Author Topic: The Reliability of the SHT-31 Humidity Sensor & What Psychrometer Should I Buy?  (Read 31623 times)

0 Members and 1 Guest are viewing this topic.

Offline jerryg

  • Forecaster
  • *****
  • Posts: 746
  • weather is never boring
    • victoria texas weather
Well what is really disappointing is it still has the same operating conditions of all the sensors, 20 to 80% above that the 3% error, etc. They just don't care about using the sensor in high humidity sites. It looks like the same old junk to me. We need Davis to change over to a sensor that is not so sensitive to high humidity.

Offline ValentineWeather

  • Forecaster
  • *****
  • Posts: 5472
Anyone else getting the sensirion cables for the SHT75, Digikey has them back in stock. Remember price 1 order includes 4 cables per order. https://www.digikey.com/product-detail/en/sensirion-ag/3M-CABLES-FOR-EK-H4/1649-1055-ND/5982944

Existing Jack needs cut off and replaced with RJ12. I just added RJ12 jack on 8 cables I have and tested. No idea what I'll do with that many cables.

Wires from sensirion cable insert into jack pins this order. The cable has a raised knob on one side this side needs 75 JD3 label up.
Green-1
White-2
Brown-3
Yellow-6


FYI anyone working with these SHT75s the pins can bend easily so be careful use plenty of light making sure you are lined up properly before inserting.  They don't insert all the way, only so far so pins are exposed.

 [ You are not allowed to view attachments ]  
 [ You are not allowed to view attachments ]




« Last Edit: October 10, 2018, 08:05:20 PM by ValentineWeather »
Randy

Offline openvista

  • Forecaster
  • *****
  • Posts: 419
    • marquetteweather.com
I recently received a new SHT31 sensor assembly. I installed it in my AC FARS on October 7th.

After performing almost 2 dozen tests with a 3 month old, factory certified Kestrel 5000 (+/- 2%) and cross referencing with the nearby airport and another SHT31 I own with a consistent bias (of about +8%), I can confidently state that the sensor arrived with an average 6% middle range wet bias. It's consistent from all the way up to the low 80 percentiles at which point it's about half that (+3%) until you reach accuracy around 90-93%.  The lowest humidity I've tested is 51%, but I have no reason to believe the error won't extend down another 20% or so into the 30 percentiles (that's the behavior of my other SHT31 sensors).

The new sensor has been maxing out at 98% but only after hours of exposure to saturated air. Effectively, 97% has been the ceiling since it was installed. I expect that maximum value to slowly decline over the next several months to around 95-96%, as has been the case with my 2 other 31s so far (along with reports of the same from many other owners and surveys of nearby VP2s).

I should point out that due to a cold snap, I've only tested in the 30-50F range. No idea how it performs in warmer weather.
« Last Edit: October 17, 2018, 02:07:45 PM by openvista »
Davis Vantage Pro2 FARS | https://marquetteweather.com | EW7933

Offline openvista

  • Forecaster
  • *****
  • Posts: 419
    • marquetteweather.com
Just to emphasize... unlike my last 31, the new one is never accurate in the middle ranges. My previous 31 would encounter a long period of 80%+ humidity and gain a wet bias that would last an indeterminate amount of time until it dropped below 80% long enough to recalibrate. There's no way you can deal with that other than to wait. Meanwhile, my readings would be 8-10% off for hours or even days. Embarrassing.

At least when the error is consistent, I am able to custom program a couple offsets below 90% and a couple others above 95% and be done. So this sensor is an improvement in that way, but still 3x outside it's spec when BRAND NEW. And who knows if its behavior remains stable over time.
Davis Vantage Pro2 FARS | https://marquetteweather.com | EW7933

Offline CW2274

  • Forecaster
  • *****
  • Posts: 4019
    • Conditions @ CW2274
The 31 I'm using I bought 3/16 and has about 8 months of service on it. I have no where near the error you're experiencing as I only use a -2% correction factor. As far as I can tell, it's very good there as I've been down to 11% and up to 96% with the -2%. Now, does that mean older ones are more reliable or that my dry climate helps or both or neither or.... #-o

Offline ValentineWeather

  • Forecaster
  • *****
  • Posts: 5472
My thinking about all of this maybe we need to add dewpoint temperature when we talk about Relative Humidity being its only relative to the ambient temperature and not a good indicator of actual moisture content in the air which IMO is the issue.
 
100% at 32F (0c) for days is still very little moisture and won't oversaturate the sensor membrane is my thinking. Now take 85% at 65+dp the sensor membrane is absorbing lots of moisture in just hours.

Randy

Offline drew1021

  • Contributor
  • ***
  • Posts: 131
This has been my exact observation in the 10 months my 31 has been in use. I used my Belford 566 as my reference. Hopefully Davis will find a solution to the problem. I would be willing to pay more for better accuracy.
VP2 with 24 hour FARS

Offline CW2274

  • Forecaster
  • *****
  • Posts: 4019
    • Conditions @ CW2274
The 96% I hit was only a few days ago, without the correction, I was at 98% with the temp at 59.9F, dew, 59.3. The dew remained there for about 11 hours, then seemed to respond fine to the drying air later the next morning.

