Author Topic: The Reliability of the SHT-31 Humidity Sensor & What Psychrometer Should I Buy?  (Read 6329 times)

ct, Dj1225 and 3 Guests are viewing this topic.

Offline ValentineWeather

  • Forecaster
  • *****
  • Posts: 4826
I was thinking about what kcidwx said about how abnormal it is to aspirate the humidity sensor. There must be reason for it...Possibly they wear out much faster...Anyway 4 days ago I started testing the Davis 24 hr shield in passive mode. 
We've only had 1 day with light enough winds I considered a good test day. It's the best passive shield I've ever tested. The sensor is well protected inside from thermal heat and painted inside flat black, radiation and reflective heat is reduced also.
Peak temperature was only 1 degree warmer than the AC fan and 1.5° warmer than the dc solar fan. The DC fan actually runs cooler than the AC on average .5F.

What I did to modify was removed fan, painted interior where sensor mounts all the way to bottom flat black added a couple more top shields, but those with rain bucket not necessary and drilled some holes on outer shield core...This may not be necessary also.  Final step I used same 5/8" bit and drilled a bottom center hole through plastic for a little more upward airflow and removed the small pyramid shaped black shield on bottom.

This is graph of the one day with wind speed light enough to consider a valid test under clear cloudless sky with high summer sun angle 43° longitude.
#3 orange is passive, #1 AC fan #5 Dc fan

Added wind speed and dewpoint
 
« Last Edit: July 12, 2018, 06:21:22 AM by ValentineWeather »
Randy

Offline Old Tele man

  • Singing in the rain...
  • Forecaster
  • *****
  • Posts: 1242
Nobody's mentioned it, but most probably know it, but there IS a possibility that DAVIS uses the combined TEMP/RH sensor because internally within the ISS they need the TEMP to compensate the RH sensor's temperature non-linearity. Anybody know for sure *IF* the SHT-31 needs temp-compensation?
• SYS: Davis VP2 Vue/WL-IP & Envoy8X/WL-USB
• CWOP: DW6988 - 2 miles NNE of Cortaro, AZ
• WU - KAZTUCSO202, Countryside

Offline jerryg

  • Forecaster
  • *****
  • Posts: 560
  • weather is never boring
    • victoria texas weather
The putty is more than enough to hold the filter on, it is so small and light it was in place for a year and i had to force it loose from the sensor. The only reason i changed it was i thought after a year it might be getting clogged but it looked ok and the new one did not show any change in performance. I used the end of a toothpick to apply a small amount to each side and mold the filter to the sensor.

Offline ValentineWeather

  • Forecaster
  • *****
  • Posts: 4826
Nobody's mentioned it, but most probably know it, but there IS a possibility that DAVIS uses the combined TEMP/RH sensor because internally within the ISS they need the TEMP to compensate the RH sensor's temperature non-linearity. Anybody know for sure *IF* the SHT-31 needs temp-compensation?

It was mentioned they are totally different circuits yesterday I believe the big debate with Davis techs why you can change temperature and RH doesn't move. 
Randy

Offline jerryg

  • Forecaster
  • *****
  • Posts: 560
  • weather is never boring
    • victoria texas weather
I was just thinking how all this discussion is about a 45 dollar sensor. When the cost is considered it is a very good deal for home use. I priced the Vaisala HMP155 sensor which is one of the better ones out there and the price was 823.00 dollars, wow. The spec for the top end stopped at 97% and it was + or -  1.8% not far from the cheap 31 spec. at 100%. I am just glad that there is a sensor that works really well at a price i can afford. I can buy a whole lot of replacement sensors for the price of the better sensor lol.

Offline openvista

  • Senior Contributor
  • ****
  • Posts: 235
    • marquetteweather.com
Nobody's mentioned it, but most probably know it, but there IS a possibility that DAVIS uses the combined TEMP/RH sensor because internally within the ISS they need the TEMP to compensate the RH sensor's temperature non-linearity. Anybody know for sure *IF* the SHT-31 needs temp-compensation?

It was mentioned they are totally different circuits yesterday I believe the big debate with Davis techs why you can change temperature and RH doesn't move.

I believe I'm the one who referenced that "debate" which was really more of an act more than anything in retrospect. Seems unlikely Davis engineers had no idea that temperature had no influence on relative humidity in the console they designed.

