Cumulus Support

Great, so I'll only need to cut the printed circuit and to solder the wires from the printed board to the potentiometer (are 10kOhm good?).

Has someone tested the mod?

Sorry for my english

Stefano

The author said his reading was 20% low, no mention was made about 'range', so it is to be assumed that his unit was 20% low across the spectrum of readings (except the very top end where all of these circuits suffer)

Looking at the manufacturers suggested circuit and the authors modification supports this.

So unless the readings are consistantly low, this mod is not for you (It can be compensated for in Cumulus anyway)

If you do need to try this modification, no cutting of PCB track is required. The authors 'final' value was 5.8MegOhm across the sensor, (That was for a 20% low error) So pick a potentiometer to suit.

this is the scheme - try it:

http://www.meas-spec.com/downloads/HS1101LF.pdf (3'th page)

Well, thanks to all...
the scheme reported by sdsi not agree with the mean reported by beteljuice.

Betel, I don't understand this of your message : 'no cutting of PCB track is required. The authors 'final' value was 5.8MegOhm across the sensor' (this is not according with the technical informations reported and... what do you mean with 'across the sensor'?)

good night

from author's site:

"The W8681 uses a humidity to frequency circuit based on the sample application given on page 3 of the" scheme http://www.meas-spec.com/downloads/HS1101LF.pdf

and

"The calibration potentometer RV1 has been omitted"

so - we ought to add the calibration potentometr or resistors (he used 5.8MegOhm but we must to try exactly values for our sensors)

"'no cutting of PCB track is required" becouse we can easy solder potentometer in parallel - I think

what du you thin about this idea?

I don't understand where I must to solder the 5.8MOhm resistor without cutting the printed circuit ... , I can't find it in the scheme where the potentiometer is present.
Many thanks

Hello,
Interesting what you made about correcting the humidity sensor with a help of a potentiometer.
I checked a lot of humidity sensors and all have certain degree of differences between them.
Resistive and capacitive sensors.
So a calibration is needed in the most used part of the sensor range in this kind of cheap meteorological stations.
If you adjust the sensor device in order to read 100 % perhaps you will have error when reading low humidity values.
That is the reason most meteorological programs have a calibration option.
I understand that we want to read the right value in the console.

Humidity and temperature sensor data can also be altered with the presence of direct sunlight which affects directly by heating the electronic box where it is located.
I have very fast changes in temperature reading when a cloud is shadowing the sun. This is not good.
The Fine Offset shell where the temperature sensor is located is not well ventilated and this is an issue design.

Despite all this we can learn a lot about monitoring the local weather where we live.

With best regards,
Daniel

The authors modification was (for his 20% low correction) a 5.8 MegOhm resistor connected (across) to each leg of the sensor. This is NOT as per the suggested diagram of the manufacturer.

This modifaction creates an 'offset', so is only worthwhile if your console is displaying
eg. 15% low through the range 10% => 85%. below this range the software usually gives up, and above this range nearly all 'home' weather stations start to read progressively 'low' upto (Max ?) 98%.

If you are unconcerned about your console display, then Cumulus can provide an 'offset' to compensate. You can also 'multiply' the value (use with caution !). You can even tell it that 98% is really 100%.

One of the other apparent problems with this unit, is the 'recovery' period from 'wet', which may be many hours ! - Which is why you may have read some of the forum users have experimented with improved ventilation. (I do hope they also changed the 'screen' / housing)

Could we have some feed-back on the success (or not) of those mods guys ?

yes - the Cumulus software has some options to calibrate e.g humidity values. but I want to display on LCD of WH1080 station the correct value for humidity parametr too.

I don't undarstand what does it means "across"? Is it connecting resistor(s) to each leg of the sensor?

I think that there is not the need of an offset. My weather station presents a 'bug' for high (>90%) values of humidity; in the range 20-80% it appears reasonably true.

sdsi wrote:... I don't undarstand what does it means "across"? Is it connecting resistor(s) to each leg of the sensor?

beteljuice wrote: .. connected (across) to each leg of the sensor.

Yes

taddeus wrote: .. My weather station presents a 'bug' for high (>90%) values of humidity; in the range 20-80% it appears reasonably true.

Then you may want to try a 'ventilation' mod if the error in the top of that range >5% low, else try the 'multiplier'
Calibration / Humidity / Multipliers eg. 1.02 (ie. x 102%).

I'd like to see the correct value on the display
I've already done a little mod on the case of the sensor (drill&drill), the measures appear improved but not enough I'd like.
A little refresh on electronic is required but... could you report here the document where the author tries the 5.8MOhm resistors? I can't find it.

Many thanks

like you - I'd like to see the correct value on the display. all photos (and nothing more) are on this page http://www.flickr.com/photos/bn37ej/set ... 701058879/

I wrote an message to their authors - I'm still waiting for reply.

beteljuice - can you write exactly what we can do with this fix? I know you prefer Cumulus settings to fix humidity values, but if you can help us with your electronic practice it'll be nice thanks

Step through the larger photos http://www.flickr.com/photos/bn37ej/259 ... 701058879/

A limited explanation is beneath each picture.

Essentially, first of all you need something to compare the RH% with !

Solder (carefully !) eg. a 10MegOhm pot across the sensor (preferably somewhere on the PCB copper track).

Adjust to suit - that's it.

Second thoughts:

The manufacturer of the sensor makes the point of saying that consideration should be given to the environment of the sensor and its reference oscillator - including the conformal coating of the PCB (affects the capacity, which is what makes it 'tick')

The ferrite rod aerial is 'hot glued' in a very slap-dash manner nearby, It might be causing the differences to the expected values.

You might as well do this modification regardless

https://cumulus.hosiene.co.uk/viewtopic.php?f=13&t=807

Then see if you really need to add a potentiometer to 'fix' your RH% !