Author Topic: The Reliability of the SHT-31 Humidity Sensor & What Psychrometer Should I Buy?  (Read 30852 times)

0 Members and 1 Guest are viewing this topic.

Offline hwcorder

  • Contract Weather Observer at KCLT
  • Senior Member
  • **
  • Posts: 73
Well the wiring in my 75 seems to have developed a short somewhere so i switched it out with a new 31 I have had lying around in a stock Davis filter.  Switched it out yesterday and was surprised to see it reach 100% last night with ease.  We did have a heavy shower come through yesterday around sunset that put down about 1/3 of an inch followed by partly cloudy skies last night with some light fog this morning.  I guess I will leave it in there for a few days to see how it tracks along with surrounding ASOS stations and to see if a substantial wet bias develops.  Still using the MeteoShield Pro for those of you wondering.

Offline jgentry

  • Forecaster
  • *****
  • Posts: 346
Well the wiring in my 75 seems to have developed a short somewhere so i switched it out with a new 31 I have had lying around in a stock Davis filter.  Switched it out yesterday and was surprised to see it reach 100% last night with ease.  We did have a heavy shower come through yesterday around sunset that put down about 1/3 of an inch followed by partly cloudy skies last night with some light fog this morning.  I guess I will leave it in there for a few days to see how it tracks along with surrounding ASOS stations and to see if a substantial wet bias develops.  Still using the MeteoShield Pro for those of you wondering.

Nice. The 31 I have in one of my field stations reaches 100% fairly easily. It does good with humidities in the 90 percentile. It’s just when it goes below 75% is when the wet bias really begins to show.
Davis Vantage Pro2. SHT-75. WU: KXALJEMI2 & KALTHORS2. CWOP/APRS: C6353 & E6358

  

Offline dendrite

  • Forecaster
  • *****
  • Posts: 360
Well the wiring in my 75 seems to have developed a short somewhere so i switched it out with a new 31 I have had lying around in a stock Davis filter.  Switched it out yesterday and was surprised to see it reach 100% last night with ease.  We did have a heavy shower come through yesterday around sunset that put down about 1/3 of an inch followed by partly cloudy skies last night with some light fog this morning.  I guess I will leave it in there for a few days to see how it tracks along with surrounding ASOS stations and to see if a substantial wet bias develops.  Still using the MeteoShield Pro for those of you wondering.
I had problems with moisture on the PCB with my first SHT75 until I put the silicone conformal coating on it. Now only the sensor and tip of the PCB that the sensor sits on is exposed.

I've been sitting at 97-98% with 1.60" of rain in 24hrs and mine is still running fine.

Offline openvista

  • Forecaster
  • *****
  • Posts: 401
    • marquetteweather.com
Looks like Onset might have figured out how to protect the Sensirion sensors from excessive moisture intake.

Check out this doc: http://www.onsetcomp.com/files/manual_pdfs/11427-N%20MAN-S-THB.pdf

It talks about the housing material over the sensor. It's quite the "cap" they devised.

I'm not sure which Sensirion model they use, but they do mention the manufacturer in at least one of their other docs detailing testing procedures (http://www.onsetcomp.com/files/15932-B-Testing-Sensirion-RH-Sensors-White-Paper.pdf).

Any Onset HOBO owners reading this thread who have tested their humidity/dew point? Is there a pronounced bias in the middle (40-70%)?
« Last Edit: August 12, 2018, 11:05:39 AM by openvista »
Davis Vantage Pro2 FARS | https://marquetteweather.com | EW7933

Offline hwcorder

  • Contract Weather Observer at KCLT
  • Senior Member
  • **
  • Posts: 73
Well the wiring in my 75 seems to have developed a short somewhere so i switched it out with a new 31 I have had lying around in a stock Davis filter.  Switched it out yesterday and was surprised to see it reach 100% last night with ease.  We did have a heavy shower come through yesterday around sunset that put down about 1/3 of an inch followed by partly cloudy skies last night with some light fog this morning.  I guess I will leave it in there for a few days to see how it tracks along with surrounding ASOS stations and to see if a substantial wet bias develops.  Still using the MeteoShield Pro for those of you wondering.
I had problems with moisture on the PCB with my first SHT75 until I put the silicone conformal coating on it. Now only the sensor and tip of the PCB that the sensor sits on is exposed.