Offline wxmanmhd

  • Member
  • *
  • Posts: 18
    • FM Weather
I have a Vantage Pro2 with 24 hr Aspiration Fan and was wondering if with a steady air movement across the SHT 31 sensor will keep it a little lower than the shield without a fan. Has anyone compared that?

Offline CW2274

  • Forecaster
  • *****
  • Posts: 4019
    • Conditions @ CW2274
Has anyone compared that?
For years and years. The fan is there to keep ambient air flowing across the sensor at all times, no matter what the wind is doing.

Offline wxmanmhd

  • Member
  • *
  • Posts: 18
    • FM Weather
I am sorry I forgot to mention that if it affects the humidity part of the sensor.

Offline CW2274

  • Forecaster
  • *****
  • Posts: 4019
    • Conditions @ CW2274
I am sorry I forgot to mention that if it affects the humidity part of the sensor.
So it is said. To what degree I think has much to do with what climate the 31 is exposed to.

Offline openvista

  • Forecaster
  • *****
  • Posts: 419
    • marquetteweather.com
My thinking about all of this maybe we need to add dewpoint temperature when we talk about Relative Humidity being its only relative to the ambient temperature and not a good indicator of actual moisture content in the air which IMO is the issue.
 
100% at 32F (0c) for days is still very little moisture and won't oversaturate the sensor membrane is my thinking. Now take 85% at 65+dp the sensor membrane is absorbing lots of moisture in just hours.

When I conduct a test, I am recording dew points and then deriving the correct humidity from that using a dew point calculator (see: http://andrew.rsmas.miami.edu/bmcnoldy/Humidity.html - the underlying formula is widely used in weather software). Otherwise, if there were a difference in temperature between the test device and the VP2, the humidity wouldn't be comparable.  Dew point should always be the same regardless of measured temp. *

Regardless, and very curiously, I nearly always arrive at a +6% difference. Same goes for my backup VP2 with an older SHT31 - it has a consistent +8% bias. Comparisons to the airport confirm the dewpoint errors and test device accuracy.


* NOTE: Just in case anyone is a stickler, normally I let the test device (Kestrel) acclimate until there's less than 1F difference between it and SHT31. This way even the reported humidities should match within about 1% and I can double check my calculations. I do not hold the device in my hand, but rather let it hang from a hook and orient the sensor into the wind on my covered, but otherwise open porch (this way I can measure in rain and snow too). I've compared the readings taken on the porch to those using a tripod next to the ISS and they are the same. The issue with using a tripod is sun exposure, which is ill-advised. I constructed an umbrella for it, but that doesn't do enough to reduce radiation errors and acts as a sail in the wind. The north-facing porch method nearly always matches the airport to within 1F when dew points are stable and well-mixed synoptic conditions prevail. 
« Last Edit: October 18, 2018, 09:03:42 AM by openvista »
Davis Vantage Pro2 FARS | https://marquetteweather.com | EW7933

Offline openvista

  • Forecaster
  • *****
  • Posts: 419
    • marquetteweather.com
I have a Vantage Pro2 with 24 hr Aspiration Fan and was wondering if with a steady air movement across the SHT 31 sensor will keep it a little lower than the shield without a fan. Has anyone compared that?

Theoretically, fan aspiration shouldn't lower or raise accurately measured humidity. Some of the best psychrometers are fan aspirated. That said, when compared to a passively ventilated shield, on some clear, still mornings the FARS may report slightly higher temperatures (and thus lower humidities) compared to some lesser quality passive shields. The FARS temp in this scenario is actually more accurate and, therefore, the humidity should be as well. However, if you observe widespread dew and the measured humidity is less than 100% (which it will be in virtually all VP2s due to an unrelated dry bias in the sensor), it's because the air and surrounding objects are not shielded from longwave radiation which has a cooling effect driving the temp down to the dew point. This is somewhat akin to the difference of how the temperature feels when you are standing in the sun vs the same temperature in the shade.

There's a more general problem with measuring humidity inside radiation shields (something kcidwx, the NWS tech, covered much earlier in this thread). If the air has been saturated for awhile, the inside of the shield can collect moisture. You can argue that a fan will make a shield wetter when the air is loaded, particularly in low wind conditions. But the opposite is also true: a fan accelerates evaporation and should return the shield to balance with the outside air once humidity begins to drop.
« Last Edit: October 18, 2018, 12:59:04 PM by openvista »
Davis Vantage Pro2 FARS | https://marquetteweather.com | EW7933

Offline jerryg

  • Forecaster
  • *****
  • Posts: 746
  • weather is never boring
    • victoria texas weather
I just did a dry/wet bulb test and it showed the humidity to be 87.9 and my console reading was 88% this is with the sht 75, looks real good, i tested several times with the same results. I found a good chart to calculate this with, it uses dry/wet/barometer inputs to calculate with.

