Relative Humidity at odds with station

[nickjwhite]

Good morning

For some time I have noticed that on the charts and in the software, the RH occasionally seems to lock at various figures for a few days, then right itself again. I had assumed it was the sensor or the station, but I have noticed today (for example) that the station is showing RH at 73%, but the software interface (and therefore everything after that) is showing 97%. All other readings between the station and the software are correct. My charts are showing a straight line at 97% for a while now.

The only change I have made this morning as an experiment is to uncheck the 'Use 100% for 98% humidity' and I will watch this to see if that actually makes a difference.

I cannot find a pattern for this, and the 'straight line' syndrome runs for anything up to a couple of days before coming back to correct readings.

Is this a known problem at all, or do I have an oddity here?

Many thanks as always.

[steve]

Please zip up the diags folder and attach it.

[nickjwhite]

Hi Steve

Diags zipped and attached.. thanks

[steve]

You've set a humidity 'spike removal' value of 4, and your pressure jumped - and stayed there - by more than 4:

28/06/2012 07:07:43.649 : Humidity difference greater than specified; reading ignored
28/06/2012 07:07:43.649 : Old value = 97 New value = 90 EWhumiditydiff = 4.0

[nickjwhite]

Hi Steve

Thanks for the reply

I understand what you say, but I have two questions.

1 Should I change the spike removal interval (do you have a recommendation?)?

2 What can I do to 'force' the RH to record correctly now (rather than waiting until it decides to show the correct reading)?

On the whole this is all working brilliantly, and I have to say, for what is now a cheapie weather station... it far out performs the old laCrosse I had at 300 quid.

Cumulus is brilliant... and I have very many positive comments about my weather site. Living right on the coast I have kite surfers, fishermen, dog walkers and all sorts use it regularly. Thank you as always for a first class piece of software. You must be due another donation from me about now. Couldn't do this without you. )

[steve]

1 Should I change the spike removal interval (do you have a recommendation?)?

Unless you know that you actually get spikes, there's no point in using a spike removal setting. When you get spikes, you'll know what value to use to remove them. You may have to make a compromise with the size of the spikes that you remove. The spike removal facility is very basic, but does work well for genuine spikes. It does carry with it this inherent risk of detecting a permanent "jump" as a spike. At some point, I will probably look into writing yet more code to remove faulty readings, in a more intelligent way.

2 What can I do to 'force' the RH to record correctly now (rather than waiting until it decides to show the correct reading)?

Restart Cumulus.

[nickjwhite]

Thanks Steve

I have removed the spike setting

the other one I guess was an obvious ) Monday morning - brain not engaged