I've been sitting at 97-98% with 1.60" of rain in 24hrs and mine is still running fine.

Found the short in the white wire. Looks like it got pulled away from the potting epoxy I used somehow. Probably since I have been messing with it pulling it in and out of different shields.  Had to order some more 4 pin connectors from mouser so I can resolder.  Like others I had little luck with the crimp style connectors. Kept getting an inconsistent connection. 

Speaking of soldering, I found these pin connectors on Mouser made by Mill Max.  They come with solder cups where it looks as though you can insert the wire then solder it in.  If it is as it looks this would be much easier for some of you who have bad luck trying to solder the 1.27mm tabs. I have left the link below.
https://www.mouser.com/ProductDetail/Mill-Max/851-13-004-10-003000?qs=sGAEpiMZZMs%252bGHln7q6pm8Vn94ktop%2fJsOvP78LpBImemlioZUsStg%3d%3d

Offline ValentineWeather

  • Forecaster
  • *****
  • Posts: 5443
Did you see what Jerry and I ordered?  I have 2 on the way may not be here until fall however. https://www.aliexpress.com/item/2pcs-Waterproof-Temperature-and-humidity-sensor-shell-SHT10-SHT21-SHT15-shT11-sht20-SHT75-sensor-protective-sleeve/32725766858.html?spm=a2g0s.9042311.0.0.69af4c4dlQrpxK

Others have some like https://www.emesystems.com/sht75pg/documents/SHT75PG.pdf

dendrite are there no worries about contamination using silicone conformal coating?
Randy

Offline ValentineWeather

  • Forecaster
  • *****
  • Posts: 5443


Speaking of soldering, I found these pin connectors on Mouser made by Mill Max.  They come with solder cups where it looks as though you can insert the wire then solder it in.  If it is as it looks this would be much easier for some of you who have bad luck trying to solder the 1.27mm tabs. I have left the link below.
https://www.mouser.com/ProductDetail/Mill-Max/851-13-004-10-003000?qs=sGAEpiMZZMs%252bGHln7q6pm8Vn94ktop%2fJsOvP78LpBImemlioZUsStg%3d%3d

I'll try these. So stick wires in solder then the SHT75 plus in, is that the idea?

Found this on how to solder with them.

https://www.eptac.com/ask/what-is-the-correct-way-to-fill-solder-cups/
« Last Edit: August 12, 2018, 11:27:53 AM by ValentineWeather »
Randy

Offline hwcorder

  • Contract Weather Observer at KCLT
  • Senior Member
  • **
  • Posts: 73


Speaking of soldering, I found these pin connectors on Mouser made by Mill Max.  They come with solder cups where it looks as though you can insert the wire then solder it in.  If it is as it looks this would be much easier for some of you who have bad luck trying to solder the 1.27mm tabs. I have left the link below.
https://www.mouser.com/ProductDetail/Mill-Max/851-13-004-10-003000?qs=sGAEpiMZZMs%252bGHln7q6pm8Vn94ktop%2fJsOvP78LpBImemlioZUsStg%3d%3d

I'll try these. So stick wires in solder then the SHT75 plus in, is that the idea?

Seems so. I would think these will be a lot easier to keep the spacing of the wires while being able to get a good solder on them.  You can do whatever weatherproofing you desire, give it time to cure then insert the 75 so there is virtually no risk of chemical contamination.

Offline openvista

  • Forecaster
  • *****
  • Posts: 401
    • marquetteweather.com
Maybe Onset hasn't figured it out? Because a quick check of one of their stations in Bryn Mawr (https://www.wunderground.com/personal-weather-station/dashboard?ID=KPAHAVER17#history) shows a dew point of 72/73 when all the surrounding airports are reporting 70.

The more I observe other Sensirion sensors in other branded weather stations, the more I realize the problem isn't limited to Davis. I also see the problem in Atlas Elites and RainWise stations.

