Gateway temperature accuracy
Hello all, So I have been trying to fix this issue for a while and I have a support ticket too but I feel like their support is not being fully honest about an issue. I only saw one other mention of this so I thought it's worth asking here. My sensor pucks have been measuring temperature very accurately and within 0.2F of other temperature sensors and pucks that I have. Recently I converted one of my sensor pucks to a gateway due to low signal strength in a room. Since then I have started noticing the gateway temperature readings are completely off. They are usually lower by anywhere from 1F to 4F of any other sensor next to it. Recently I had one hour of data where it was off by 7F. I am pretty sure it's off because the room is quite hot and the gateway is reading a very low temperature. Based on what support said I tried using calibration but I don't think it's working as the readings on gateway are not increasing correctly. As the room gets hotter gateway needs a higher calibration. I am thinking this might be a bad Puck but two things make me think otherwise 1. This same Puck was super accurate as a sensor. 2. Another gateway has a similar issue but the difference varies from 0.1F to 1F depending on how high the temperature is. All sensor pucks are always within 0.2F of the numerous offline sensors that I have. I am continuing to work with support but I felt I should check here. This seems like a design issue and was wondering if others had seen it. Thanks
-
Official comment
Hi Everyone,
A Puck can be calibrated by going into the Puck settings in the Flair app. When you calibrate - or convert between Sensor <-> Gateway - give the Puck some time to adjust.
If you’re seeing temperature spikes, get in touch with our Support Team at
flair.co/contact
All the best,
Finn
Comment actions -
I had 3 sensors in the same room over night. In the morning they are all wildly inaccurate - 71f, 74f, and 83f. I feel this issues may be related to the sensors themselves not being very quality, coupled with the slow refresh rate of the e-ink screen. I'm unsure if this could even be fixed with a firmware update.
I find is also sad and terrible that i'm responding before a rep from flair is.
-
Every time I'm about to pull the trigger on a Flair vent, I come check the forums and find a lack of response and support from the company. I have to weigh this against my desire to automate the heating/cooling of certain rooms and ultimately end up just waiting.
Flair claims to be financially solvent and doing just fine, but there still hasn't been any sort of update to the cloud software vs local option thread in several months, despite many users replying. I wish we had more options or at the least, better ones
-
Thanks for the replies.
A bit of an update - their support has been responding throughout this process and I created this post to see if it happened to others on a consistent basis. I don't really need them to respond here when they are giving me elaborate answers on my support ticket.
In my case this specifically happened after converting to a gateway.
So I ran a test and their support verified the graph. I did indeed have a bad puck which for some reason showed varying temperatures at specific times and was consistent otherwise.
They sent me a return label and are shipping me a replacement.On flair offline mode I completely agree. I did my fair share of research before picking flair. I also have a ton of devices that work even if the internet connection is gone and I prefer that for a simple register. Also once flair hits market saturation their business model can't justify the cost of running the cloud service and I am worried they will switch to a subscription model. Look at all the security cameras that started with unlimited cloud storage and changed later on. So it's really important they provide cloud free control of at least the vents.
Here's my take on why I still chose flair
1.The Hvac guys I talked to don't care about my house. All they recommend is mini splits everywhere which can cost anywhere from 10k to 20k depending on the rooms. Very few (just 1 of 4 I spoke to) was interested in zoning the house but was honest that it meant ripping apart a newer house in several places.
2.Nest is the worst. They don't even allow you to choose a specific sensor at a specific custom time window. Their fan settings are minimal.
The worst part is their differential temperature is 1F and can't be changed. My system is turned on and off almost 20 times a day for 8 hours.
With flair and a custom nest schedule I brought it down to 8 times a day and a total of 4 hours a day.
3. I looked at keen but I don't see any activity on any of their sites or forums. They have no availability of products either.So far flair has been working great for me if not for this temp inconsistency. The vent build quality is very good. Love the stats it provides and proof that it works.
So for a under 1k investment I will be glad if this system lasts 3 years. -
I just converted from Keen to Flair. The Keen system would be down for days and no-one would respond to support requests. I got tired of having to manually adjust the vents. Flair is AMAZING compared to Keen.
I have the same issue described initially in this post. When I was setting up my Flair system I converted some pucks from sensors to gateways and I noticed a 1-4 degree lower temperature difference immediately. I tried the calibration but that didn't seem to make a difference. I just adjusted the room temperature down 1-4 degrees.
-
Unfortunately I never did overcome this issue completely. And I resorted to buying a couple of extra pucks as dedicated sensors.
This is just my guess as an engineer. The networking stack needed to run wifi connectivity generates sufficient heat to impact the Puck. Instead of redesigning I think they made software based calibrations to account for the warming.
I don't think the calibration works linearly so that's why we see temperature discrepancies at different temperatures.Honestly I gave up on having the perfect system and use it for controlling the vents manually.
-
We just purchased flair vents for our downstairs. So far we have the four bedrooms installed with pucks and vents to work with our nest thermostat. 3 of the bedrooms are gateways and 1 is not. For two of the gateway rooms, flair shows a temperature reading that is 10 and 12 degrees too low. This makes it impossible to get these rooms to cool since I can't get accurate readings to trigger the vents to cool. The room without the gateway does not have this problem. Today I set the desired set point really low to compensate for the 10-12 degree false differential. The vents opened but no air came out. The pucks have been calibrated and the rooms have been deleted and readded with no improvement. We have a ticket started with customer support but so far they have just asked us to make sure the temperature readings provided by flair are actually incorrect. Flair says one room is 60 degrees and it feels about 75. I know 60 would feel freezing. I've ordered separate thermometers to get a true reading.
Any suggestions on what to try while I wait for customer support to help fix this? It seems unlikely that two of our pucks would be duds.
-
My understanding (this is my own opinion) is the WiFi module in the gateway produces enough heat to affect the sensor reading. I think (in my opinion) this was fixed by offsetting the sensor reported temperature by some value and this is not a linear fix. I think and based on my observations the calibration helps if your room temperature doesn’t fluctuate much. But if there’s a wide swing and with WiFi on, the calibration stops working correctly.
I decided to add more pucks as sensors and kept gateways to help sensors and vents reach them. The pucks are expensive and if you are using ecobee it might be better to use their sensors instead.
Anyway tl;dr pucks as gateways - doesn’t work well in all situations. Pucks as sensors (with batteries) has been super accurate.
Please sign in to leave a comment.
Comments
8 comments