It sure sounds defective

Hello
The accuracy of a thermostat is a relative thing when you are talking HVAC. One may read different depending on where it is placed. The objective is to control the system. Using a thermometer to check a thermometer is not very accurate. If you asked me to correct the problem, I would charge you for a trip and time spent. It may or may not be defective. It doesnt matter if the 2 other thermometers are reading the same, It just don't matter.

Ok, Bob, if you don't recommend using a thermometer to check a thermometer, how WOULD you go about determining whether a thermostat is accurate? I just paid $6o for a digital thermostat that allegedly displays the temperature to a tenth of a degree. If it's off by four degrees, then it matters to me.

Hi

I'm not Bob but you may want to check to
see if your thermostat has an offset adjustment
Mine does and I adjusted it to agree with a
Weather Instrument in the same area. I also
make sure that the air is constantly circulating in
that room and that I've sealed the opening where
the wire enters the thermostat from behind.
They typically agree to within one degree now.

HTH
Chris

Thanks, Chris. There is no offset adjustment and there is nothing about calibration in the user manual or installer manual. The only other thing I haven't done is plug the hole in the wall. I don't remember it being especially big, but I'll have to check it.

Hello,
If you are really concerned about the exact reading, I would purchase a calibration meter that is accurate. Have it certified before each test. I would imagine you could purchase such a rig for under $1000 and recertify it for a couple of hundred dollars. Other than a certified meter, you are not accurate with any measurement. Call the manufacturer and find out what is suggested for calibration.

True story: many years ago my sister in law worked for Cooper Thermometer in Middlefield, CT. For Christmas she gave me a display box set up of 24 digital pocket thermometers. (She probably stole em) When I turned them on there were not any two that read the same even when putting them back in the display box. My point, which two would you like to use to check your digital thermostat?
As mentioned previously, check and plug the hole where the wires come through the wall and if you had your hands on or near the t stat I'd wait to give it time to settle out. Oh yeah, and put the digital thermometers right on top of the stat too.

Guys give this guy a little slack! The hole may be your answer, or more specifically, the other hole. When the thermostat was intalled, at least 2 holes were drilled. The air infiltration from an unconditioned space may be causing the discrepancy. If it were me, I'd expand your test by removing the mounting bracket, plug the hole with a rag and setup a fan nearby on a low setting. Wait 15 min check, if still off by mo than a degree, wait another 15. If it's still faulty, either return it still within limits or call RS service directly for a replacement.

Thanks, razl, but I have no time to investigate the hole -- had to take a second job to save up for that $1,000 calibration meter that bob brown so kindly suggested.
Gotta run--

I believe that a difference of 2.5 to 4 degrees between the thermostat and digital thermometers is too much to be acceptable and ignored. I can see +/- 1 degree between the 2, but not 2.5 - 4 degrees. I understand what Bob Brown is saying about how the thermostat's function is to control the system, but the temperature display should be fairly accurate.

I noticed a strange thing with my digital programmable Honeywell stat for my heat pump while in heating mode: Sometimes, if I change the setting one degree (reading is 69 and I knock it down to 68 setpoint), the display temp will move up by one degree to 70. I am not sure if by changing the setpoint, I am forcing an "instant read" instead of the stat calculating the temperature over a small period of time. Maybe it is moving the display temperature up one degree to insure that the heat pump does not come on for a while since I reduced the setpoint temperature. The same thing happens in cooling mode when I move the setting up one degree, the display drops one degree. This doesn't happen all of the time, but often enough that I notice it.

Hello,
Once again in simple terms. The accuracy is determined with a calibrated standard. If your stat is different than your other stats, this does not infer the stat is accurate or inaccurate. The only conclusion I can make is it is different. If you test the stat with a calibrated standard and you find it is still out of spec, then you have a reason to complain.

Its kinda funny how this is going to turn out. We, as contractors, know that all thermostats give differant readings. I have seen them as much a 7 degrees off, right out of the box. Yet we strive to give the homeowners an exact measurement(to the best of our abilities) of the exact amount of BTU's required to heat and cool each room in a house, using a heatload calculation of some fashion. So we decide that a imaginary living room requires 450 cfm to cool it properly, and i run 2- 8" drops to it. 8" @ .08 static only moves about 200 cfm. So that would only be about 400 cfm. Not enough. So what about 1-8" and 1- 9". 9" @ .08 static moves about 275 cfm so 200 + 275 = 475cfm. Too much. It seems no matter how close we would like to come to "Perfect" it never happens due to what we have available from our suppliers. I have never seen flex made in 1/2" sizes but i have seen new t-stats way off.

No matter how hard we try, it will never be perfect.
Take the thermostat back and replace it untill you find one that matches your $10.00 walmart thermostats.

Good post fluffybunnysiu, you're right on. But you forgot to add on thing, "the customer is always right"

I the temperature is 75 degrees and I set the thermostat to 74 degrees the AC continues to run until the temperature is BELOW 74 degrees (reads 73 degres. Why?

Is there a thermostat that will control to the temperature set point?

Here is a link that might be useful: thermostat accuracy

ive seen hot attic air leak thru the hole in the wall behind the stat (where the wires enter) and throw off the temp reading. i plug em with perma gum (hvac equivalent to plumbers putty) before i go messing with temp offsets. kind of a low tech fix but usually works

Digital stats are funny. Depending on who makes the stat, temperature and call for heat/cool is derived and maintained severals ways, including humidity, pressure, and sometimes even the surface they are mounted on has an effect. Even with the old mercury stats there was a 2F difference. A 70F stat got you a 72F reading. I will say the Honeywells I am using are usually right on the money when it comes to accuracy. There are so many variables with individual installations and brands it's sometimes just trial and error as others have said.

@asoh_frontiernet_net, this might be because the program is written with a hysteresis algorythm. That means that when the temperature is around the setpoint (the target temperatuere), the thermostat will continue its current phase until some time in order to reduce the number of times that the heater should be turned on and off. This will obviously create bigger inaccuracy in the room temperature but the moving parts in relays and in the heater will be less likely to wear out.