It probably goes back to what Ron (kcidwx) said about Sensirion sensors being designed for indoor use and HVAC primarily and not as meteorological instruments.

Doesn't seem like the cap matters, at least not for 11/15/31 sensors. Seems it's more about how often they've been in saturated air.
« Last Edit: August 12, 2018, 11:46:25 AM by openvista »
Davis Vantage Pro2 FARS | https://marquetteweather.com | EW7933

Offline ValentineWeather

  • Forecaster
  • *****
  • Posts: 5443
How are these sht75's working if interface is I2c? Thought that wouldn't work.

https://www.mouser.com/ProductDetail/Sensirion/SHT75?qs=%2fha2pyFaduhV3NYEe1MP5y%2fYDjvd38ndllSCUE15w5k%3d
Randy

Offline hwcorder

  • Contract Weather Observer at KCLT
  • Senior Member
  • **
  • Posts: 73
How are these sht75's working if interface is I2c? Thought that wouldn't work.

https://www.mouser.com/ProductDetail/Sensirion/SHT75?qs=%2fha2pyFaduhV3NYEe1MP5y%2fYDjvd38ndllSCUE15w5k%3d

Unless there is an I2C version not advertised by Senserion of the 75 then this must be a typo.  I don't order my 75s from Mouser but from Newark 14. Main reason being they have a distribution warehouse in South Carolina so I get them within a day or two even with UPS gound.

Offline openvista

  • Forecaster
  • *****
  • Posts: 401
    • marquetteweather.com
How are these sht75's working if interface is I2c? Thought that wouldn't work.

https://www.mouser.com/ProductDetail/Sensirion/SHT75?qs=%2fha2pyFaduhV3NYEe1MP5y%2fYDjvd38ndllSCUE15w5k%3d

It's a typo. I ordered my 75 from them and had it working briefly before I broke it.
Davis Vantage Pro2 FARS | https://marquetteweather.com | EW7933

Offline openvista

  • Forecaster
  • *****
  • Posts: 401
    • marquetteweather.com
OK, after looking at a bunch of Rainwise stations, I should back up and say I only occasionally see the wet bias problem.

Whereas with Davis VP2s, well, it's uncommon to find one that's accurate in the middle range of humidity (40-70%).

Also, to be fair, the Acurite Atlases listed on Wunderground may very well be other stations at this point since the "beta" test period ended, I believe, and the product has yet to be released.


Davis Vantage Pro2 FARS | https://marquetteweather.com | EW7933

Offline CW7491

  • Member
  • *
  • Posts: 23
OK, after looking at a bunch of Rainwise stations, I should back up and say I only occasionally see the wet bias problem.

Whereas with Davis VP2s, well, it's uncommon to find one that's accurate in the middle range of humidity (40-70%).

Also, to be fair, the Acurite Atlases listed on Wunderground may very well be other stations at this point since the "beta" test period ended, I believe, and the product has yet to be released.

Do we know for sure that Rainwise uses Sensirion? Maybe just for humidity? I noticed their specs for temp go down to -67F and I have tried to research online what they use for temp and humidity, but came up without much info.

Offline openvista

  • Forecaster
  • *****
  • Posts: 401
    • marquetteweather.com
The Rainwise III uses the SHT15. I don't think they use separate temp & humidity sensors, although someone correct me if I'm wrong.
Davis Vantage Pro2 FARS | https://marquetteweather.com | EW7933

Offline CW7491

  • Member
  • *
  • Posts: 23
The Rainwise III uses the SHT15. I don't think they use separate temp & humidity sensors, although someone correct me if I'm wrong.

Good to know and thanks for the info. I've been trying to figure that one out for a while.

Offline ValentineWeather

  • Forecaster
  • *****
  • Posts: 5443
OK, after looking at a bunch of Rainwise stations, I should back up and say I only occasionally see the wet bias problem.

Whereas with Davis VP2s, well, it's uncommon to find one that's accurate in the middle range of humidity (40-70%).