Sure, from the console's perspective temperature and humidity are two separate values that seem unrelated. But as for the SIM, I'd assume it uses temperature for no other reason than to calculate saturation vapor pressure as a denominator for vapor pressure (the ratio that produces humidity). Maybe it has a lookup table instead? As to whether the SIM also uses temperature as a parameter for a calibration routine, I would not know.
Davis Vantage Pro2 | https://marquetteweather.com

Offline openvista

  • Senior Contributor
  • ****
  • Posts: 235
    • marquetteweather.com
I was just thinking how all this discussion is about a 45 dollar sensor. When the cost is considered it is a very good deal for home use. I priced the Vaisala HMP155 sensor which is one of the better ones out there and the price was 823.00 dollars, wow. The spec for the top end stopped at 97% and it was + or -  1.8% not far from the cheap 31 spec. at 100%. I am just glad that there is a sensor that works really well at a price i can afford. I can buy a whole lot of replacement sensors for the price of the better sensor lol.

Perhaps the reason you've had good luck with these sensors is you knew to put the Sensirion cap on them. Did you notice a difference in their long-term accuracy once you started doing that?
Davis Vantage Pro2 | https://marquetteweather.com

Offline WheatonRon

  • Forecaster
  • *****
  • Posts: 851
    • WUnderground
I was just thinking how all this discussion is about a 45 dollar sensor. When the cost is considered it is a very good deal for home use. I priced the Vaisala HMP155 sensor which is one of the better ones out there and the price was 823.00 dollars, wow. The spec for the top end stopped at 97% and it was + or -  1.8% not far from the cheap 31 spec. at 100%. I am just glad that there is a sensor that works really well at a price i can afford. I can buy a whole lot of replacement sensors for the price of the better sensor lol.

Jerry,

I agree with you, but I wish Davis informed us of the requirement to change sensors every 12 to 24 months depending on where you live (climate dryness) to ensure accuracy. This would have done two things—made me feel better about Davis and would have forced me to do a job I hate—cleaning my ISS. It ranks right below sanitizing my reverse osmosis drinking system as the worst home maintenace job I have.
« Last Edit: July 11, 2018, 09:48:59 PM by WheatonRon »
Davis VP2 with 24 hour FARS, SHT31 (3 complete systems-2 for uploading to the internet the other system for test and play); CWOP--CW5020 and FW3075; WU--KILWHEAT17 and KILWHEAT36; WeatherCloud.net; CoCoRaHS--IL-DP-132; Rainwise 111

Offline ValentineWeather

  • Forecaster
  • *****
  • Posts: 4826
Here is link to Mouser the filter is 1.06 plus shipping.https://www.mouser.com/Sensirion/Sensors/Sensor-Hardware-Accessories/_/N-11ei3?P=1ytt2dn  It is the sf2 filter.

Jerry is this the filter you cut the legs off and glue on?
When you say sealing putty would any glue work?
Randy

Offline jgentry

  • Senior Contributor
  • ****
  • Posts: 190
    • Gentry Field Station
I was just thinking how all this discussion is about a 45 dollar sensor. When the cost is considered it is a very good deal for home use. I priced the Vaisala HMP155 sensor which is one of the better ones out there and the price was 823.00 dollars, wow. The spec for the top end stopped at 97% and it was + or -  1.8% not far from the cheap 31 spec. at 100%. I am just glad that there is a sensor that works really well at a price i can afford. I can buy a whole lot of replacement sensors for the price of the better sensor lol.

TBH, I think the high prices of these sensors like Vaisala is due to more of supply/demand than in terms of the quality of the sensors. Sure, they will be higher due to the components and their quality of materials but the supply/demand plays a bigger role. I’m sure mass production of chips is way cheaper than making one Vaisala sensor. That’s basically my opinion & I could be wrong.

Now I’m curious if the SHT-35 is any better than the 31?
Davis Vantage Pro2. SHT 31. 24hr FARS. WU: KXALJEMI2 & KALTHORS2 CWOP: C6353 & C6358

  

Offline kcidwx

  • Forecaster
  • *****
  • Posts: 318
As far as why you shouldn't aspirate a humidity sensor goes.... Two things that will kill a humidity sensor prematurely, contaminants/pollution and condensation. With the SHT-31 in the Davis FARS you are throwing both at it. At or near saturation, there's no heater to keep the sensor temperature a degree or two above the dew point so condensation doesn't form on the sensor and, you are blowing dirty air across it 24/7. Definitely not best practice for sensor longevity. It comes down to how much abuse the SHT-31 can handle. I have my Vaisala HMT337 humidity sensor in a passive shield and the filter never gets very dirty. On my RM Young FARS for my temperature sensor it gets incredibly dirty from pulling air through it 24/7. If Sensirion offers their own filter that's superior over the Davis filter, then I would go that route. Ideally for temperature compensation you can use the onboard temperature sensor for the humidity sensor temperature in a passive shield. Then use a separate temperature probe in a FARS for ambient air temperature. That's best practice.
Meteorological Technician
NWS Certified Aviation Weather Observer
B.S. Meteorology

Offline CW2274

  • Forecaster
  • *****
  • Posts: 3551
    • Conditions @ CW2274
As far as why you shouldn't aspirate a humidity sensor goes.... Two things that will kill a humidity sensor prematurely, contaminants/pollution and condensation.
Well, according to Senserion, this filter, when properly sealed around the sensor should eliminate all these factors.

