CO2 readings stop working "Get CO2 failed"

[jsclayton]

For the last many months I’ve had to frequently power cycle my AG ONE I-9PSL unit when the CO2 readings stop working. Sometimes it comes back on its own by restarting, but I usually find it in the pictured state and power cycle it.

I finally unmounted it from the wall to get logs and the only thing that’s showing up is the line “Get CO2 failed” and an incrementing count. Before that…works fine. At some point it just stops working.

The only reference that I can find online is this post about elevation compensation. I live at 3123ft (~950m). I can understand the compensation affecting the reading, but the sensor should continue to function shouldn’t it?

Thanks!

Logs

TVOC index: 72
TVOC raw: 31394
NOx index: 1
NOx raw: 16356
CO2 (ppm): 1281

PM1 ug/m3: 0
PM2.5 ug/m3: 0
PM10 ug/m3: 0
PM0.3 Count: 0
PM firmware version: 151

TVOC index: 73
TVOC raw: 31384
NOx index: 1
NOx raw: 16348
Temperature in C: 20.86
Relative Humidity: 33
Temperature compensated in C: 20.86
Relative Humidity compensated: 33

TVOC index: 73
TVOC raw: 31376
NOx index: 1
NOx raw: 16354

PM1 ug/m3: 0
PM2.5 ug/m3: 0
PM10 ug/m3: 0
PM0.3 Count: 0
PM firmware version: 151

TVOC index: 74
TVOC raw: 31383
NOx index: 1
NOx raw: 16350
Temperature in C: 20.87
Relative Humidity: 33
Temperature compensated in C: 20.87
Relative Humidity compensated: 33

TVOC index: 74
TVOC raw: 31392
NOx index: 1
NOx raw: 16352
CO2 (ppm): 1280

PM1 ug/m3: 0
PM2.5 ug/m3: 0
PM10 ug/m3: 0
PM0.3 Count: 0
PM firmware version: 151

TVOC index: 75
TVOC raw: 31372
NOx index: 1
NOx raw: 16351
Temperature in C: 20.86
Relative Humidity: 33
Temperature compensated in C: 20.86
Relative Humidity compensated: 33

TVOC index: 75
TVOC raw: 31384
NOx index: 1
NOx raw: 16351

PM1 ug/m3: 0
PM2.5 ug/m3: 0
PM10 ug/m3: 0
PM0.3 Count: 0
PM firmware version: 151

TVOC index: 75
TVOC raw: 31385
NOx index: 1
NOx raw: 16353
Temperature in C: 20.86
Relative Humidity: 33
Temperature compensated in C: 20.86
Relative Humidity compensated: 33

TVOC index: 76
TVOC raw: 31395
NOx index: 1
NOx raw: 16351
CO2 (ppm): 1280

PM1 ug/m3: 0
PM2.5 ug/m3: 0
PM10 ug/m3: 0
PM0.3 Count: 0
PM firmware version: 151

TVOC index: 76
TVOC raw: 31395
NOx index: 1
NOx raw: 16353
Temperature in C: 20.83
Relative Humidity: 33
Temperature compensated in C: 20.83
Relative Humidity compensated: 33

TVOC index: 76
TVOC raw: 31389
NOx index: 1
NOx raw: 16356

PM1 ug/m3: 0
PM2.5 ug/m3: 0
PM10 ug/m3: 0
PM0.3 Count: 0
PM firmware version: 151

TVOC index: 77
TVOC raw: 31391
NOx index: 1
NOx raw: 16357
Temperature in C: 20.85
Relative Humidity: 33
Temperature compensated in C: 20.85
Relative Humidity compensated: 33

TVOC index: 77
TVOC raw: 31391
NOx index: 1
NOx raw: 16358
Get CO2 failed: 1

PM1 ug/m3: 0
PM2.5 ug/m3: 0
PM10 ug/m3: 0
PM0.3 Count: 0
PM firmware version: 151
Temperature in C: 20.87
Relative Humidity: 33
Temperature compensated in C: 20.87
Relative Humidity compensated: 33

TVOC index: 79
TVOC raw: 31394
NOx index: 1
NOx raw: 16351
Last process timeout, clear buffer and last handle package

TVOC index: 80
TVOC raw: 31398
NOx index: 1
NOx raw: 16358

PM1 ug/m3: 0
PM2.5 ug/m3: 0
PM10 ug/m3: 0
PM0.3 Count: 0
PM firmware version: 151
Temperature in C: 20.85
Relative Humidity: 33
Temperature compensated in C: 20.85
Relative Humidity compensated: 33

TVOC index: 80
TVOC raw: 31401
NOx index: 1
NOx raw: 16355

TVOC index: 80
TVOC raw: 31403
NOx index: 1
NOx raw: 16357
Get CO2 failed: 2

[Achim_AirGradient]

Please contact our customer support on this.

[Rev]

It seems I have a similar error on my airgradient one. Today it showed very low (45 to 65) CO2 values. After I restarted it it shows the same CO2 — as the OP’s airgradient. It is only 6 weeks old. What should I do?

[Rev]

I contacted support, as was suggested to the OP.

[Achim_AirGradient]

Did you try a manual CO2 calibration from the dashboard?

[Rev]

There has not been a change from my side, either in hardware or software or position. Why would I try to calibrate now? It took 8 hours for the sensor to “catch itself”, showing no or too low values. Now the last 8 hours look relatively “normal” in gain and values. I keep an eye on it and reply again if something out of order happens, like yesterday.

And here is the chart of the last 24 hours. See the choppiness since 9pm after the “closed window” phase.

[luckman212]

I have this problem now with my I-9PSL-DE (running 3.3.6)

I’ve had the unit plugged in and stable for about 2 months now, suddenly the CO2 reading is null. I haven’t power cycled yet, is there anything I could/should do?

Last 8h

[MallocArray]

There is a current issue from the Dashboard site that is causing reboots for some users. The CO2 sensor take a bit to startup, so could be that your is looping, preventing from getting a valid value

[Rev]

For me it was a firmware update that restarted the device and “lost” CO2 calibration. AirGradient support told me that this is not supposed to happen. But it did. Repeatedly. So the only solution is to recalibrate the sensor again. In summer that’s OK, but in Winter opening the Windows for 20 minutes…

[RQQQQQ]

This happened to me too. I have three AirGradients (two indoor, one outdoor) and one indoor and one outdoor both stopped sending CO2 readings at the exact same time yesterday. Just the CO2 reading stopped showing up on the dashboard., otherwise they updated normally. I have a custom script that reads data from the local API too, and confirmed that the CO2 readings were just missing from the JSON. A reboot of them both fixed it. But, the fact it happened at the same time on multiple devices (yet not all the devices) is interesting?

[TeMpL]

sme here on 1 of 2 two…

ok… reboot fixed… the problem!! (so is not that critical) it’s working now… but only one died after upgrade… from, 3.3.6 to 3.3.7

[MMMurphy]

LIkewise. CO2 readings stopped MAY11. I have not rebooted.

[washley]

I just noticed my CO2 values stopped reporting last Wednesday (so ~5 days ago), the other values have been reporting.

Are there any mechanisms currently in the software to try and recover from these situations (at least reboot if the condition persists)?

[MallocArray]

I’m not aware of a feature on the Dashboard to request a reboot, but one method that may work is to adjust the Firmware so you force it to downgrade to an earlier version and then after 30 minutes or so, set it back to the version you want or Latest.

This will do a soft reboot, but sometimes these devices need a physical loss of power to fully reboot.

[MachineCode]

I also wanted to report that I have lost CO2 measurements post 3.3.7 patch. Going to downgrade and report back