old one up or down when we left for work.

The programming on it can be set for almost any parameter you wish -- ours is set for 78°F in summer, 66°F in winter. That's where SWMBO is comfortable, or at least can live with it.

Regards,

Leon Jester
Leon

(post #112082, reply #21 of 24)

Leon,

I don't know when you got your Hunter thermostat, but I had several back in the late 80's/early 90's and found that they failed rather quickly. My HVAC person confirmed that that was his experience too. For quality (price not considered) he considered the Honeywells to be best. For price/quality tradeoff, the Whites.

(post #112082, reply #23 of 24)

Thanks, Bryan, I'll keep an eye on it.

IIRC, it's got a multi-year warranty.

Got it at the Blue Borg. At the time, it was the most flexible of what was on the rack. The next-up in price offered daily control settings, this one has M-F and Sat/Sun which works fine for us.

It replaced a White, which was flat-slap wore out. I've some suspicions that it might not have been installed correctly myself, as quite a bit that the previous owner did was half-a**ed at best.

We won't go into the attic fan and why I have 250 ft of 12/2 Romex lying around unused. Bad for the blood pressure.

Leon Jester
Leon

(post #112082, reply #4 of 24)

Where it is located can have a significant effect.

Back in the late 70's early "energy crisis" days (back when we knee jerk touchy feely types were learning how to tree hug <G>) I was living in an old VT farmhouse.

I was being really good, keeping my thermostat set in the mid 60's or something.

Man, I was GOOD! Broke my arm in three places patting myself on the back!

One day I happened to leave a cheap thermometer (intended for another purpose) sitting on the kitchen table until I got around to using it. Glanced at it after a couple of hours and it was showing low 70's!

??????

Checked the thermostat: Nope- set for mid 60's?!?!?!?

Quid malborg en plano - consternation turned to elucidation - the thermostat was on the basement stairs wall, the "basement" was fieldstone lined cellar barely running above freezing!

Since then, I've heard people complain their apartments/houses were always cold - usually turns out they have a lamp near or a TV under the thermostat!

======================================== "Man's capacity for justice makes democracy possible, but man's inclination to injustice makes democracy necessary." Reinhold Niebuhr: 'The Children of Light and the Children of Darkness' http://rjw-progressive.blogspot.com/ ========================================

(post #112082, reply #18 of 24)

were always cold - usually turns out they have a lamp near or a TV under the thermostat!


LoL!  I know of a series of "fourplex" apartment units scattered all over the county, that, due to their design, have a hall light lamp fixture right over their round stats.  I've helpped innumerable maintainence types & managers with "heat broke" compliants by telling them to unscrew the 75W (and larger) bulb out of the hall light fixture.


This cures some $200-250 electric bills (for 800sf apartments) for a/c in the summer, too.


 


Occupational hazard of my occupation not being around (sorry Bubba)
I may not be able to help you Occupational hazard of my occupation not being around (sorry Bubba)

(post #112082, reply #20 of 24)

Among some of the Old House people, there is much debate about whether set-backs save any money or energy. The theory being that it takes more energy to raise (lower) the temperature than it does to maintain it. Have you done any research on this, like how much of a set back for how long at a given energy price will result in savings?

It seems to me that there should be graphs of set back time vs set back degree difference that shows how much energy it takes to get back to the desired temperature, with a reference line of maintaining the temperature, ie " a set back of 4 degrees C from 32 to 28 C for 6 hours requires Z BTUs, holding it at 32 requires W BTUs". I know it would vary based on heat loss (gain) of the building, but those computer systems used in office buildings must work on about the same principal, don't they?

(post #112082, reply #22 of 24)

The amount of energy that is lost is based on Delta-t and time.

Reduce the delta-t by lowering the thermostat and you will use less energry for that time period.

But when you raise the temp you not only have the higher energy usage because of the higher delta-t, but also the energy needed to raise the temp of the building and contents.

However, that is not "lost energy".

The trick is "use" that stored energy to heat the building before you set the thermostate back.

For example if you leave at 7 am, you would set the thermostate to go back at, say, 6:30.