Offline WheatonRon

  • Forecaster
  • *****
  • Posts: 851
    • WUnderground
As far as why you shouldn't aspirate a humidity sensor goes.... Two things that will kill a humidity sensor prematurely, contaminants/pollution and condensation.
Well, according to Senserion, this filter, when properly sealed around the sensor should eliminate all these factors.

So does the Davis 24 hour FARS seal it like this to eliminate “these factors”?
Davis VP2 with 24 hour FARS, SHT31 (3 complete systems-2 for uploading to the internet the other system for test and play); CWOP--CW5020 and FW3075; WU--KILWHEAT17 and KILWHEAT36; WeatherCloud.net; CoCoRaHS--IL-DP-132; Rainwise 111

Offline CW2274

  • Forecaster
  • *****
  • Posts: 3551
    • Conditions @ CW2274
As far as why you shouldn't aspirate a humidity sensor goes.... Two things that will kill a humidity sensor prematurely, contaminants/pollution and condensation.
Well, according to Senserion, this filter, when properly sealed around the sensor should eliminate all these factors.

So does the Davis 24 hour FARS seal it like this to eliminate “these factors”?
No.

Offline jerryg

  • Forecaster
  • *****
  • Posts: 560
  • weather is never boring
    • victoria texas weather
Well i am no expert on what to use to secure the filter to the sensor i used the putty because i had it on hand and it was sticky and water proof plus it didn't have any chemical spell to it that might be bad for the sensor. I have some stuff that is used to seal electrical connections that is non corrosive that i could have used but wanted to be able to remove the filter for cleaning.
What a mess, need fars for good temp reports but bad for humidity, no fars good for humidity but bad for temp with no wind.  ](*,)

Offline klschmidt

  • Member
  • *
  • Posts: 15
Would switching the aspirator fan off when wind exceeds a certain speed (2MPH) be advantageous for sensor life?

Offline CW2274

  • Forecaster
  • *****
  • Posts: 3551
    • Conditions @ CW2274
but wanted to be able to remove the filter for cleaning.
Why? If the inside is sealed, why would removal be necessary?

Offline jerryg

  • Forecaster
  • *****
  • Posts: 560
  • weather is never boring
    • victoria texas weather
Because i had no info as to how well the filter would perform until i tried one and with the seal all the way around the outside of the filter the only way the air can be sampled is through the membrane. When you compare the big Davis shield to the really small fl2 filter and its little opening it looks like it would plug up in no time but even out here in the country surrounded by farm land that is plowed and harvested all the time with crop dusting a surface spraying it has held up real well. My humidity runs over 90% nearly every night unless in a major drought lol. Two of my 3 sensors are over 2 years old in this bad, humid and dusty area and are still tracking with the new sensor i just put in. As far as i can tell i have not had any wet bias.

Offline jgentry

  • Senior Contributor
  • ****
  • Posts: 190
    • Gentry Field Station
Because i had no info as to how well the filter would perform until i tried one and with the seal all the way around the outside of the filter the only way the air can be sampled is through the membrane. When you compare the big Davis shield to the really small fl2 filter and its little opening it looks like it would plug up in no time but even out here in the country surrounded by farm land that is plowed and harvested all the time with crop dusting a surface spraying it has held up real well. My humidity runs over 90% nearly every night unless in a major drought lol. Two of my 3 sensors are over 2 years old in this bad, humid and dusty area and are still tracking with the new sensor i just put in. As far as i can tell i have not had any wet bias.

You might have solved our problems! Lol  Does your sensors struggle to reach 100%?

I hope the filter cap is the answer. I would’ve love to see the DP readings match up well with the Airport stations. Only difference I want to see is due to transpiration and not from a wet biased sensor.
Davis Vantage Pro2. SHT 31. 24hr FARS. WU: KXALJEMI2 & KALTHORS2 CWOP: C6353 & C6358

  

Offline openvista

  • Senior Contributor
  • ****
  • Posts: 235
    • marquetteweather.com
The problem with maintaining two sensors, one in passive for humidity and one aspirated for temp, is that there isn't a Davis console (I'm aware of) that can feed temperature and humidity on separate IDs (what some folks call "channels"). You could add a 2nd console to receive the 2nd SIM signal, but that doesn't solve how you feed your weather software both consoles' data simultaneously.  While it may be possible to run two different, incompatible weather apps concurrently and somehow feed humidity from one and temperature from the other to your website (custom job but do-able), I can't see how you would be able to send both to the same statistics database, let alone online services like CWOP & WU, without some serious custom programming.

