Topic: replacement humidity element

[jerryg]

I have a bunch of sensors that only have problems with the humidity, the temp is ok, and i was wondering if anyone knows if you can just get the element to repair the unit, it has the number 3 on them.

[wxtech]

I have 3 different types of T/H sensors for the VP & VP2.  I have one VP humidity sensor that dashes out on the console when the temperature is very hot, but works OK at night.  Some of the old humidity sensors have potting compound on the bottom solder side of the board.  Maybe to prevent attempting an unauthorized repair.

[DanS]

Chances are you may have the type of temp/humid sensor with one of these inside since you said that humidity only went bad   . It may have a plastic grill/case enclosure snapped on over it. These are what read the humidity from a special hygroscopic gel/compound applied on it's face. This gel over time can lose it's properties or change causing the humidity readings to go bad. I have been unsuccessful locating a source for replacement gel but did discover certain Polyethylene Glycol gels have the right properties. Polyethylene Glycol is found in many hair gels and it's properties vary as indicated in the listed ingredients, such as PEG-8 or PEG-45, etc. It appears the number has some relation with the moisture absorption and evaporation rate. I have found gels with around PEG-8 to 12 (or lower) seem to work best. I've read a study that states the higher the number the lower the hygroscopic property.   http://www.arpc.ir/Portals/0/Pdf/Products/Chemical/PEGChemicalGrade.pdf
So if you do find you have these resistive humidity sensor chips inside, you can clean off the old gel carefully with a q-tip soaked in alcohol and apply thin layers of hair gel (with PEG), letting it dry between applications, while watching the console/display readings until you're back in business.

There are also the Sensirion SHTxx type of sensors used in many units today and measure both temp and humidity together on the chip. Since you mention only humidity readings are suffering you may have the first type I mentioned above. Good luck and post results if you do go for it.

[jerryg]

That first type is what os uses in their sensors. I guess if you could find a source for new ones they problem sell by the hundred or thousands unit. It seems they do not last very long with any kind of accuracy and the temp and transmitters seem to not fail very often that they would have them available to buy for repair instead of having to buy a new sensor everytime the humidity goes south. I will try that fix if i can find the right stuff just for fun if nothing else. Thanks for the info.

[DanS]

That first type is what os uses in their sensors. I guess if you could find a source for new ones they problem sell by the hundred or thousands unit. It seems they do not last very long with any kind of accuracy and the temp and transmitters seem to not fail very often that they would have them available to buy for repair instead of having to buy a new sensor everytime the humidity goes south. I will try that fix if i can find the right stuff just for fun if nothing else. Thanks for the info.

My older La Crosse ws-2310 has this type of humidity sensor in it and it did fine until I shipped it over here. Something in the shipping caused both the outside and the inside sensors to go bad (inside the console has the same for inside humidity measurement). Anyway after I did this fix several years ago, it's been reading correctly. I've been interested in trying Lithium Chloride with it but don't want to mess with something working so well. If/when it does go out again then I'll try it. Good luck!

[jerryg]

Is there one kind of gel that is better than another? What brand did you use?

[DanS]

Look for the ones with the lowest PEG-##'s. I settled on L'Oreal Studio Wetlook when I saw the results it gave. It has PEG-8 and works great. I also found a Paul Mitchell batch that seem to work with PEG-12 but the upper portion of the humidity scale didn't respond with readings as well. So, look for low PEG-##'s. Sometimes you may see ingredients listed like this (attachment) where PEG is listed several times and combined with other ingredients. Just pay attention to where it's listed alone or as PEG-# Polyethylene Glycol, ...  

[jerryg]

How many layers did you need to get to the right point? I got some peg8 and am doing it now.

[DanS]

I don't recall how many layers I applied but it was somewhere around 3 or 4. The important things where to have the element cleaned good first, as indicated by the lowest readings on the display/console. Apply each layer with a thin coat and watch the console numbers jump up then slowly drop as the gel dries. Once it levels off then apply another thin coat and watch then numbers again. Once you have the right amount of layers applied and the element is completely dried it should stabilize at the actual humidity level for the room. Applying more coats won't keep increasing these readings, it will still jump up and drop as is dries to the same humidity level of the room. That is, you can have not enough applied but you won't overshoot by applying too much. If you have to smear on more than 7 or 8 layers and the readings still drop down below the actual room humidity level then something's not right, probably the PEG content. You should have no problem if the gel is the right stuff. Good luck.

[jerryg]

Well i did not have any luck with the gel but i did find a replacement element on the web. I took a chance and ordered 5 of them at 2 bucks apiece with 3 bucks shipping i fiqured it would be no real loss. They are about 2 times the size of the os element and i had to trim the leads to get them to fit inside the sensor housing. I have replaced three units and have been checking the calibration on them as well as puting them outside in my stevenson screen with the Davis sensor for comparison and so far they have been working quite well. I checked at 33 and 75 percent points plus the wet towel check and it came in right on with all the check points, i think it is more accurate than the original unit. In the shield it compares real close with the Davis except of course when the humidity gets above 90% the it goes higher faster but all os sensor are very poor at the higher humidities. I am well pleased that i can repair the sensors now, in fact i just ordered some more for spares.

[sam2004gp]

Well i did not have any luck with the gel but i did find a replacement element on the web. I took a chance and ordered 5 of them at 2 bucks apiece with 3 bucks shipping i fiqured it would be no real loss. They are about 2 times the size of the os element and i had to trim the leads to get them to fit inside the sensor housing. I have replaced three units and have been checking the calibration on them as well as puting them outside in my stevenson screen with the Davis sensor for comparison and so far they have been working quite well. I checked at 33 and 75 percent points plus the wet towel check and it came in right on with all the check points, i think it is more accurate than the original unit. In the shield it compares real close with the Davis except of course when the humidity gets above 90% the it goes higher faster but all os sensor are very poor at the higher humidities. I am well pleased that i can repair the sensors now, in fact i just ordered some more for spares.

Please provide some ordering details on the elements.   

[jerryg]

Here is where i got mine from. http://iteadstudio.com/store/index.php?main_page=product_info&cPath=4&products_id=212

[jerryg]

Just a follow up on my sensor repair. First i was not aware how poor the specs are for the os sensor. It is shown to be good from 25 to 90% and accuracy is poor also. The sensor i got is from 20 to 95% accuracy of 5%. I have it installed in my stevenson screen next to my vp2 sensor for comparison. It does well in tracking the humidity below 40% and lags some as the humidity rises compared to the vp2 but seems to catch up around 50% then tracks pretty good and does not seem to run away when the humidity goes above 90% like the os one does. This am for example the new sensor showed 96% humidity while the vp2 showed 95%, pretty good for os reading lol.