is this normal?

[mcrossley]

My fault. I've fixed that now, and also added the missing Temp & Hum for the CO2 sensor that I added to the historic catch-up for all Ecowitt stations, but forgot to add the Ecowitt Cloud station current data decode.

I'll ping you a test build to try soon.

[mcrossley]

Screenshot 2024-01-24 165703.png

[Nossie]

Thanks Mark!

seems fixed so far!

[mcrossley]

Ok, keep an eye on it please

[Nossie]

Sorry Mark, thought it might have been good last night but Hans just informed me PM values (both 2.5 and 10) are still goosed

[mcrossley]

In what way? What does the dashboard graph look like?

[Nossie]

apologies - here we go -

looks like a rollercoaster

[Nossie]

hold up - I made the change last night

but if you look here it might be sorting itself out (can be seen in the uploaded screenshot too

go to www.stewartonweather.com / extra / extra sensors / extra co2 -- I'll keep an eye on it but as you can see it becomes less drunk just before 7:30 - might be sobering up before work? HR referral for a disciplinary

Edit: - nope, I think it's still drunk.

[mcrossley]

Could you zip up the MXdiags folder for me please.

[Nossie]

here we are

strange zip is just under 10mb but wont let me attach - I've uploaded it here

https://www.stewartonweather.com/MXdiags.zip


cheers.

[mcrossley]

Thanks, I have found a couple of issues - one me, the other a result of the data sent.

If you have missing values in history, then for example Ecowitt send "-" rather than null (The fact that Ecowitt send numerics as strings is poor anyway).
The JSON parser is expecting a number (integer in this case) or null and curls its toes up at the string "-" so no history for that sensor gets loaded

I'll have to pre-massage the JSON text that Ecowitt send and replace all "-" with null before trying to parse it as JSON.

Ecowitt do not send the 24hr average values for the Combo PM values in the historic data (they do in the current data ), currently they default to zero during catch-up. Not sure what else I can do other than maybe set them to the current value, but then it isn't an average. I think leaving them at zero is best as it is obvious there is missing data.

[Nossie]

no problem glad to help - although once this is sorted I might slink away from the api and go back to a direct feed

PM me a new build when you have the time and I'll check it out