Author Topic: The Reliability of the SHT-31 Humidity Sensor & What Psychrometer Should I Buy?  (Read 40943 times)

0 Members and 1 Guest are viewing this topic.

Offline ValentineWeather

  • Forecaster
  • *****
  • Posts: 5854
    • LiVE WEATHER
Wow, I'm happily shocked about returns. I figured it would be a "pound sand " approach, nature of the beast thing. Nicely done Randy. =D>

That is the approach I've had from them before but Ryan is different and reliable for sure.

Still nothing back from Davis Inst.
They may or may not get back I've had it both ways and been flat lied to before when I pointed out an software issue for cup size with Vue console using VP2 wind sensor.

They claimed they ran wind tunnel test which I know wasn't true if they had they would of seen problem immediately.
It wasn't just me but multiple people with VP2 and Vue consoles were seeing different wind speeds.
 
Long story short they did a Vue firmware update, not long after and first one in years (how convenient).
With no mention in update but it fixed issue. I almost didn't bother with it but someone said it fixed the wind speed issue and sure enough it did.
Now both Vue and Vp2 consoles follow each other perfectly.
 
Randy

Offline CW2274

  • Forecaster
  • *****
  • Posts: 4728
    • Conditions @ CW2274 West Tucson-Painted Hills Ranch
Wow, I'm happily shocked about returns. I figured it would be a "pound sand " approach, nature of the beast thing. Nicely done Randy. =D>

That is the approach I've had from them before but Ryan is different and reliable for sure.

Yes, my comment was geared towards Davis, not Ryan. Honestly, not holding my breath with Davis, I think they'll say we did our R&D, take it or leave it, if they comment at all for that matter.

Offline openvista

  • Forecaster
  • *****
  • Posts: 453
    • marquetteweather.com
Wow, I'm happily shocked about returns. I figured it would be a "pound sand " approach, nature of the beast thing. Nicely done Randy. =D>

That is the approach I've had from them before but Ryan is different and reliable for sure.

Still nothing back from Davis Inst.
They may or may not get back I've had it both ways and been flat lied to before when I pointed out an software issue for cup size with Vue console using VP2 wind sensor.

They claimed they ran wind tunnel test which I know wasn't true if they had they would of seen problem immediately.
It wasn't just me but multiple people with VP2 and Vue consoles were seeing different wind speeds.
 
Long story short they did a Vue firmware update, not long after and first one in years (how convenient).
With no mention in update but it fixed issue. I almost didn't bother with it but someone said it fixed the wind speed issue and sure enough it did.
Now both Vue and Vp2 consoles follow each other perfectly.

Sounds familiar.

Last year I pointed out to them that temperatures could be adjusted in the VP2 console (via offsets) without a corresponding change in relative humidity and they acted like that was news to them (on a 12 year old console). The head of Technical Support had to call in senior engineers and the staff meteorologist to get to the bottom of it!  Ultimately, after over a dozen emails, I requested they simply make a note in the manual in the section with the offset procedures so users are aware of this fact. I didn't even demand they fix it. Yet, they would not commit to even editing a damn PDF and, as far as I know, they have not notified their customers of this issue. I guess they see this as an upgrade opportunity ($$) for anyone who has (what they assume to be without understanding all use cases) an out-of-calibration sensor?

And, of course, on the issue with cold temperature signal drop outs they were nowhere to be found even though several people reached out to them directly and johnd, one of their authorized dealers, was quite involved in that discussion.

So, no, I don't expect them to really acknowledge this publicly either.

Now, having said that, I hope Davis will prove me wrong. Don't care as long as there's a solution.
« Last Edit: July 10, 2018, 06:29:06 PM by openvista »
Davis Vantage Pro2 AC FARS | https://marquetteweather.com | EW7933

Offline CW2274

  • Forecaster
  • *****
  • Posts: 4728
    • Conditions @ CW2274 West Tucson-Painted Hills Ranch
Wow, I'm happily shocked about returns. I figured it would be a "pound sand " approach, nature of the beast thing. Nicely done Randy. =D>

That is the approach I've had from them before but Ryan is different and reliable for sure.

Still nothing back from Davis Inst.
They may or may not get back I've had it both ways and been flat lied to before when I pointed out an software issue for cup size with Vue console using VP2 wind sensor.

