Author Topic: The Reliability of the SHT-31 Humidity Sensor & What Psychrometer Should I Buy?  (Read 107060 times)

0 Members and 4 Guests are viewing this topic.

Offline CW2274

  • Forecaster
  • *****
  • Posts: 6731
    • Conditions @ CW2274 West Tucson-Painted Hills Ranch
Understood. It’s an open forum and I’m replying to you.
Fine. My response was not meant for your issues as I have zero experience with either the 75 or 85. My "fiasco" comment was strictly meant in regards to the 31 issue, which I have plenty of experience with.

Offline jgentry

  • Forecaster
  • *****
  • Posts: 533
Is there any way to make adjustments to the PCB of the Davis 31 temp/hum sensor that will allow maximum airflow to the sensor without damaging the sensor? My thinking is that if you bake the sensor to put it back into calibration, along with the sensor getting maximum airflow, it will fix any prolong wet bias issues of the sensor along with it struggling to reach 100%.
Davis Vantage Pro2 & WeatherFlow Tempest. WU: KXALJEMI2, KALJEMIS7, KFLPANAM363 & KALTHORS2. CWOP/APRS: C6353 & E6358

Offline jgentry

  • Forecaster
  • *****
  • Posts: 533
Just to add to the mix i have an old 31 that i had out for some time and it showed the bias of 3% high which is keeping with the specs of the sensor which is the sensor will read high after being in a high humidity setting of more than 80% for three days are more and will only return to the normal reading after being below the 80% for some time as it says the return will be very slow. I put the new 31 out and it has been very good with the low and high end readings with a lot of 100% readings in heavy fog. I put the 31 back out in a different shield and sure enough after a long period of drying out it read 1% lower than the new 31 and has been holding that for several weeks. We haven't got into a run of high humidity yet so won't know what happens until i get a good run of above 80% to set the stage for the error to show up. Now i think the baking works well because of the very low humidity in the oven and makes for a faster recovery time but just keeping it below 80% for sometime works the same just slower. This time of year i can get periods of above 80% for a week at a time so i will just have to wait for that to happen.

I’m curious if the 31 was mounted correctly, it would return back into calibration quicker than the Davis’ setup?
If you recall, the Acurite device that utilized the 31 had no bias compared to the Davis mounted 31. I kinda thought we were done with this fiasco. Yes, I know, I'm in the desert, but my current 31 is 18 months old and I remain very happy with it. I still think a sealed SF2 cap is a difference maker in the big scheme of things.

I’m sure the cap helps. But I think the big issue is that the sensor isn’t getting maximum airflow, which causes moisture to collect for long periods of time and permanently offset the sensor. WeatherFlow Tempest doesn’t use the Sensirion cap for the 31 and humidity sensor is exposed. No prolong wet bias issues with the 31 for them either.

But having the Sensirion filter instead of the Davis one is a major plus.
Davis Vantage Pro2 & WeatherFlow Tempest. WU: KXALJEMI2, KALJEMIS7, KFLPANAM363 & KALTHORS2. CWOP/APRS: C6353 & E6358

Offline CW2274

  • Forecaster
  • *****
  • Posts: 6731
    • Conditions @ CW2274 West Tucson-Painted Hills Ranch
Just to add to the mix i have an old 31 that i had out for some time and it showed the bias of 3% high which is keeping with the specs of the sensor which is the sensor will read high after being in a high humidity setting of more than 80% for three days are more and will only return to the normal reading after being below the 80% for some time as it says the return will be very slow. I put the new 31 out and it has been very good with the low and high end readings with a lot of 100% readings in heavy fog. I put the 31 back out in a different shield and sure enough after a long period of drying out it read 1% lower than the new 31 and has been holding that for several weeks. We haven't got into a run of high humidity yet so won't know what happens until i get a good run of above 80% to set the stage for the error to show up. Now i think the baking works well because of the very low humidity in the oven and makes for a faster recovery time but just keeping it below 80% for sometime works the same just slower. This time of year i can get periods of above 80% for a week at a time so i will just have to wait for that to happen.

I’m curious if the 31 was mounted correctly, it would return back into calibration quicker than the Davis’ setup?
If you recall, the Acurite device that utilized the 31 had no bias compared to the Davis mounted 31. I kinda thought we were done with this fiasco. Yes, I know, I'm in the desert, but my current 31 is 18 months old and I remain very happy with it. I still think a sealed SF2 cap is a difference maker in the big scheme of things.

I’m sure the cap helps. But I think the big issue is that the sensor isn’t getting maximum airflow, which causes moisture to collect for long periods of time and permanently offset the sensor.
That may or may not be true. However, with an SF2 cap that is properly sealed, it is a water tight solution that only allows water vapor to penetrate. That an a case fan should alleviate any potential moisture build up over ambient AFAIC. One thing for certain, I'll never go without an SF2 cap again.

