Vacuum cleaning the air sensor stabilizes abnormal readings again
38 views
Skip to first unread message
Alberto Villa
Jun 27, 2017, 5:46:41 AM6/27/17
to Safecast Air Quality Discussion
Hi to the team,
as recently posted (June 19th), I've got abnormal PM values again. This time it was a limited perdiod with sequence of zeros followed by bursts of really huge NEGATIVE peaks. Furthermore, since the end of these extremely unstable events the PM values remained fluctuating at too high levels.
So, after an electrical check of the Alphasense/Motherboard connections, recently (June 19th) I repeated the gentle vacuum-cleaning of the air sensor, exactly like described in my previous post "PM abnormally high values issue. Apparently solved", April 30th. Same results were obtained, as shown in the attached graphs: vacuum-cleaning brings the PM readings back to the expected values range. It seems to confirm that cleaning the sensor can have a positive effect on the accuracy of the readings (and that the first time it was not a mere coincidence...).
At this point some questions arise:
a) Is there something wrong in my specific sensor?
b) Is there a potential weakness in the Alphasense product?
c) How reliable and meaningful are the readings generated by my air kit?
d) What happens if the sensor is "overloaded" by dust?
e) Should it be periodically vacuum-cleaned?
f) Are there special precautions that must be followed in the installation/protection/orientation of the sensor air intake?
Thank you very much and best regards,
Alberto
as recently posted (June 19th), I've got abnormal PM values again. This time it was a limited perdiod with sequence of zeros followed by bursts of really huge NEGATIVE peaks. Furthermore, since the end of these extremely unstable events the PM values remained fluctuating at too high levels.
So, after an electrical check of the Alphasense/Motherboard connections, recently (June 19th) I repeated the gentle vacuum-cleaning of the air sensor, exactly like described in my previous post "PM abnormally high values issue. Apparently solved", April 30th. Same results were obtained, as shown in the attached graphs: vacuum-cleaning brings the PM readings back to the expected values range. It seems to confirm that cleaning the sensor can have a positive effect on the accuracy of the readings (and that the first time it was not a mere coincidence...).
At this point some questions arise:
a) Is there something wrong in my specific sensor?
b) Is there a potential weakness in the Alphasense product?
c) How reliable and meaningful are the readings generated by my air kit?
d) What happens if the sensor is "overloaded" by dust?
e) Should it be periodically vacuum-cleaned?
f) Are there special precautions that must be followed in the installation/protection/orientation of the sensor air intake?
Thank you very much and best regards,
Alberto
ta...@kithub.cc
Jul 15, 2017, 6:27:32 PM7/15/17
to Safecast Air Quality Discussion
Hi Alberto,
What kind of power plug are you using? Have you tested the electrical noise on it?
-Tara
Alberto Villa
Jul 17, 2017, 2:43:31 PM7/17/17
to Safecast Air Quality Discussion
Hi Tara,
The power plug is the one I got with the kit, model SFE-5V2AU. I had to modify the AC interface, because it came with the NEMA blades instead of the EU cylindrical pins (see attached photos).
No, I have not checked the switchng noise of this power supply. Do you think some low-pass filtering should be placed along the 5 V input path?
Take into account that my AC mains distribution network is surge-protected.
By the way, recently I have measured a similar device in a different application (12 V - 2A). In that case the ripple was 25-30 mVpp at 28-29 kHz.
Thank you very much, and best regards.
Alberto
The power plug is the one I got with the kit, model SFE-5V2AU. I had to modify the AC interface, because it came with the NEMA blades instead of the EU cylindrical pins (see attached photos).
No, I have not checked the switchng noise of this power supply. Do you think some low-pass filtering should be placed along the 5 V input path?
Take into account that my AC mains distribution network is surge-protected.
By the way, recently I have measured a similar device in a different application (12 V - 2A). In that case the ripple was 25-30 mVpp at 28-29 kHz.
Thank you very much, and best regards.
Alberto
Reply all
Reply to author
Forward
0 new messages