They claimed they ran wind tunnel test which I know wasn't true if they had they would of seen problem immediately.
It wasn't just me but multiple people with VP2 and Vue consoles were seeing different wind speeds.
 
Long story short they did a Vue firmware update, not long after and first one in years (how convenient).
With no mention in update but it fixed issue. I almost didn't bother with it but someone said it fixed the wind speed issue and sure enough it did.
Now both Vue and Vp2 consoles follow each other perfectly.
Last year I pointed out to them that temperatures could be adjusted in the VP2 console (via offsets) without a corresponding change in relative humidity and they acted like that was news to them (on a 12 year old console).
I'm confused, changing the temp does effect the RH, why would you not want the offset to effect it?

Offline openvista

  • Forecaster
  • *****
  • Posts: 453
    • marquetteweather.com
Wow, I'm happily shocked about returns. I figured it would be a "pound sand " approach, nature of the beast thing. Nicely done Randy. =D>

That is the approach I've had from them before but Ryan is different and reliable for sure.

Still nothing back from Davis Inst.
They may or may not get back I've had it both ways and been flat lied to before when I pointed out an software issue for cup size with Vue console using VP2 wind sensor.

They claimed they ran wind tunnel test which I know wasn't true if they had they would of seen problem immediately.
It wasn't just me but multiple people with VP2 and Vue consoles were seeing different wind speeds.
 
Long story short they did a Vue firmware update, not long after and first one in years (how convenient).
With no mention in update but it fixed issue. I almost didn't bother with it but someone said it fixed the wind speed issue and sure enough it did.
Now both Vue and Vp2 consoles follow each other perfectly.
Last year I pointed out to them that temperatures could be adjusted in the VP2 console (via offsets) without a corresponding change in relative humidity and they acted like that was news to them (on a 12 year old console).
I'm confused, changing the temp does effect the RH, why would you not want the offset to effect it?

No, that's what I'm saying. Currently, changing the temperature via an offset in the console does NOT affect the RH. That's not correct. I pointed it out and Keystone Cops ensued over at Davis HQ.
Davis Vantage Pro2 AC FARS | https://marquetteweather.com | EW7933

Offline CW2274

  • Forecaster
  • *****
  • Posts: 4728
    • Conditions @ CW2274 West Tucson-Painted Hills Ranch
Wow, I'm happily shocked about returns. I figured it would be a "pound sand " approach, nature of the beast thing. Nicely done Randy. =D>

That is the approach I've had from them before but Ryan is different and reliable for sure.

Still nothing back from Davis Inst.
They may or may not get back I've had it both ways and been flat lied to before when I pointed out an software issue for cup size with Vue console using VP2 wind sensor.

They claimed they ran wind tunnel test which I know wasn't true if they had they would of seen problem immediately.
It wasn't just me but multiple people with VP2 and Vue consoles were seeing different wind speeds.
 
Long story short they did a Vue firmware update, not long after and first one in years (how convenient).
With no mention in update but it fixed issue. I almost didn't bother with it but someone said it fixed the wind speed issue and sure enough it did.
Now both Vue and Vp2 consoles follow each other perfectly.
Last year I pointed out to them that temperatures could be adjusted in the VP2 console (via offsets) without a corresponding change in relative humidity and they acted like that was news to them (on a 12 year old console).
I'm confused, changing the temp does effect the RH, why would you not want the offset to effect it?

No, that's what I'm saying. Currently, changing the temperature via an offset in the console does NOT affect the RH. That's not correct. I pointed it out and Keystone Cops ensued over at Davis HQ.
You gotta be kidding me........ :shock:   I didn't even check for such a thing, that's an absolute assumable! Talk about wx 101...:roll:

Offline WheatonRon

  • Forecaster
  • *****
  • Posts: 1018
    • WUnderground
Brett Lane head of Davis support as it relates to the VP2, just acknowledged my email to him this morning where I pointed out the “gist” of this thread. Next steps? Who knows.
Davis VP2 with SHT31 (3 complete VP2 systems—2 with a daytime fan and 1 that has a 24 hour fan); CWOP--CW5020, FW3075 and FW4350; WU--KILWHEAT17, KILWHEAT36 and KILWHEAT39; WeatherCloud.net; CoCoRaHS--IL-DP-132; and Weatherlink 2.0

Offline ValentineWeather

  • Forecaster
  • *****
  • Posts: 5854
    • LiVE WEATHER
This temp offset not changing humidity is interesting. I need to test on my Envoy8x.