Offline Jester

  • Pony Up!
  • Senior Contributor
  • ****
  • Posts: 171
  • US NAVY Retired
    • Castleton Weather, Virginia Beach, VA.
CW, how did you modify your SF2 cap to work on the SHT31 Davis PCB?

MADIS D4557
CWOP DW4557
Wunderground KVAVIRGI57
CoCoRaHS VA-VBC-13
Davis VP2+ 6163
WiFi LoggerV2


Offline johnd

  • Forecaster
  • *****
  • Posts: 4826
    • www.weatherstations.co.uk
Not directly relevant, but I see that the Sensirion SHT4x series are starting to becoming available:

https://www.sensirion.com/en/environmental-sensors/humidity-sensors/humidity-sensor-sht4x/

Just SHT40 so far, but 41 and 45 apparently due later this year. Adafruit have stock of the SHT40 (true) I2C version in a STEMMA board:

https://www.adafruit.com/product/4885

Absolutely no word or hint from Davis about any changes, so this is just for general background info.
Prodata Weather Systems
Prodata's FAQ/support site for Davis stations
Includes many details on 6313 Weatherlink console.
UK Davis Premier Dealer - All Davis stations, accessories and spares
Cambridge UK

Sorry, but I don't usually have time to help with individual issues by email unless you are a Prodata customer. Please post your issue in the relevant forum section here & I will comment there if I have anything useful to add.

Offline Jester

  • Pony Up!
  • Senior Contributor
  • ****
  • Posts: 171
  • US NAVY Retired
    • Castleton Weather, Virginia Beach, VA.
Not directly relevant, but I see that the Sensirion SHT4x series are starting to becoming available:

https://www.sensirion.com/en/environmental-sensors/humidity-sensors/humidity-sensor-sht4x/

Just SHT40 so far, but 41 and 45 apparently due later this year. Adafruit have stock of the SHT40 (true) I2C version in a STEMMA board:

https://www.adafruit.com/product/4885

Absolutely no word or hint from Davis about any changes, so this is just for general background info.


Interesting.  I kind of like this idea https://www.adafruit.com/product/4099 instead of the cap filter. What is the VCC operating range of the Davis SHT31?
MADIS D4557
CWOP DW4557
Wunderground KVAVIRGI57
CoCoRaHS VA-VBC-13
Davis VP2+ 6163
WiFi LoggerV2


Offline galfert

  • Global Moderator
  • Forecaster
  • *****
  • Posts: 6822
You'll likely never be able to run anything other than the SHT31 with Davis VP2. Unless Davis comes out with a new ISS. This is because Sensirion has stopped making sensor variant with the Legacy Sensirion Sensibus protocol. The SHT31 was the last sensor to be made with the required LSS version.

Ecowitt GW1000 | Meteobridge on Raspberry Pi
WU: KFLWINTE111  |  PWSweather: KFLWINTE111
CWOP: FW3708  |  AWEKAS: 14814
Windy: pws-f075acbe
Weather Underground Issue Tracking
Tele-Pole

Offline johnd

  • Forecaster
  • *****
  • Posts: 4826
    • www.weatherstations.co.uk
You'll likely never be able to run anything other than the SHT31 with Davis VP2. Unless Davis comes out with a new ISS.

100%. But, and this is just an errant thought: If someone wanted to do so badly enough and assuming that the LSS protocol details are in the public domain, then it wouldn't it be reasonably easy for a suitably skilled person to write some code for a little microcontroller that sat in between an SHT35 or 4x and translated the true I2C data  into LSS to respond to the SIM boards requests for T/H data every 10 secs?
Prodata Weather Systems
Prodata's FAQ/support site for Davis stations
Includes many details on 6313 Weatherlink console.
UK Davis Premier Dealer - All Davis stations, accessories and spares
Cambridge UK

Sorry, but I don't usually have time to help with individual issues by email unless you are a Prodata customer. Please post your issue in the relevant forum section here & I will comment there if I have anything useful to add.

Offline galfert

  • Global Moderator
  • Forecaster
  • *****
  • Posts: 6822
Yes in theory reading in an I2C sensor and then replacing the protocol with LSS sounds interesting. This has been mentioned before but it never saw results.
https://www.wxforum.net/index.php?topic=33157.msg355110#msg355110

It doesn't mean someone else can't succeed.
Ecowitt GW1000 | Meteobridge on Raspberry Pi
WU: KFLWINTE111  |  PWSweather: KFLWINTE111
CWOP: FW3708  |  AWEKAS: 14814
Windy: pws-f075acbe
Weather Underground Issue Tracking
Tele-Pole

Offline CW2274

  • Forecaster
  • *****
  • Posts: 6731
    • Conditions @ CW2274 West Tucson-Painted Hills Ranch
I'm almost positive that a member here got a 35 to work with their VP2 ISS. Obviously he had a very knowledgeable background, and I believe some of us said we would pay him for his service, but nothing ever came of it. That was like four years ago or so.

