WXforum.net
Weather Software => Cumulus => Topic started by: PaulMy on December 24, 2018, 09:11:41 AM
-
Mark has posted the new current build https://cumulus.hosiene.co.uk/viewtopic.php?f=27&t=17213
and further https://cumulus.hosiene.co.uk/viewtopic.php?f=27&t=17214
Enjoy,
Paul
-
Any changes for humidity calibration? I wouldn't hesitate if I could find a workaround for the high Davis sensor humidity bias.
Mark mentioned he would look at making the calibration more like WD with a better slope.
-
Any changes for humidity calibration? I wouldn't hesitate if I could find a workaround for the high Davis sensor humidity bias.
Mark mentioned he would look at making the calibration more like WD with a better slope.
Yes, I added a second order (slope) correction factor for humidity, it works the same as the existing temperature second order factor - you have to add it to the ini file manually.
-
Any changes for humidity calibration? I wouldn't hesitate if I could find a workaround for the high Davis sensor humidity bias.
Mark mentioned he would look at making the calibration more like WD with a better slope.
Yes, I added a second order (slope) correction factor for humidity, it works the same as the existing temperature second order factor - you have to add it to the ini file manually.
Thankyou =D>
Now to make the changeover.
-
Yes, I added a second order (slope) correction factor for humidity, it works the same as the existing temperature second order factor - you have to add it to the ini file manually.
So it is a multiplier now, rather than an offset?
-
It is a quadratic function. CMX already had a multiplier (gradient) and offset for humidity, this adds a multiplier of the square value. So maybe my using 'slope' is confusing, the new mult2 will transform the curve...
CalcHum = (hum*hum*mult2) + (hum*mult) + offset
You will need to use something like Excel or MathCAD to find the correct values for your sensor to plug-in to get the correction you desire to match a calibrated value.
-
CMX already had a multiplier (gradient) and offset for humidity
I didn't realise this - didn't see it on the Wiki (I checked both sections for both versions of Cumulus). Can you supply a link to it on the Wiki, or the tags to use if it really isn't there, please?
-
That's good information, I downloaded and was clueless how it worked still am actually...LOL First time I ever looked at MX.
I'll need to play with it, so my understanding the INI file comes over with the original 1.9.4 Cumulus because I don't see it anywhere else. Both versions are mixed in the same folder and looks like MX has adopted the same settings as before.
Even after I added the new offset in INI file editor I don't see in MX window, hope that's normal.
-
Freddie do you need this info?
- Added second order humidity correction factor, works the same as the temperature. It has to be set manually in Cumulus.ini.
- Cumulus.ini
[Offsets]
HumMult2=0.0
You need to add the HumMult2=0.0 to offsets in INI file I believe.
Link is here: https://github.com/mcrossley/CumulusMX/blob/master/Updates.txt
-
Hi Randy,
Yes, that's exactly what I was after. Thank you!
-
Hi Randy,
Yes, that's exactly what I was after. Thank you!
If you figure it out and what numbers to use please share.
-
Hi,
Recently updated to ver 3045, and I have noticed that my twitter account no longer updates, I have changed no settings and have not had this issue in the past with updating Cumulus files.
Any ideas would be gratefully received.
Thanks.
Nick. dw7240.com
-
I think it is to do with twitter's change in limits as CumulusMX is classed as an app and limit is per app not per user tweeting.
More discussion about this here:
https://cumulus.hosiene.co.uk
Sent from my ONEPLUS A6003 using Tapatalk
-
Hi Randy,
Yes, that's exactly what I was after. Thank you!
If you figure it out and what numbers to use please share.
I actually need a cubic equation - not a quadratic :-(
-
Hi Randy,
Yes, that's exactly what I was after. Thank you!
If you figure it out and what numbers to use please share.
I've had a play with this and used Excel Solver to work out the coefficients. I have used the following:
HumMult2=0.005535
HumMult=0.484336
HumOffset=-0.619714
This seems to be working very well. It corrects a slight under-read above 85-90%, and (more importantly) corrects the larger over-read below 75-80%.