Offline openvista

  • Forecaster
  • *****
  • Posts: 419
    • marquetteweather.com
I just did a dry/wet bulb test and it showed the humidity to be 87.9 and my console reading was 88% this is with the sht 75, looks real good, i tested several times with the same results. I found a good chart to calculate this with, it uses dry/wet/barometer inputs to calculate with.

OK, but just to be clear, I'm not questioning whether the console's (or your weather software's) dew point calculation is off. I'm saying my new SHT31 relative humidity sensor runs wet below about 90%. I only mentioned the dew point formula just in case anyone was questioning my method of calculating humidity discrepancies from dew points.
Davis Vantage Pro2 FARS | https://marquetteweather.com | EW7933

Offline jerryg

  • Forecaster
  • *****
  • Posts: 746
  • weather is never boring
    • victoria texas weather
When it comes to dp and wet bulb stuff i get semi lost, but the formulas on the web sure help. I also noticed with my 31 as the humidity dropped below 90% the readings compared to the 75 would slowly get larger with the 31 getting up to 3% higher around 80%. I see the same thing with the 15 but haven't seen it on the 75 so far. I just don't like the fact that the operating range is 20 to 80% everything else is up for grabs as far as accuracy goes. I am not going to hold my breath until Davis uses a sensor without this problem.

Offline Old Tele man

  • Singing in the rain...
  • Forecaster
  • *****
  • Posts: 1372
Sensor design & technology is good, but NOT perfect. Sometimes, "close enuf" is all you're gonna get...unless you don't mind mortgaging the house.
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 jerryg

  • Forecaster
  • *****
  • Posts: 746
  • weather is never boring
    • victoria texas weather
 :lol:

Offline wxmanmhd

  • Member
  • *
  • Posts: 18
    • FM Weather
Thanks for all the great info. I also was thinking of some evaporation with the FARS and agree sensors are pretty good vs the cost.

Offline openvista

  • Forecaster
  • *****
  • Posts: 419
    • marquetteweather.com
Well, just to make it interesting, we finally had a day above 55F. And wouldn't you know it, the sensor is now running at +4% (rather than +6%) in the warmer air!  ](*,) That still equates to about a +2.5F dew point error.

Not looking for perfection, but when the manufacturer states that their sensor is +/- 2% *through the entire range* and that's not even close to true (not even at a given temp like 25C), I take issue. If it was consistently even +/- 4%, I could live with it. But having a varying wet bias from 35-85% (!) followed by a dry bias is unacceptable. This is my 3rd SHT31 and 2 of them have run wet at all humidities below 90% and the 3rd ran wet after extended periods above 80%. It's become near impossible to get these sensors to operate within specification and that's coming from a professional software developer.

Kestrel's custom designed humidity sensor is *damn good* (truly within +/- 2% at every temp I've tested so far) and it's in a $250 device. So, no, you don't need to 2nd mortgage your house to afford a good humidity sensor.
« Last Edit: October 18, 2018, 11:11:05 PM by openvista »
Davis Vantage Pro2 FARS | https://marquetteweather.com | EW7933

Offline drew1021

  • Contributor
  • ***
  • Posts: 131
I don't know about everyone else but I am getting impatient waiting for Davis to come up with a solution to the infamous wet bias of the 31. Its frustrating when all around me there are <200 dollar stations, ambient weather, showing reasonable humidity values as compared to my Belfort 566 and local ASOS.  Example, my Belfort was reading 67%, local ASOS 65%, my davis 73% ](*,)
« Last Edit: November 04, 2018, 10:46:18 AM by drew1021 »
VP2 with 24 hour FARS

Offline jerryg

  • Forecaster
  • *****
  • Posts: 746
  • weather is never boring
    • victoria texas weather
I don't think we will get anything better until Davis changes over to the i2c format and thus gets a larger variety of sensors to choose from. They have no options right now with the ls format. I fixed my problem by switching to the 75 sensor and making my own but even that is going to be a problem with the 75 being done away with. The maker of the sensor admits to the sensor being a lemon above 90% and should not be used in a wet setup. I blame Davis for picking a sensor for their stations with that type of warning from the maker.

Offline drew1021

  • Contributor
  • ***
  • Posts: 131
Agree, but is Davis going to listen to us and begin working on the fix? If enough people complain or care enough to voice their concerns to Davis then maybe. The bottom line is if sales of the VP2 suffer then we have a better chance.
VP2 with 24 hour FARS

Offline CW2274

  • Forecaster
  • *****
  • Posts: 4019
    • Conditions @ CW2274
Agree, but is Davis going to listen to us and begin working on the fix? If enough people complain or care enough to voice their concerns to Davis then maybe. The bottom line is if sales of the VP2 suffer then we have a better chance.
I think that the VP2 owners here on this forum are a minuscule segment of their business and not worth a "drop everything that you're doing and fix it for these whiners" type of scenario. If it was a problem that was so obvious to ALL, then it might be a different story, but I think the vast majority are oblivious, or don't care enough to voice a complaint.
I very well could be and hope I'm wrong, but doubtful because I'd think there'd be a fix already.

 

anything