At this point, I'd have to side with jerryg. Does it make sense to spend hundreds in initial and ongoing expenses and invite untold headaches all in an effort to save a $45 sensor from annual or semi-annual replacement?

Food for thot. Old Teleman provided a link several pages ago in this thread to an academic study which concluded that naturally aspirated humidity readings are, in fact, less accurate, on the whole, than fan aspirated readings. Why? Same reason as with temp readings: on low wind days an artificial environment can build up within the shield. I remember reading that elsewhere too.

If the cap solution work out, then maybe we can have our cake and eat it too?
Davis Vantage Pro2 | https://marquetteweather.com

Offline kcidwx

  • Forecaster
  • *****
  • Posts: 318
Here's the spec on that Filter Cap SF2.

Filtration efficiency 99.99% 
0.1μm particles at 0.05m/s air flow.

I'm not sure what the Davis fan pulls for air flow. Keep in mind as you increase air flow through the chamber the filter becomes less efficient but there's no data table to show by how much. I think this filter is your best bet.

Anyway, that's why I don't like FARS on humidity sensors. You are constantly ramming air through the chamber 24/7 and there's always some contaminants/pollution that's getting through the filter to the sensor.
« Last Edit: July 12, 2018, 12:34:58 PM by kcidwx »
Meteorological Technician
NWS Certified Aviation Weather Observer
B.S. Meteorology

Offline CW2274

  • Forecaster
  • *****
  • Posts: 3551
    • Conditions @ CW2274
I think I'll still use the stock filter as well considering my strong fan and dusty climate, my sensor chamber gets filthy. I'm sure it'll slow response slightly, but I'm okay with that for the added protection.

Offline kcidwx

  • Forecaster
  • *****
  • Posts: 318
Food for thot. Old Teleman provided a link several pages ago in this thread to an academic study which concluded that naturally aspirated humidity readings are, in fact, less accurate, on the whole, than fan aspirated readings. Why? Same reason as with temp readings: on low wind days an artificial environment can build up within the shield. I remember reading that elsewhere too.

That depends on the design of the passive shield. The open air passive shields such as the Vaisala one used on the DTS1 for ASOS doesn't suffer from that. I always bring that up because PWS owners are always comparing their humidity readings to ASOS and ASOS doesn't use a fan aspirated shield with the humidity sensor. It isn't needed. I have an open air shield as well for my humidity sensor and love it. Since the Davis FARS is an enclosed chamber, then I can see why a fan is needed. The open air shields are not the same as the gill plated passive shields that are notorious for trapping stagnate air.

At this point, I'd have to side with jerryg. Does it make sense to spend hundreds in initial and ongoing expenses and invite untold headaches all in an effort to save a $45 sensor from annual or semi-annual replacement?

If $45 is all it costs then yes. Buy the sensor, put the filter on it, install it and replace as needed.
« Last Edit: July 12, 2018, 01:10:48 AM by kcidwx »
Meteorological Technician
NWS Certified Aviation Weather Observer
B.S. Meteorology

Offline johnd

  • Forecaster
  • *****
  • Posts: 2697
    • www.weatherstations.co.uk
The problem with maintaining two sensors, one in passive for humidity and one aspirated for temp, is that there isn't a Davis console (I'm aware of) that can feed temperature and humidity on separate IDs (what some folks call "channels").

If you can still find an Envoy8X then that would be one option of course. A Meteobridge Pro Red (or whatever Ed calls it in the US) or Meteostick-based station would be another.
Prodata Weather Systems
Prodata's dedicated Davis EnviroMonitor website
UK Davis Premier Dealer - All Davis stations, accessories and spares
Littleport, Ely, Cambs UK

Offline ValentineWeather

  • Forecaster
  • *****
  • Posts: 4826
I would like to see the SHT31 with some side by side temperature data against vaisala temperature sensor.
The ASOS here has been running around 2° warmer than my instruments. I've taken nearby data with portable FARS and Davis console nearby and come up 1.7° lower vs what the ASOS 5 minute average shows.  I've cross checked the SHT31 with NIST certified and they are very close within .1-.2F.
Does anyone know of a study where this has been conducted?
Randy

 

anything