I'm not really liking the deadband/hysteresis settings on the Pearl thermostat. When I have the heat set to 71, for example, the heat comes on <70.5 and runs until it reaches 72.5. Is there any way to change this and have it run like a normal thermostat?
WAF is low as well, not good 
I dont think so, at least it's not anything the driver is doing.
Bummer. I wasn't implying that this is a driver issue BTW, just wondering if there was some hidden config setting in the tsat or something.
You can change it on the thermostat itself. Although I don't know what the mix/max value you can set it to is... That isn't listed.
Go into the thermostat Programming Mode and it is item #9. Default is 1.8 DegF/1 DegC.
fwiw, link to manual featuring Programming Mode descriptions...
On a vaguely related note, for heat pumps that have 2nd-stage or Aux heat, that will kick in when the temperature is 3C less than the heat set point. (That's the answer I got from Centralite support when I wrote to ask how that's governed - since there's no indicator of when it's being used).
Thank you both! I should have RTFM
Default setting is 1.8 deg F, looks like minimum is 1.0 deg, so that's where it is now. Hopefully that will help.
I wouldn't be happy with 1.8 DegF either. 1 DegF is more reasonable (and more normal in the US).
Maybe 1 DegC is a standard setting outside of the US? Don't know, but it is too big for systems with forced air heating/cooling. Probably OK for radiator/steam heat, though.
Luckily, it is changeable.
Interesting, I just realized, the manual I have does list deadband, however I remember it being an option when I installed my Pearl.
If I recall correctly it defaulted to 1 degree and I reduced it to 1/2 degree (I don't recall if it was C or F).
At the time I went through every item and checked or changed the entry. The deadband migh have been under the heating type.
If I get some time I'll try to find it again (and document it this time).
John
I just looked this up the other day for the same reasons. Almost a 2 degree swing is wayyyyy too much.
I contacted Centralite support today (and received an answer) regarding the dead band and Parameter 19 (some counter). The response is below.
According to engineering, the deadband is in the thermostat as future upgrade should the platform it is connected to ever support auto mode via Zigbee. Today, the thermostat only operates in Heat OR Cool mode and the deadband is not used.Parameter 19 is an option that is not used for anything at this time.
Interesting they say deadband isn't used. I'm curious now if I'll see any difference.
I sent a follow up eMail asking them to describe how the heating function works..... we'll see what they say.
BTW in an earlier eMail they told me the contacts were rated for 2 amps. Seems odd they don't put this in the literature. Or maybe they think its too much for the average bear** to think of.
** Yogi
This might fit my description given the company around here. 
I have been lurking at this thread all day and was curious if it would work. But being the "average bear" I figured I would let more advance members go first. I have to admit I am accustom to a more complete thermostat application like some of the ST Nest integrations. So following stuff like this is helpful and hopeful a more advanced solution comes along. I switched from nest because I really want to see my own data over time. I however do understand their model and that is my main grip. But so far I am happy with the Pearl.
Happy to be the "average bear". Can I get a handle change? LOL
I think in most cases I think it's not the ability to understand but the desire. Some folks aren't interested in the inner workings, and rightly so.
With a heating setpoint of 72, my heater shut off at 73.2 this morning. So I think it made a difference? I'm looking through logs and seeing some entries previous to the change where the temp was 73.6 or 73.7 a few minutes after the heat shut off, which aligns with the default setting of 1.8.
Although now it's reporting a temp of 74.1 a few minutes after turning off, so maybe it's not any different. Hm.
I just changed the deadband to 1.0 this morning and it is making a difference. TStat set to 67 and heat kicked on at 66.1 instead of 65.2+-. And...after kicking on, it ran until 68.0. On to calibration time now. I'm going to dunk a DS18B20 into a 'pot O ice water', make sure it's at 32, offset it if necessary and then hang it on the Pearl to see how accurate it is reporting.
I am concerned about the same thing. I had a Honeywell in a different spot an it reports a rock solid 69 degrees. However the Honeywell only reports temperature to the whole degree.
On my to do list is to setup a recording temperature device, run the Pearl for a week then swap the Honeywell and Pearl and repeat the week long test.
BTW I got into all this thermostat stuff because my heating system has 3 zones and 3 thermostats. I find (hear) my boiler often turn on for only 10 to 20 seconds then go off.
For anyone interested, I calibrated the DS18B20 to both extremes of freezing and boiling. It was within .50 so offset it and then calibrated the Pearl. This Pearl is almost perfect...1/10th of a degree low. I'll take that as acceptable tolerance!! 
The response you received is consistent with my own observations of said Deadband. IOW it's a Placebo Setting. No matter how I set Parm 19, the temperature swing/deadband, for me, is still exactly 1.8 Degrees. This translates to 0.9 Degrees below my Setpoint, and 0.9 Degrees above.
Now to see near 0.1 degree increments while running on Battery only, I have to set a rule to Refresh my thermostat every 5 minutes. For good measure I also have another Refresh Rule when the status changes to better capture the exact temp when heat is satisfied. The log, for me, shows the exact same swing regardless of the Parm 19 setting; 1.8 Degrees.
So, it makes sense to me that for a true 1.0 Degree Deadband, I'd fully expect to see 0.5 Degrees below Setpoint when Heat turns on, and 0.5 Degrees above when Heat turns off. I just don't see it. Does anyone actually have a log that shows a 1.0 degree on/off swing?
Hi,
I've yet to measure the room temperature swing with an external device, however early indications suggest the temperature control for this thermostat is below par. My Honeywell would hold the temperature right at 69 °F. Although the Honeywell would only display temperature in whole degrees F.
I'll report when I finally get to measure the two in the same room with an external thermometer.
John
Download the Hubitat app
