Page 1 of 1

Humidity showing high value compared to local stations

Posted: Fri 08 Jun 2012 7:33 am
by Solorize
I’ve recently noticed that my Humidity value seems to be hitting 99% where other stations nearby are at
a lower value, i.e 85%, 89%.

As you can see from the below screen capture of a comparison graph. My station’s humidity is shown with
the Blue line and the other station is shown in green, which is way higher!

Image

I am unsure whether there is a problem with my sensor or something needs resetting?

Does anyone know if you can reset the value or sensor?

Re: Humidity showing high value compared to local stations

Posted: Fri 08 Jun 2012 8:59 am
by Super-T
moisture build up....blocked vents?

Re: Humidity showing high value compared to local stations

Posted: Fri 08 Jun 2012 10:26 am
by Charlie
Given your station seems to be tracking the reference station with a fixed offset, you might simply put a correction factor in Cumulus in the calibration section. The spec for your sensor is 5%, however it will be impacted by temperature and battery voltage to beyond that range. It is not unusual for humidity sensors to be off by 10% even under ideal conditions, but things like proximity to plants or standing water can impact your readings. Humidity tends to be very local. Do you really have multiple references that all agree and yours is the odd man out, or is it a single reference that might be off a bit in the other direction?

Re: Humidity showing high value compared to local stations

Posted: Fri 08 Jun 2012 12:23 pm
by Solorize
Terry,

I will have a look and see if there is any excess water around the sensor when I get home tonight.

Charlie,

I have compaired my humidity readings with 4 other station which are located close to my station
and they all seem to be lower.

My Station = 80%

Other Stations = 77%, 76%, 71%, 75%

My Sensor is within a homemade radiation screen, which is located over grass. see below.

Image Image

Re: Humidity showing high value compared to local stations

Posted: Fri 08 Jun 2012 5:23 pm
by ironeagleuk
I've always had this trouble with my humidity sensor....mine has always neared 99% whereas local stations might be at around 80%.

Mine is currently about 3m above the ground.

I've asked about the offset before, but if you do, it will never reach 100%


Here's mine right now:

Image

Re: Humidity showing high value compared to local stations

Posted: Sat 09 Jun 2012 9:58 am
by Charlie
Your location and set-up looks fine. I presume you are using a standard pair of batteries (1.5 V each for a total of 3.0 V) and not rechargables or a homemade power supply or something. If that is true, and based on your example, it seems like you are reading only about 4% to 5% high which is within the specification of the unit. If it really bothered me, I'd put in a correction factor and not lose any sleep over it.

Richard - if that graph is real, your unit is dead. 5% is one thing, even 10% is occasionally seen, but you are way over that at times, and your value never changes. Soldering iron or bin for that one I'm afraid.

Re: Humidity showing high value compared to local stations

Posted: Sat 09 Jun 2012 10:16 am
by ironeagleuk
Charlie wrote: Richard - if that graph is real, your unit is dead. 5% is one thing, even 10% is occasionally seen, but you are way over that at times, and your value never changes. Soldering iron or bin for that one I'm afraid.
It does move....it just didn't yesterday....but today it's taken a sudden drop

Image

Here's a close up of the humidity graph since 7am

Image

Re: Humidity showing high value compared to local stations

Posted: Sat 09 Jun 2012 9:11 pm
by Solorize
Hi Charlie, I have two lithium non rechargeable batteries powering the unit.

Also I put some mastic sealant around the plenum box, for the solar fan, where it joins the top bowl to make sure no water gets in and drops onto the sensor.

Re: Humidity showing high value compared to local stations

Posted: Sun 10 Jun 2012 12:28 pm
by Charlie
Which lithium batteries are you using?
Energizer Advanced lithium EA91 starts out at 1.5V and hold it nicely for many years in this application, and have been recommended on this site by several people, including me.
Energizer Ultimate lithium L91 (recently introduced) starts at 1.6V and could stay there several months in this application, before hitting 1.5 V for the bulk of it's life. I'm not sure if any other company makes this type of battery.
This 0.2 V difference could be sufficient to put humidity off a bit, but still within it's rather wide tolerance. The humidity module is calibrated at 3.0 V and the calibration values are stored in ROM, then read by the FO station and used to correct the actual reading. (The module itself will survive up to 5.5V, but the readings will be way off at those levels).
Honestly, I'd stick a 5% offset in Cumulus and not worry about it.
Advanced lithium should last 9-10 years, ultimate lithium 14-15 years. Truthfully, either will likely outlive your F.O. station.

Re: Humidity showing high value compared to local stations

Posted: Sun 10 Jun 2012 1:59 pm
by Solorize
Hi Charlie,

I took off my radiation screen this afternoon and had a good look around the sensor and it
all looked nice and dry.

I popped the batteries out of the unit and then reset everything to see if that made a
difference, so am still watching to see how it performs over the next couple of days. If
it is still reading high then I will take your advice and put a 5% offset into Cumulus.

btw.
The batteries I am using are the Energizer Ultimate lithium L91 ones.

Re: Humidity showing high value compared to local stations

Posted: Sun 10 Jun 2012 4:57 pm
by ironeagleuk
Solorize wrote: If
it is still reading high then I will take your advice and put a 5% offset into Cumulus.
Won't that mean it will only ever record 95% humidity?

Re: Humidity showing high value compared to local stations

Posted: Mon 11 Jun 2012 3:15 am
by ejay
Have a look at: Fixing 99% humidity problem on W-8681 / WH1080. In your case, your humidity is not always stuck at 99%, but rather like a meter that reads too high and keeps hitting the end of the scale. Your meter is still working, but your readings are useless when the actual humidity is over 85%. Thus, when your station says 99%, the actual real humidity could be anywhere from 85% to 99%.

The manufacturer has not calibrated your sensor properly. The capacitance-to-frequency conversion table stored in the EEPROM is incorrect. This explains why your humidity readings are still linear when compared to other stations when the humidity is less than 85%. Adding an offset or scaling value to Cumulus wouldn't really solve your problem as you have still lost your 85% to 99% readings.

If you're good at electronics, you could probably try correcting it by soldering capacitors in series and/or parallel with the humidity sensor to try to bring the readings down. You may need to experiment with various values from 1pF to 100pF. Good luck.

Re: Humidity showing high value compared to local stations

Posted: Tue 12 Jun 2012 12:35 pm
by Solorize
I've just adjusted my offset for the Humidity to a value of "-4" to see if that sorts my problem.

One thing now that I have adjusted it, when it was previously showing 99% it will show 95%,
but does that now mean that when the sensor is indicating 99% it will show 95% in cumulus
and it will never get any higher than 95% due to the sensor never getting over 99%.

If so I may just put it back to NO offset and just live with the fact that it may top out
at 99% every now and again.

Re: Humidity showing high value compared to local stations

Posted: Mon 06 Aug 2012 11:35 pm
by Super-T
My Outside Humidty has been on 99% for a week or so. Always thought it was reading high but now it seems to have decided to stay at 99%. Other local stations show quite a variation with only a couple showing 99%, mine and another fine offset station.
The question is, has anyone found a way to purge the sensor itself? Or perhaps what sort of liquid to purge with?
It's not a question of wrong initial calibration, although that may be out too, as the station has been running since July 2008.
I could play around with the sensor "total capacitance" but probably better to just to replace the transmitter.
Last weekend I took it all apart but there was very little corrosion, surprisingly enough.