Humidity Sensor Troubleshooting?

[JayMan07]

I have a thermometer/hygrometer in my reptile cage that has failed.  It is one of 3 and a really good one so I'd like to repair it if possible.  It is reading 65% when the actual humidity is 95% (I used a very accurate probe that I use for work to verify this, and it also makes sense given the environment).  I did all the simple stuff like replace the battery and I cleaned the sensor with iso-propyl.  The temperature is spot on, it's just the humidity.

Here's a pic of the sensor and I assume the square is humidity and the ball is temperature.  Can I buy that square piece somewhere and just replace it or is there a way to troubleshoot it.  It may have gotten wet but it's in perfect condition and it doesn't seem like a little water should have destroyed it.

[DrG]

That could be difficult to do without further information. I don't know, but that looks like an analog sensor that outputs either a frequency or capacitance proportional to RH, could even be a voltage out but I am guessing capacitance (and I am guessing).

Looks a bit like this one https://www.rapidonline.com/b-b-sensors-con-efs10-through-hole-electrolytic-capacitive-humidity-sensor-64-3498

Thing is, you really need to know which one it is and if it, or a drop in replacement, is available. Do you have a manual? Can you provide brand and any other information? What about the interface part?

Digital devices that do both temp and humidity are more common now days - at least in my view.

[JayMan07]

Thanks for the reply.  I will try and take a voltage reading at the sensor when I get a chance to see if it is a voltage based design.

The device is a th-100.

[JayMan07]

Also that link you posted looks spot on!  I bet that is the sensor or at least I kinda know what to search for now.  If it is capacitive then maybe it getting a little wet is what ruined it since I'm guessing if water somehow soaked into it then the capacitance value may be skewed.

[DrG]

Take a look here and see if you can identify the model a little better: https://www.sentronic.com/data/product_datasheets/KIMO-Transmitter-TH100.pdf

That seems to be a voltage out sensor - if I am reading correctly. This will take some digging, but I think you might be able to find the right one.

[DrG]

Of course if it is basically this one: https://www.reptilebasics.com/thermometers/TH-100/ I would be tempted to just spring for the $12 and be done with it

Edit: apparently many units are called a TH-100 from what I see online - weird.

[floobydust]

My guess is generic HR202L (has 3 fingers) humidity sensor.
Very cheap, order some to replace it. There is an impedance table to compare to in the datasheet.

[JayMan07]

Of course if it is basically this one: https://www.reptilebasics.com/thermometers/TH-100/ I would be tempted to just spring for the $12 and be done with it

Edit: apparently many units are called a TH-100 from what I see online - weird.

I know!  Initially I thought I'd just buy a new one but the way the wiring is run it would be much easier to replace the sensor.  This one takes a beating so if I can just replace the sensor every so often that would be ideal!  Also covid has really screwed up supply chains.  It's horrible with pet supplies.  Most websites are saying they have stuff in stock and you order it and wait a week then they tell you it's not in stock lol.

[JayMan07]

My guess is generic HR202L (has 3 fingers) humidity sensor.
Very cheap, order some to replace it. There is an impedance table to compare to in the datasheet.

Thanks floobydust!  You were the one who also helped me with that variac volt meter mod!!! 

Yes that really looks like it!  It is 5mm x 7mm like the data sheet describes.  I can't get a good resistance reading on it but maybe I need to desolder it first.  I've got another piece of the puzzle though.  So I have a known good one and I scoped it.  It updates the display every 5 seconds or so.  When it does that this is what I see...



I need to scope the broken one to see if the signal is there, but I think it probably is because it's reading humidity, it's just off.  The broken on is really hard to get to but I could maybe use my portable scope that I use for work to take a look at it.

[JayMan07]

Here's what the timing of the signal looks like.  It's a pulse every 9 seconds or so...

[JayMan07]

Here's one more clue I think to what is going on.  When I scope the sensor the humidity goes up like 5% so maybe that's evidence for it being resistive.  If the impedance of the scope is causing the controller to see somewhat of a lower resistance than it would increase the humidity according to the datasheet.

[floobydust]

These kind of sensors are used with a MCU generating an RC ramp and measuring time, instead of a real A/D converter. So a little bit of work to go backwards from the scope traces.
Another trick is to put in a resistor (or a potentiometer) in place of the sensor, the display should match the HR202L curves.

[JayMan07]

These kind of sensors are used with a MCU generating an RC ramp and measuring time, instead of a real A/D converter. So a little bit of work to go backwards from the scope traces.
Another trick is to put in a resistor (or a potentiometer) in place of the sensor, the display should match the HR202L curves.

Awesome!!! So I put in my variable resistor dialed into 88.5K at 22.7C it reads 53% so it's following the table exactly.  So that's the sensor I need.  Thanks I really appreciate it once again!  I think as an experiment I'll just swap sensors with the good one maybe to get me by until I can get some sensors ordered.  I'll update this post with the results once I swap the sensors.

[JayMan07]

Cut out the bad sensor and open circuit it reads 10% humidity just like the other one so that was a good sign.  Soldered in the good sensor and everything is working well.  Got some on order from China so that'll probably take like 3 months but at least I'll have them.