Offline CW2274

  • Forecaster
  • *****
  • Posts: 6731
    • Conditions @ CW2274 West Tucson-Painted Hills Ranch
CW, how did you modify your SF2 cap to work on the SHT31 Davis PCB?
I wasn't thrilled about the glue option so I trimmed the four feet down to be flush with the base and used very thin stainless steel wire to wrap around the cap and cinch it down with pliers twisting from the back side. The wire doesn't come in contact with the membrane but arches over. With the "gooey" substance that surrounds the sensor it allows the cap to seat firmly in what I believe is a true seal.

Offline jgentry

  • Forecaster
  • *****
  • Posts: 533
Not directly relevant, but I see that the Sensirion SHT4x series are starting to becoming available:

https://www.sensirion.com/en/environmental-sensors/humidity-sensors/humidity-sensor-sht4x/

Just SHT40 so far, but 41 and 45 apparently due later this year. Adafruit have stock of the SHT40 (true) I2C version in a STEMMA board:

https://www.adafruit.com/product/4885

Absolutely no word or hint from Davis about any changes, so this is just for general background info.

My biggest pet peeve about the Sensirion sensor (and I’ve heard others do the same) is that after being exposed to humidities over 80% for a long period of time, it goes out of calibration until it recalibrates itself.
Davis Vantage Pro2 & WeatherFlow Tempest. WU: KXALJEMI2, KALJEMIS7, KFLPANAM363 & KALTHORS2. CWOP/APRS: C6353 & E6358

Offline CW2274

  • Forecaster
  • *****
  • Posts: 6731
    • Conditions @ CW2274 West Tucson-Painted Hills Ranch
Not directly relevant, but I see that the Sensirion SHT4x series are starting to becoming available:

https://www.sensirion.com/en/environmental-sensors/humidity-sensors/humidity-sensor-sht4x/

Just SHT40 so far, but 41 and 45 apparently due later this year. Adafruit have stock of the SHT40 (true) I2C version in a STEMMA board:

https://www.adafruit.com/product/4885

Absolutely no word or hint from Davis about any changes, so this is just for general background info.

My biggest pet peeve about the Sensirion sensor (and I’ve heard others do the same) is that after being exposed to humidities over 80% for a long period of time, it goes out of calibration until it recalibrates itself.
Is it high humidity or high dew point? If it's 0F outside and a dew point of -1F, that's 98ish% humidity, but bone dry air. I'm thinking high dew is the culprit, it's not fond of being "wet".

Offline CW7491

  • Senior Contributor
  • ****
  • Posts: 202
The datasheet specifies humidity. Sensirion’s documents say that prolonged exposure to high humidity will induce the reversible high bias which has been discussed ad nauseam here, but it also mentions extreme low humidity for prolonged periods will introduce a low bias, which hasn’t been discussed because most of us probably don’t experience it. Finally, the sensor outputs relative humidity with reference to saturated water vapor pressure over liquid water even for temperatures below freezing. The result is as temperatures get colder and colder below 0C, maximum humidity output by the sensor decreases. So at 0F (-18C) in your example, maximum humidity output from the sensor would be about 85% and that reading would indicate fully saturated air.

Offline CW2274

  • Forecaster
  • *****
  • Posts: 6731
    • Conditions @ CW2274 West Tucson-Painted Hills Ranch
it also mentions extreme low humidity for prolonged periods will introduce a low bias, which hasn’t been discussed because most of us probably don’t experience it.
Good post, I didn't know this. Obviously I do experience regular extreme low humidity, and sure enough, when I get to the 4% or lower range it starts to get a little over the top with tanking dew points, especially down at 1%, but even I don't get that but a few days a year.

Offline Jester

  • Pony Up!
  • Senior Contributor
  • ****
  • Posts: 171
  • US NAVY Retired
    • Castleton Weather, Virginia Beach, VA.
CW, how did you modify your SF2 cap to work on the SHT31 Davis PCB?
I wasn't thrilled about the glue option so I trimmed the four feet down to be flush with the base and used very thin stainless steel wire to wrap around the cap and cinch it down with pliers twisting from the back side. The wire doesn't come in contact with the membrane but arches over. With the "gooey" substance that surrounds the sensor it allows the cap to seat firmly in what I believe is a true seal.