Brett Lane head of Davis support as it relates to the VP2, just acknowledged my email to him this morning where I pointed out the “gist” of this thread. Next steps? Who knows.
'
Good deal Ron.
Randy

Offline ValentineWeather

  • Forecaster
  • *****
  • Posts: 5854
    • LiVE WEATHER
Yep you can change temperature and humidity doesn't budge...Very odd.
Cumulus does move however for those using 3rd party software.

Humm.... this just have me an idea. Many of us had to subtract .9 with new SHT31's and older transmitters. I did it using Davis software maybe I need to do it using the Cumulus software and humidity will adjust.
 [ You are not allowed to view attachments ]
Randy

Offline openvista

  • Forecaster
  • *****
  • Posts: 453
    • marquetteweather.com
Humm.... this just have me an idea. Many of us had to subtract .9 with new SHT31's and older transmitters. I did it using Davis software maybe I need to do it using the Cumulus software and humidity will adjust.
 [ You are not allowed to view attachments ]

Well, in that conversation with Davis it came out that the humidity is calculated by the SIM transmitter, NOT the console. The older SIMs add 0.9F to the temp which is why you need to subtract it out at the console. But the humidity is pinned to the ACTUAL temperature measured by the sensor before the SIM adds the 0.9F. So, in other words, you don't need to worry about the humidity if you only have a temperature offset to compensate for an older SIM. It should be correct.

You can verify that by plugging the adjusted temp and humidity into a dew point calculator and see if it matches the dew point. Or you can take the dew point and temp and find the humidity. Either way. I've done it many times on my older SIM. It always matches up. I also have a newer SIM that doesn't need the offset and that humidity verifies as well.

Now, the problem is that if you have any additional temperature offset (say you discovered the temperature sensor has a bias of some sort) then your humidity and dew point and heat index WILL be off and there's no good way to fix it.
« Last Edit: July 10, 2018, 09:07:38 PM by openvista »
Davis Vantage Pro2 AC FARS | https://marquetteweather.com | EW7933

Offline ValentineWeather

  • Forecaster
  • *****
  • Posts: 5854
    • LiVE WEATHER
Okay makes sense I think.  #-o
Randy

Offline DaleReid

  • Forecaster
  • *****
  • Posts: 1492
    • Weather at Eau Claire, WI
Why the 0.9 degree 'correction' in the SIM?
ECWx.info
&
ECWx.info/t/index.php

Offline ValentineWeather

  • Forecaster
  • *****
  • Posts: 5854
    • LiVE WEATHER
Heard back from Ryan he doesn't think there is an issue with Davis covering for replacements.
If you purchased through him send defective units back.
As far as solution which I would rather have than keep replacing nothing yet but hopefully it will now get addressed as they say the ball is rolling with appropriate people notified.

This may not be as open closed as expected for replacements. I told Ryan I'm okay with replacing sensor myself before summer hits but would like Davis to dig into issue. Surprised they didn't already see it. I was looking at their station in Hayward Cal. yesterday afternoon and dewpoint was running +2° over airport about what I see all the time here also.

Anyway the questions Davis is asking Ryan and myself. I understand they need to start somewhere but WOW. Had I known I would of kept better track and logged everything which I didn't . Like I previously said I just used Cumulus to correct issue.
Randy

Offline openvista

  • Forecaster
  • *****
  • Posts: 453
    • marquetteweather.com
Why the 0.9 degree 'correction' in the SIM?

If you're asking why does an older SIM (pre 2017?) add 0.9F to the temp, I'm not exactly sure. I just know that it does and Davis instructs owners to subtract it out with an offset in the console.
Davis Vantage Pro2 AC FARS | https://marquetteweather.com | EW7933

Offline openvista

  • Forecaster
  • *****
  • Posts: 453
    • marquetteweather.com
This may not be as open closed as expected for replacements.

I kinda figured. Sounds like Ryan has started having second thoughts? If Davis won't accept the returns then he'd be exposing his business to significant losses potentially.

