I received the outdoor monitor a few weeks ago and the temperature is never close to correct. I tried using the default formula and custom offsets but even when I offset it to be correct, its way off come night time out day time. When using the default formula it’s usually 5 to 12 degrees Fahrenheit to warm. I have it attached to the side of my house under a covered porch. Sun light doesn’t shine directly on it.
I may have the same issue – I also get too-warm values from an outdoor unit, a pretty constant 3.5C. Two AirGradient indoor units are each 1C too high.
I had the outdoor sensor outside until it became obvious it was too high. I have a bunch of other sensors (uhoo, aranet4, six lumi sensors, two indoor airgradient and one outdoor airgradient). As a test they are now all sitting next to each other on a table inside. The non-airgradient sensors all agree within 0.5C.
Switching between “Standard AirGraident Open Air Calibration” and “Raw Data” has no apparent effect. Setting x 1.181 - 5.113 as a custom formula matches the “standard” calibration. I added a custom -1C offset to the two indoor airgradient sensors and a custom -3.4C offset to the outdoor sensor, which is clearly a hack, but brings the current reading in line with the others (currently 74F). The other thing I tried is unplugging the outdoor sensor for 20 minutes and plugging it back in. It comes back cool and warms back up over about 15 mins, which I think confirms the issue is electronics heating the unit.
I did my initial setup with the Home Assistant integration, now using app.airgradient. Is this a known issue? What else can I try?
On the Outdoor AirGradient that you have inside for comparison, it is raised up so airflow can happen under the device? just sitting on the table will cause false readings.
Similar with the indoor AirGradients, they should be upright using their stands or mounted to a wall. If laying flat on the back, it can impact readings due to heat building up inside of the enclosure.
If you used HomeAssistant first, it changes the configuration source to Local, so none of the AirGradient corrections are going to be in effect even if you change it on the Dashboard site. You can try changing it in Home Assistant under Settings>Devices>AirGradient and finding your device and changing it to Cloud
Then it will start using what you have configured in the Dashboard app site. If you want the Local configuration source, there are API calls you can make to enable it on the device.
How are you adding the custom offsets? Making a template sensor in Home Assistant so it gets a corrected value? Or anything else special you are doing?
Similar to the other poster on this thread, are you using Home Assistant integration? If so, it likely set the Configuration Source to Local which doesn’t apply the correction algorithms by default.
You could try changing the Source to Cloud in Home Assistant if applicable and see what happens
Thanks for the tips. I did start with Home Assistant, but before posting I found another forum post and added them to app.airgradient and set configuration to cloud – so that’s already taken care of.
I’m looking at the charts in HA, but the numbers match the numbers reported on app.airgradient, and I am changing the “Temperature Calibration Formula” there. I know it’s working because I can set a custom offset and I see it take effect everywhere.
I changed the mounting as you suggested – I have them all mounted to a vertical sheet of plywood now, a few inches apart from each other, a foot above the table. It didn’t significantly change the readings.
The big issue is the O-1PST’s temperature reading is about 6F too high and I can’t figure out why.
I turned off my custom adjustments so the two indoor units are set to “Raw Data” and the outdoor is set to “Standard AirGradient Open Air Calibration”. (Setting it to “Raw Data” gives the same numbers, which seems wrong…)
Here’s the three AirGradient sensors sitting next to each other:
At 11am I tried turned the O-1PST off for 15 minutes, then back on. It starts cold and heats up by about 12F.
I appreciate any ideas I can try.
I’ve also come to conclude that the indoor Airgradient ONE should have default calibration for temperature, raw is not acceptable. I don’t know what the full formula should be but at around 15C I need to subtract 1 from it to match two ordinary thermometers I have from different manufacturers that closely agree with each other. (Device sits on table in vertical orientation using the supplied feet, away from walls.) There are after all a few watts of heat dissipating within a partially vented enclosure and potentially conducted on the PCB. For now I have:
"atmp": { "correctionAlgorithm": "custom", "slr": { "intercept": -1.0, "scalingFactor": 1.0 } }
But I’m wondering if, like one blog article suggests, the effect is temperature dependent (and thus not observed in warm conditions) and a slope other than 1 should be used.
so is there any solution for ourside unit and temperature… as it is +1 to +3 off …mybe to add a real tempeature sesor to it? which would be able to measure correctly… not all of us is living in tropical area…
raw data is even +7 off… as today as i write calibrate hanna +/- 0.1… reads -6.2 and outdoor unit on same spot - 2.7 that simply not acceptable…
should i calibrate it using indoor unit as reference… and mount it to the same spot fo a 12h ?
Was your Indoor device one that you assembled? Any chance that your temp sensor was installed in the upper right corner instead of the lower left? That can often result in temperature readings that are off by a few degrees.
I don’t have other temp sensors I’m comparing to, so can’t say for sure if mine is off, but we have seen in the forums multiple times were they get installed in the wrong spot accidentally
In your Settings, do you have the Temperature Calibration Formula set for the Open Air Calibration
https://app.airgradient.com/settings/place?tab=1
yes, i have it. if i switch to raw then it show even another 4C >n + direction…
does it low wifi signal heat more?
no both indoor devices are ok… it is outdoor unit that it has a problem… it ha no real tem sensor… it user temerature senso inside particle senor … i fahr i understand this correctly…
You are correct that the outdoor Open Air model has a temperature sensor in the particle sensor, but I don’t know that it isn’t a real sensor, it is just inside of another, plus these sensors are inside of the Open Air unit itself. They have a temperature correction that should be compensating for this, but if you don’t think it is accurate, then discussing here is the right place.
Their science team can try to see if an updated algorithm is appropriate or if there is something else with your particular setup that is causing a mismatch.
Any chance that your temp sensor was installed in the upper right corner instead of the lower left?
There’s always a chance I suppose, but the instructions were really emphasizing to be careful and get this right and I did so. In addition I thought there were other symptoms of the swap which I haven’t seen. VOCs readings have behaved as expected.