I have a Honeywell with Adaptive Setback. It figures how long the furnace is running and you enter the actual times you want it to change setpoints and it adjust the setpoint times to compensate.

. William the Geezer, the sequel to Billy the Kid - Shoe

(post #112082, reply #24 of 24)

there is much debate about whether set-backs save any money or energy


Which is one of those arguments that's easy to "win" by setting all the presumptions in favor of "your" argument.


The real world "trick" of a set back, is to have a middling consistant schedule in the space conditioned.  If no one is home, the temp can be allowed to get colder during heating season, and warmer during cooling season by 3-4º to 5-6º (or 7-8º if you believe the stat makers).


Now, the next trick of it, is to have a consistant time for the system to come back on.  I have found that most folks can "stand" to have the system come on an run while they are just getting home.  In an ideal world, maybe that's half an hour before you (spouse, kids, whatever) get home.  Also, you can get a few degrees warmer or cooler after everyone gets to bed. 


Reducing the temperature setting reduces the run time of the system, which generally reduces costs.  There's a "gotcha" in that you genreally need to run the system a few minutes in every hour for air quality, but that's why the stat temp is set "back," not set "off."


Now, if it's a building with people coming in and going out at all hours, then, sure, the setback won't help a lick (the users will be "defeating" the set back for comfort more than the schedule will run).


Now, the USN (all fortran on punchcards no less) energy modeling program I was using in the early 80's suggested (and only suggested) that there was some economy in creating a "bubble" of conditioned air in advance of need.  That presumed a lot, too, like knowing the need schedule, and how well that bubble would "hold," and whether or not there was as real cost benefit from "off peak" conditioning.  It still really only works for threaters and very large industrial facilites in that sense.


Occupational hazard of my occupation not being around (sorry Bubba)
I may not be able to help you Occupational hazard of my occupation not being around (sorry Bubba)

(post #112082, reply #5 of 24)

Two degrees is probably pretty good. The old round Honeywell mercury thermostats could be off by a lot just because they weren't level. And many mechanicals had an adjustment screw that could be fiddled with to correct (or throw off, depending on your preference) the scale.

Another thing to keep in mind is that having the anticipation setting wrong can contribute an error. (And what % of those do you think are actually set right?)

I frankly wouldn't expect any home thermostat to be accurate within more than 2 degrees F.


Of all the preposterous assumptions of humanity over humanity, nothing exceeds most of the criticisms made on the habits of the poor by the well-housed, well-warmed, and well-fed.  --Herman Melville

(post #112082, reply #6 of 24)

Sorry, guys. I should have provided more detailed information. The thermostat is a White-Rodgers unit mounted shoulder-height just inside the master bedroom door and directly under that room's air return.


I took the faceplate off and on the faceplate is:


Type No. 1F80-51


On the PCB itself is '593866' ink-stamped in the upper-right corner, and '0019B E5' ink-stamped in the lower-left corner. Its an ICD readout and non-mercuty (according to the back of the faceplate. Runs on three (3) AAA batteries.


I have an identical unit downstairs for that zone.

(post #112082, reply #7 of 24)

So the difference between your two thermometers is 1.3 degrees C. The tolerance on the "lab" unit is +/- 0.5. Now that's "lab" quality so let's assume the thermostat is only halfas accurate and +/- 1.0 C. Combined that is +/- 1.5 C. The difference is actually within the tolerance band of the two instruments.

Stu

(post #112082, reply #8 of 24)

I'm not complaining, but rather noting the offset, which is what I presume it is. I'm also noting the wifey's perception that its a number above what she'd rather not have. Now, if I could get it comfortable for me and have it read 80ºF ... :)

(post #112082, reply #9 of 24)

In my parents' old house, which had electric baseboard heat, the room thermostats were designed such that you could pull the knob off and move the little metal spring to different grooves inside the perimeter of the knob. As installed the thermostats weren't accurate within 5 degrees, but by comparing to a thermometer and adjusting the spring you could get them within two or so. (Or you could fudge the other way, if you and your mother had differing opinions as to what a proper temp setting was.)