[DrG]

Cut out the bad sensor and open circuit it reads 10% humidity just like the other one so that was a good sign.  Soldered in the good sensor and everything is working well.  Got some on order from China so that'll probably take like 3 months but at least I'll have them.

Good deal! 

I have never used a resistive one like that and did not even know they were around. Aosong makes the DHT11 and 22 and AHT-10, which I have used but never an HR202L so thanks @floobydust .

With the digital sensors that I have used - if they are in a very high humidity for a long time, they will get saturated and they can be baked as per the datasheet. I have an outdoor one (Sil7020) which got wet, thanks to my poor packaging, and became completely unreliable regardless of baking, but the temp still works).

[CatalinaWOW]

Cut out the bad sensor and open circuit it reads 10% humidity just like the other one so that was a good sign.  Soldered in the good sensor and everything is working well.  Got some on order from China so that'll probably take like 3 months but at least I'll have them.

Good deal! 

I have never used a resistive one like that and did not even know they were around. Aosong makes the DHT11 and 22 and AHT-10, which I have used but never an HR202L so thanks @floobydust .

With the digital sensors that I have used - if they are in a very high humidity for a long time, they will get saturated and they can be baked as per the datasheet. I have an outdoor one (Sil7020) which got wet, thanks to my poor packaging, and became completely unreliable regardless of baking, but the temp still works).

All humidity sensors I have looked at can be ruined by too much moisture, particularly moisture which might have extra ingredients like one in a reptile cage.  Most warn against cleaning with solvents.  I have had mixed results baking them out and never had success on one which was contaminated with anything but water.

I think your plan to have spares ready is sound.  Store them in the driest place you can.

[JayMan07]

Hmmm that's interesting!  I didn't know that baking them was an option.  Do you happen to remember the temp and time used to do that?

And yeah I have a feeling that I'll be seeing a lot of failed sensors.  The location of this one stays at 99% so the sensor is taking tons of abuse, but it's necessary for it to be in that location because when the humidity drops it lets me know the substrate is going dry.

[DrG]

Hmmm that's interesting!  I didn't know that baking them was an option.  Do you happen to remember the temp and time used to do that?

And yeah I have a feeling that I'll be seeing a lot of failed sensors.  The location of this one stays at 99% so the sensor is taking tons of abuse, but it's necessary for it to be in that location because when the humidity drops it lets me know the substrate is going dry.

To answer directly, in the case of the Sil7020 https://www.silabs.com/documents/public/data-sheets/Si7020-A20.pdf that I mentioned, see:

4.6.  Bake/Hydrate Procedure
After  exposure  to  extremes  of  temperature  and/or  humidity  for  prolonged  periods,  the  polymer  sensor  film  can become either very dry or very wet, in each case the result is either high or low relative humidity readings. Under normal  operating  conditions,  the  induced  error  will  diminish  over  time.  From  a  very  dry  condition,  such  as  after shipment  and  soldering,  the  error  will  diminish  over  a  few  days  at  typical  controlled  ambient  conditions,  e.g.,48 hours  of  45≤%RH≤55.  However,  from  a  very  wet  condition,  recovery  may  take  significantly  longer.  To accelerate recovery from a wet condition, a bake and hydrate cycle can be implemented. This operation consists of the following steps:

Baking the sensor at 125 °C for ≥ 12 hours
Hydration at 30 °C in 75% RH for ≥ 10 hours

BUT, I would definitely not assume that the one you have will respond in the same way. Do that at your own risk

Note also, that re-hydration is recommended when it is used/stored in a very dry environment for s long time.

Finally, if you have not done so already, sear on checking the accuracy of humidity sensors using the salts method. It is not too hard to do and you can check them before they are completely out of whack.

What kind of reptiles or you keeping?

[JayMan07]

Wow 125C for 12hrs!  I guess I'll be replacing them since the electricity to bake it is more than the sensor hahaha.  But I think I will try some experiments of maybe higher heat for shorter time.  I've got 20 sensors on the way so as far as I'm concerned I couldn't care less if I destroy them.

I've got two leopard geckos.  They don't need a high humidity vivarium.  But they like to have a humid hide.  It really helps them shed and they actually really like it.  They split their time between one that is 90F and 40% humidity and the humid hide is 90F and 99% humidity.  99% is a little on the high side but there's really no way to maintain anything else since the hide is either wet or dry.  It's not really practical to keep it at something in between that.  The humid hide is enclosed and at 90F the moss dries out relatively quickly so the humidity sensor helps me keep an eye on that.

[Raj]

my dht22 has been used for literal years, still reads the same as my analogue (hair and needle) meter which I calibrate regularly using nearby university reports ( they have a big display showing the same at their entrance, as a novelty.)
The thing about sensors is, the problem might not always be where you think it is. What if your reference is incorrect but the sensor is OK?
Even a little bit of organic material ingressing into your probe can make our sensor laggy or make it read higher. (actually had this problem when I was using a wooden housing for my sensor and every time it rained, the sensor read 99% for a whole day, while the one without housing dropped to 85% immediately.
it got as responsive as it can be when I switched to using plastic enclosure.