Thanks!
MADIS D4557
CWOP DW4557
Wunderground KVAVIRGI57
CoCoRaHS VA-VBC-13
Davis VP2+ 6163
WiFi LoggerV2


Offline jgentry

  • Forecaster
  • *****
  • Posts: 533
I think the wax stuff that Davis uses is the main culprit of their humidity sensor woes. The wax holds moisture and keeps the sensor from fully “drying” out.
Davis Vantage Pro2 & WeatherFlow Tempest. WU: KXALJEMI2, KALJEMIS7, KFLPANAM363 & KALTHORS2. CWOP/APRS: C6353 & E6358

Offline CW7491

  • Senior Contributor
  • ****
  • Posts: 202
One of the things I noticed in the SHT4X datasheet is an onboard heater. This is incorporated into the SHT3X as well, but the SHT4X datasheet specifically mentions one of the functions of the heater is for condensing environments to eliminate creep, which is the name they give to the high bias we see and they mention in prolonged high humidity. They also mention that there will be dedicated application notes elaborating on the use of the heater for this purpose.

It would be interesting if this function could be utilized so the Sensirion sensor was used to calculate and output dew point and then incorporate a separate temperature measuring element (thermally isolated from any self-heating), maybe a standard thermistor, for air temperature. The console could then calculate RH based on the two temperatures instead of calculating dew point. I posted a bit ago about a new version .80 temp/hum sensor Rainwise uses for their MKIII that I was speculating might actually do this as it appears to have both a thermistor and a Sensirion sensor on their temp/hum board
« Last Edit: February 21, 2021, 04:34:53 PM by CW7491 »

Offline jgentry

  • Forecaster
  • *****
  • Posts: 533
One of the things I noticed in the SHT4X datasheet is an onboard heater. This is incorporated into the SHT3X as well, but the SHT4X datasheet specifically mentions one of the functions of the heater is for condensing environments to eliminate creep, which is the name they give to the high bias we see and they mention in prolonged high humidity. They also mention that there will be dedicated application notes elaborating on the use of the heater for this purpose.

It would be interesting if this function could be utilized so the Sensirion sensor was used to calculate and output dew point and then incorporate a separate temperature measuring element (thermally isolated from any self-heating), maybe a standard thermistor, for air temperature. The console could then calculate RH based on the two temperatures instead of calculating dew point. I posted a bit ago about a new version .80 temp/hum sensor Rainwise uses for their MKIII that I was speculating might actually do this as it appears to have both a thermistor and a Sensirion sensor on their temp/hum board

That might be the solution to the problem. Sensirion isn’t the only ones that have the issue. Bosch deals with a similar issue also.
Davis Vantage Pro2 & WeatherFlow Tempest. WU: KXALJEMI2, KALJEMIS7, KFLPANAM363 & KALTHORS2. CWOP/APRS: C6353 & E6358

Offline CW2274

  • Forecaster
  • *****
  • Posts: 6731
    • Conditions @ CW2274 West Tucson-Painted Hills Ranch
The wax holds moisture and keeps the sensor from fully “drying” out.
I really can't see that. If anything, wax is hydrophobic.  That's why you put it on your car's finish for example.

Offline jgentry

  • Forecaster
  • *****
  • Posts: 533
The wax holds moisture and keeps the sensor from fully “drying” out.
I really can't see that. If anything, wax is hydrophobic.  That's why you put it on your car's finish for example.

It’s my guess and I could be totally wrong.
Davis Vantage Pro2 & WeatherFlow Tempest. WU: KXALJEMI2, KALJEMIS7, KFLPANAM363 & KALTHORS2. CWOP/APRS: C6353 & E6358

Offline galfert

  • Global Moderator
  • Forecaster
  • *****
  • Posts: 6822
Wax has both hydrophobic and hydrophilic properties. It is a little complicated. It is sort of like soap where it starts off being hydrophobic but then as it gets worn it becomes hydrophilic.
Ecowitt GW1000 | Meteobridge on Raspberry Pi
WU: KFLWINTE111  |  PWSweather: KFLWINTE111
CWOP: FW3708  |  AWEKAS: 14814
Windy: pws-f075acbe
Weather Underground Issue Tracking
Tele-Pole

Offline CW2274

  • Forecaster
  • *****
  • Posts: 6731
    • Conditions @ CW2274 West Tucson-Painted Hills Ranch
Don't even know for a fact what the substance is, but it's certainly gummy. Well, just another reason to pimp the SF2. It's just big enough to cover the sensor itself.

Offline CW2274

  • Forecaster
  • *****
  • Posts: 6731
    • Conditions @ CW2274 West Tucson-Painted Hills Ranch
After playing with my spare 31, I think the substance isn't wax at all, but some form of rubber.

 

anything