Also, to be fair, the Acurite Atlases listed on Wunderground may very well be other stations at this point since the "beta" test period ended, I believe, and the product has yet to be released.

Do we know for sure that Rainwise uses Sensirion? Maybe just for humidity? I noticed their specs for temp go down to -67F and I have tried to research online what they use for temp and humidity, but came up without much info.

It looks like they may have changed sensor at sometime. The RWIII now says -40 to 60c range and use to be different like -68F. There was a forum member who lived in Alaska that would come here and brag about how his station would still work at those cold temperatures.   https://www.rainwise.com/products/attachments/6801/20150710121434.pdf
Randy

Offline Old Tele man

  • Singing in the rain...
  • Forecaster
  • *****
  • Posts: 1343
Did you see what Jerry and I ordered?  I have 2 on the way may not be here until fall however. https://www.aliexpress.com/item/2pcs-Waterproof-Temperature-and-humidity-sensor-shell-SHT10-SHT21-SHT15-shT11-sht20-SHT75-sensor-protective-sleeve/32725766858.html?spm=a2g0s.9042311.0.0.69af4c4dlQrpxK

Others have some like https://www.emesystems.com/sht75pg/documents/SHT75PG.pdf

dendrite are there no worries about contamination using silicone conformal coating?

Anybody else notice that the 'dewpoint accuracy' graph at the bottom of page 4 [ https://www.emesystems.com/sht75pg/documents/SHT75PG.pdf ] shows a VERY interesting anomaly: the dp 'error' suddenly kinks UPWARD at and above 90%RH.
« Last Edit: August 12, 2018, 02:07:57 PM by Old Tele man »
• SYS: Davis VP2 Vue/WL-IP & Envoy8X/WL-USB;
• DBX1 Precision Digital Barograph
• CWOP: DW6988 - 2 miles NNE of Cortaro, AZ
• WU - KAZTUCSO202, Countryside

Offline ValentineWeather

  • Forecaster
  • *****
  • Posts: 5443
Yeah that's the correct 75 spec graph. Why everyone was so high on the 31's being flatline 2% originally (now it changed some with updated specs) and we've found out Davis version is nowhere near specs  +/- 8% is more inline and why others are going back to the SHT75 that hasn't been destroyed with the waterproofing process Davis uses melting chip into board. My Acurite 31 still looks good so not blaming sensirion at all for this mess. 
Randy

Offline dport

  • Senior Contributor
  • ****
  • Posts: 169
Maybe Onset hasn't figured it out? Because a quick check of one of their stations in Bryn Mawr (https://www.wunderground.com/personal-weather-station/dashboard?ID=KPAHAVER17#history) shows a dew point of 72/73 when all the surrounding airports are reporting 70.

The more I observe other Sensirion sensors in other branded weather stations, the more I realize the problem isn't limited to Davis. I also see the problem in Atlas Elites and RainWise stations.

It probably goes back to what Ron (kcidwx) said about Sensirion sensors being designed for indoor use and HVAC primarily and not as meteorological instruments.

Doesn't seem like the cap matters, at least not for 11/15/31 sensors. Seems it's more about how often they've been in saturated air.

My station is just a couple miles from the Onset station you mentioned.  I’m running a 24 hr FARS VP2 that was just installed in June.  We had heavy rains yesterday and my station hit 99% humidity on many occasions last night (this is stock with no calibration adjustments).  The Onset hit a max of 97%. Currently, my dew is 72 while the Onset is 73. The nearby airports (mainly philly international) have a weird dry bias that doesn’t make sense. 

Here is my station if you want to compare to the Onset:
https://www.wunderground.com/personal-weather-station/dashboard?ID=KPABRYNM2#history
« Last Edit: August 12, 2018, 02:38:20 PM by dport »

Offline jgentry

  • Forecaster
  • *****
  • Posts: 346
OK, after looking at a bunch of Rainwise stations, I should back up and say I only occasionally see the wet bias problem.

Whereas with Davis VP2s, well, it's uncommon to find one that's accurate in the middle range of humidity (40-70%).

Also, to be fair, the Acurite Atlases listed on Wunderground may very well be other stations at this point since the "beta" test period ended, I believe, and the product has yet to be released.