Of all the preposterous assumptions of humanity over humanity, nothing exceeds most of the criticisms made on the habits of the poor by the well-housed, well-warmed, and well-fed.  --Herman Melville

(post #112082, reply #10 of 24)

Many, most from the dozen or two specifications I have recently read, thermostats and AC controllers have a differential, the amount they allow the temperature to slack off, of 3 to 5 degrees F. Without some differential the AC would kick on and off too often and wear the starter circuit out. The short cycling would also be annoying as the thermostat called and then stopped calling a degree below the set point only to kick back on as the temperature rose a degree.

Also consider the placement of the thermostat. Interior versus exterior walls tend to show significant differences. Two exterior walls can be several degrees different because of sun or wind exposure. Interior walls vary because interior areas and heat, or cool sources or sinks vary.

An incandescent light bulb, better at heating than lighting, can change the temperature in the surrounding area. A computer, especially ones with CRT monitors, put out a lot of heat. A heavy stone fireplace, unless the fire is lit, can absorb a lot of heat as can tiled floors.

I'm surprised to read that the alcohol thermometer is certified to be accurate within 0,5 degrees. Must be a laboratory model. I noticed the other day at a local retailer that the twenty or so thermometers on sale, all within a foot of each other varied by about seven degrees Fahrenheit. There were not noticeable drafts, heat sources or heat sinks. By rights they should all agree. All were made in China.

On the other hand a thermostat that was off five degrees or so wouldn't surprise me. Most are there to give you a rough idea only. A lot of such thermometers have some adjustment but be careful. Like the person with two watches that don't agree you can't be sure which is more correct.

Perhaps you wife has access to certified thermometers but even there she would be better off with two of them. One to confirm the first which could be compared to the thermostat.

(post #112082, reply #11 of 24)

"I'm surprised to read that the alcohol thermometer is certified to be accurate within 0,5 degrees. Must be a laboratory model. I noticed the other day at a local retailer that the twenty or so thermometers on sale, all within a foot of each other varied by about seven degrees Fahrenheit. There were not noticeable drafts, heat sources or heat sinks. By rights they should all agree. All were made in China."

I have designed some tempature sensing equipment. Much the same range accuracy of a thermostat.

http://www.cselectronicsinc.com/computemp5.php

It is hard to get accurate measurements when testing without having them all in a thermomass (water, oil) and in an insulated container.

Now with something identical, such as the display of glass thermometsers it is not as bad. But take a tiny thermister mounted in a thermostat on the wall, a glass thermometer, and pocket dail thermometer and you will go crazying trying to get them to read the same. Between the time response and air currents they will be all over the place. For my purpses when calibrating that means 1-3 degrees. And when testing endpoints, say from 30 to 120 it was real noticable.

"Many, most from the dozen or two specifications I have recently read, thermostats and AC controllers have a differential, the amount they allow the temperature to slack off, of 3 to 5 degrees F. Without some differential the AC would kick on and off too often and wear the starter circuit out."

The mechanical ones did have a differential and also an anticipator, ie a small resitor that heated it up. That was to allow for the heat from FWA units after the burner turned off.

But I it appears that electronic ones use a time delay instead. You can program them for cycles per hour.

. William the Geezer, the sequel to Billy the Kid - Shoe

(post #112082, reply #12 of 24)

Bill, do you know of anyone that makes a wireless thermostat? I've often wondered why they need batteries when there is a LV line already running to it.

(post #112082, reply #13 of 24)

"Bill, do you know of anyone that makes a wireless thermostat?"

http://www.smarthome.com/91371w.html

This is the only thing that I came up with quickly. I think that Smarthome used to have more.

But you migth want to go through these in some detials

http://www.smarthome.com/hvactherm.html
http://www.smarthome.com/sectemp.html

A google brings these up as the first couple of sites.

http://www.smarthome.com/30401.html
http://www.store.yahoo.com/air-n-water/withprhothsy.html
http://www.wirelessthermostats.com/
http://www.energyautomationinc.com/

http://www.houseneeds.com/shop/HeatingProducts/HydronicHeating/thermostatindexpage.asp
http://gadgets.engadget.com/entry/1234000963023623/

"I've often wondered why they need batteries when there is a LV line already running to it."

