cancel
Showing results for 
Search instead for 
Did you mean: 

CO₂ sensor

h_nakamura
New Contributor

 

Actual Measurement Result
When CO2 was injected into the closed space with the PAS CO2 Sensor, the CO2 concentration in the GUI increased immediately.
16:10:49(JST)/500ppm ⇒ 16:10:59(JST)/1500ppm
After that, it stabilizes at 32618ppm.
Leave PAS CO2 from enclosed space to open space (≈400ppm) at 16:16:39(JST)
16:21:49(JST)/32765ppm
16:21:59(JST)/32671ppm
16:22:09(JST)/18915ppm
16:22:19(JST)/7576ppmm

16:31:39(JST)/436ppm, Finally back to the CO2 concentration of the open space. Why does it take so long, 15 minutes, to get back to normal?

 

Translated with www.DeepL.com/Translator (free version)

0 Likes
8 Replies
AjayB_76
Moderator
Moderator

Hi, 

I have a question on the below comment: 

16:31:39(JST)/436ppm, Finally back to the CO2 concentration of the open space. Why does it take so long, 15 minutes, to get back to normal?

Question: The response time for sensor spec is 15sec. In your measurement, can you please let us know if 63% of final value is reached within 75 sec, We will analyze the log and get back to you. 

AjayB_76_0-1634880442510.png

 

0 Likes
AjayB_76
Moderator
Moderator

Hi, 

can you provide additional information on how the testing is being done? Please let us know the source of CO2 provided while testing. Since CO2 reading is showing as 32765PPM, the assumption is that sensor is getting saturated with CO2. 

If there is any confidential information that is to be shared, let us know and we can work through a secure channel as well. 

-Ajay

0 Likes
h_nakamura
New Contributor

Did I inject too much CO2 because I didn't know how much it was?

Once saturated, does it take such a long time to measure the concentration normally?
If so, what is the mechanism?

0 Likes
h_nakamura
New Contributor

I attached the document regarding the test condition.

0 Likes
AjayB_76
Moderator
Moderator

Hi, 

Thanks for the detailed explanation of the test setup. 

eCC2 is typically not accurate since it determines CO2 through another means by measuring alternate parameters and is not a direct representation of CO2 in the chamber. The accuracy of CO2 may be different. 

Can you provide additional details of the gas present in the can of gas? Does the can of gas has other components? 

The 100,000 PPM of gas may be injected into the chamber if the system is full of CO2. this is an area which we have not tested and maybe saturating the chamber. 

We recommend to do performance measurement of the system using the method in the app note - https://www.infineon.com/dgdl/Infineon-PASCO2_recommended_performance_evaluation_methodology-Applica...

Also for a more confidential discussion, can we move this to a secure channel to discuss additional details. Please let us know. 

-Ajay

0 Likes
mht
Employee

What about the following hypothesis?

The testing bag has a small volume. Using the CO2 gas can is not so precise, most likely the sensor cavity is filled with CO2, therefore the sensor is saturated and a high value reading.

After taking the sensor out of the plastic bag, the sensor opening is pointing upwards. CO2 density is ~1.5 heavier than air. On the desk there is not so much air movement, therefore the only mechanism(except diffusion) helping to reach an equilibrium with the ambient is the IR emitter, heating up during a measurement and cooling down in between. When the IR emitter heats up, it heats up the gas inside the sensor and "pushes" out some CO2 molecules which are then replaced with air. This mechanism alone will take a long time to reach equilibrium with the ambient air. 

Additionally there is also the gas diffusion which helps reaching an equilibrium with the ambient. The white membrane (PTFE?), which seals the sensor is slowing down this process. 

In the end, when the sensor cavity  is filled with CO2, I don't know which mechanism is dominant (diffusion or "IR pumping").

Another idea to try out: maybe creatig some air flow around the sensor will also help improving the response time. 

0 Likes
AjayB_76
Moderator
Moderator

We dont think "IR pumping" is the reason for slow return of normal value of CO2. Even in ambient conditions where air flow is at the desk, the response time spec of the sensor should be valid (T63, as specified in datasheet). Above 32k PPM is an area where we have not tested the device and is outside the operating condition. 

0 Likes
h_nakamura
New Contributor

Thanks for your reply. I understood it. I'll try to evaluate within the operating condition. This time, I'll be careful not to saturate it.

0 Likes