Do we know for sure that Rainwise uses Sensirion? Maybe just for humidity? I noticed their specs for temp go down to -67F and I have tried to research online what they use for temp and humidity, but came up without much info.

It looks like they may have changed sensor at sometime. The RWIII now says -40 to 60c range and use to be different like -68F. There was a forum member who lived in Alaska that would come here and brag about how his station would still work at those cold temperatures.   https://www.rainwise.com/products/attachments/6801/20150710121434.pdf

The email I received from them, which was around a month ago, stated that they use the SHT-15 for MK III stations
Davis Vantage Pro2. SHT-75. WU: KXALJEMI2 & KALTHORS2. CWOP/APRS: C6353 & E6358

  

Offline jgentry

  • Forecaster
  • *****
  • Posts: 346
Did you see what Jerry and I ordered?  I have 2 on the way may not be here until fall however. https://www.aliexpress.com/item/2pcs-Waterproof-Temperature-and-humidity-sensor-shell-SHT10-SHT21-SHT15-shT11-sht20-SHT75-sensor-protective-sleeve/32725766858.html?spm=a2g0s.9042311.0.0.69af4c4dlQrpxK

Others have some like https://www.emesystems.com/sht75pg/documents/SHT75PG.pdf

So basically the SF1 filter caps are not able to protect the sensor during extreme humid conditions?
Davis Vantage Pro2. SHT-75. WU: KXALJEMI2 & KALTHORS2. CWOP/APRS: C6353 & E6358

  

Offline openvista

  • Forecaster
  • *****
  • Posts: 401
    • marquetteweather.com
The nearby airports (mainly philly international) have a weird dry bias that doesn’t make sense. 

When I checked (around 11:30AM EDT), every nearby airport was reporting 70 -- in all cardinal directions. It's unusual to see agreement like that. Again, not saying definitively that the station in question runs wet (verification would require a trustworthy instrument next to it).

Pulling up a map of the area now (3:45PM EDT) shows light winds from the N & NE and 10 mi visibility. There are 9 airports averaging about 70 again, a couple RAWS stations about the same, and state department of transportation stations (in northern Delaware) avg around 70/71. Meanwhile dew points at area PWS's are averaging about 73.
« Last Edit: August 12, 2018, 04:32:12 PM by openvista »
Davis Vantage Pro2 FARS | https://marquetteweather.com | EW7933

Offline ValentineWeather

  • Forecaster
  • *****
  • Posts: 5443
Did you see what Jerry and I ordered?  I have 2 on the way may not be here until fall however. https://www.aliexpress.com/item/2pcs-Waterproof-Temperature-and-humidity-sensor-shell-SHT10-SHT21-SHT15-shT11-sht20-SHT75-sensor-protective-sleeve/32725766858.html?spm=a2g0s.9042311.0.0.69af4c4dlQrpxK

Others have some like https://www.emesystems.com/sht75pg/documents/SHT75PG.pdf
So basically the SF1 filter caps are not able to protect the sensor during extreme humid conditions?

My understanding the whole sensor fits inside waterproof case was idea. Davis encapsulates connections in plastic/rubber material on board. 
Something needs to protect the connections or they won't last exposed to elements. May get by for awhile but eventually failure.
« Last Edit: August 12, 2018, 04:12:07 PM by ValentineWeather »
Randy

Offline dendrite

  • Forecaster
  • *****
  • Posts: 360
Anybody else notice that the 'dewpoint accuracy' graph at the bottom of page 4 [ https://www.emesystems.com/sht75pg/documents/SHT75PG.pdf ] shows a VERY interesting anomaly: the dp 'error' suddenly kinks UPWARD at and above 90%RH.
Most of them do it. They have the 31 at a steady 2% through the full range, but my hunch is it's a little higher than that in that 80/90-100% range. I mean, why would the 30 and 35 have a decrease in accuracy in high RH and not the 31?

https://www.sensirion.com/en/environmental-sensors/humidity-sensors/digital-humidity-sensors-for-various-applications/