There isn't a LV supply, in general, The most basic is for a furnace where you have the transformer connected with to the gas valve and a one lead from the gas valve to the thermostat and the other lead from the transformer to the thermostat.

Yes, you have 24 volts at the thermostat, but when the thermostat closes you don't have any. And when it is open you can't draw too much power or it will activate the gas valve.

A basic AC/heat is similar, but at an relay for the AC and one for FAN ON. But still no common 24 v line.

Then you get in to units with multi-stage furnace, multi-stage AC, and heat pumps with dual fuel or multi-stage aux heat.

And some of them have separate transformers for AC then for heat.

And sometime they can "steal" power from the circuit and use it to recharge batteries, but it can be trickly to have it work reliably with all different kinds of equipment.

Some do allow you to use an extra wire, if available, and connect to the transformer hot so that it always has power.

. William the Geezer, the sequel to Billy the Kid - Shoe

(post #112082, reply #14 of 24)

The standard deviation is 1/2 degree either side of setpoint. Anything more and most occupants can feel a temp difference.  The anticipator adds a little amount of heat on a mechanical stat to shut off the burners early, allowing the residual heat in the appliance to be used. With an auto change-over stat, the deadband between heating and cooling is usually a minimum of 1 1/2 degrees between the heating and cooling setpoint.


If a stat has high swings between off and on, or is shortcycling, there can be different problems.


The anticipator or cycles per hour can be set wrong.


The stat is being affected by external conditions.


The equipment is under or oversized. This can be a big problem on a furnace that is grossly oversized to begin with, and when only a few btus are needed, it still delivers the max. (insert plug for hydronics here LOL)


The stat could be screwed.

(post #112082, reply #15 of 24)

According to my first HVAC instructor all the numbers and readouts on T-stats cause to many trouble calls and domestic disputes.


His perfect stat. would have a cool to a warm setting on the face and incremental marks between the two end point. No numbers or temperature readings would be shown.


Just set the thing to your personal comfort level and let it do its job of turnig the equipment on/off to maintain that comfort level. It's those numbers that cause all the problems.  "Honey, I'm cold, the thermostat says it's 68 degrees in here, and you know I freeze below 72."


Set it where you are comfortable and ignore those damn numbers.


 


Dave

(post #112082, reply #16 of 24)

You don't even want marks on it.

Just a shaded dail that goes from solid blue on one side to solid red on the other.

Otherwise 2 mark from the center is now "72". Gradual shading is hardere to pin down.

. William the Geezer, the sequel to Billy the Kid - Shoe

(post #112082, reply #19 of 24)

Just a shaded dail that goes from solid blue on one side to solid red on the other


LoL!


Too many seem to think that the hvac unit produces air at the temperature selected, anyway . . .


Just the situation for a "pyschostat" which seems to have almost as much effect as a placebo.


Now, the tough part is explaining "delta T" to people; that with an outside temp around 98º it can be difficult to get the interior temp much below 78º without running the hvac system continuously ("But I turned it down to 66º, why isn't it colder?").


My Hunter setback reads 2º higher than my digital thermometer; I just adjust the program accordingly.


Occupational hazard of my occupation not being around (sorry Bubba)
I may not be able to help you Occupational hazard of my occupation not being around (sorry Bubba)

(post #112082, reply #17 of 24)

I suppose none of this helps when its all applied to a cookie-cutter home where 50% of the cooling capacity is centered on the master suite and 50% of the rest of the floor. Unfortunately, I've got one aux. bedroom as an office with a couple of computers running, and another aux. bedroom with a 145-pound CRT projector, subwoofer, and a bunch of other heat-producing components.


And I'm betting the AC is sized for the minimal requirements per trade & code.


I can live with all this for now knowing I am not staying in this house, but my next house will have more than two zones. I'm tired of the Mom+Dad+2.5children home design. I don't have kids, don't want kids, and the rooms the wife and I do want have special cooling needs that no cookie cutter home is going to begin to provide.


Of course, this also means no easy or thrifty home loans as what mortgage company wants to mortgage a home that they cannot worse-case resell to that Mom+Dad+2.5children scenario? :)