AirGradient Forum

ONE v9: SHT4x/SGP41 sub-board spacing, measured SGP41 heating effect, and S8 lock-up (#320)

Hardware extensions
#1

Hi all, I’m researching enclosure and PCB thermal design for an indoor air quality monitor and have been reading the ONE v9 KiCad files and this forum. Three things I couldn’t pin down from the docs:

  1. SHT4x vs SGP41 sub-board placement. What is the physical spacing between the SHT4x and SGP41 sub-board sockets on the ONE v9 main board, and was that distance chosen from a measurement? Sensirion’s design-in guide asks for ≥2 cm plus milled slots, so I’m curious how much the SGP41 heater actually raises the SHT40 reading in the final enclosure.

  2. Built-in temperature compensation. The calibration docs say the ONE’s raw SHT4x data is already close to reference, yet a lot of threads here report +2 °C or more. Does the current firmware apply any offset or model to SHT4x on the ONE, or is it raw? Any idea how much of the spread is orientation (vertical vs flat) vs unit-to-unit?

  3. S8 lock-up (GitHub #320). The issue shows the S8 stops answering (“Unexpected length”) until a full power cycle; an ESP warm reset doesn’t recover it, so it looks like the S8 itself hangs rather than the ESP UART. Has anyone scoped the 5 V rail or UART lines when it enters that state, or heard back from Senseair about S8 FW 1.2? For a new design I’m planning a MCU-controlled load switch on the S8’s 5 V so firmware can power-cycle it — does that match what AirGradient has seen?

Thanks — happy to share my findings back here.

#2

Hi @Ling_Lei - @Tai_AirGradient will check on this with the team and get back to you during the working week.

#3

Hi @Ling_Lei, welcome to our community!

I have discussed #2 and #3 with our firmware engineer. He confirmed that the SHT4x on the AirGradient ONE has no correction algorithm applied by default. And for the S8 issue, this should already be fixed in the current firmware. Would you mind sharing with us what firmware version your AirGradient ONE is running?

For question #1, please let me consult further with our team, and I will let you know.

Have a good day!

#4

Thanks Tai, and thanks for checking with the firmware engineer.

I should have been clearer: I don’t have a ONE myself. I’ve been going through the v9 KiCad files and this forum for my own monitor design, so I can’t give you a firmware version, sorry about that.

On the S8 issue, could you point me to the firmware release or commit that fixed #320? The thing I’d like to know is whether the fix recovers the S8 purely in firmware, or whether it still needs the S8’s supply to be cycled. That decides whether I put a load switch on its 5 V rail.

Thanks for confirming #2, and I’ll wait for the team on #1.

1 Like
#5

Good day @Ling_Lei! Sorry for my late reply. I’m back from my holiday.

Thank you for your clarification! Please give our team some time, and I will get back to you.