C'mon Davis. Replacing defective sensors will create a heap of good will.
Davis Vantage Pro2 AC FARS | https://marquetteweather.com | EW7933

Offline Jim's Weather

  • Forecaster
  • *****
  • Posts: 329
So my email has been getting dinged that there's been replies to the SHT-31 thread.  :lol: My thoughts on the SHT-31 have already been expressed. I have not read this entire thread but I want to address the topic of humidity sensor replacement. In the professional world, it is standard practice to replace electronic humidity sensors at a minimum of every 24 months. My standard practice has been every 18 months. For the SHT-31, I would recommend in humid environments, every 9-12 months minimum. Dry environments, every 12-15 months. Extreme dry environments every 15-18 months.

I only approve of two ways for checking humidity sensor accuracy. Using a humidity sensor calibration chamber is my #1 choice. My #2 choice is a sling psychrometer out in the field. It's very important to have a matched set of thermometers in the psychrometer. It's also important to have the psychrometer certification tested so you know precisely what the temperature correction is on the dry bulb and wet bulb thermometers. If you break one of the thermometers, you need to replace both with another matched set. Also, you need to use the psychrometer correctly. I'm amazed when I'm at a station and see the tech using the sling incorrectly. I've seen them sling with the sling in direct sunlight, only 6 inches away from their body, using tap water on the wick, etc.  :lol:

As far as using another electronic humidity sensor to check another one? I wouldn't.
« Last Edit: July 11, 2018, 11:34:45 AM by kcidwx »
Meteorological Technician
NWS Certified Aviation Weather Observer
B.S. Meteorology

Offline ValentineWeather

  • Forecaster
  • *****
  • Posts: 5854
    • LiVE WEATHER
So my email has been getting dinged that there's been replies to the SHT-31 thread.  :lol: My thoughts on the SHT-31 have already been expressed. I have not read this entire thread but I want to address the topic of humidity sensor replacement. In the professional world, it is standard practice to replace electronic humidity sensors at a minimum of every 24 months. My standard practice has been every 18 months. For the SHT-31, I would recommend in humid environments, every 9-12 months minimum. Dry environments, every 12-15 months. Extreme dry environments every 15-18 months.

I only approve of two ways for checking humidity sensor accuracy. Using a humidity sensor calibration chamber is my #1 choice. My #2 choice is a sling psychrometer out in the field. It's very important to have a matched set of thermometers in the psychrometer. It's also important to have the psychrometer certification tested so you know precisely what the temperature correction is on the dry bulb and wet bulb thermometers. If you break one of the thermometers, you need to replace both with another matched set. Also, you need to use the psychrometer correctly. I'm amazed when I'm at a station and see the tech using the sling incorrectly. I've seen them sling with the sling in direct sunlight, only 6 inches away from their body, using tap water on the wick, etc.  :lol:

As far as using another electronic humidity sensor to check another one? I wouldn't.

Appreciate your comment kcidwx and yes your old review of the SHT31 and thoughts came up as some are realizing  how fast they are going high. I'm taking the advice on change out and do highly recommend others do the same being you are the expert on the subject.

Now we have Davis Inst. involved not sure this will go anywhere. They are rather defensive of their products as they should be.

« Last Edit: July 11, 2018, 12:42:32 PM by ValentineWeather »
Randy

Offline jgentry

  • Forecaster
  • *****
  • Posts: 440
So my email has been getting dinged that there's been replies to the SHT-31 thread.  :lol: My thoughts on the SHT-31 have already been expressed. I have not read this entire thread but I want to address the topic of humidity sensor replacement. In the professional world, it is standard practice to replace electronic humidity sensors at a minimum of every 24 months. My standard practice has been every 18 months. For the SHT-31, I would recommend in humid environments, every 9-12 months minimum. Dry environments, every 12-15 months. Extreme dry environments every 15-18 months.

I only approve of two ways for checking humidity sensor accuracy. Using a humidity sensor calibration chamber is my #1 choice. My #2 choice is a sling psychrometer out in the field. It's very important to have a matched set of thermometers in the psychrometer. It's also important to have the psychrometer certification tested so you know precisely what the temperature correction is on the dry bulb and wet bulb thermometers. If you break one of the thermometers, you need to replace both with another matched set. Also, you need to use the psychrometer correctly. I'm amazed when I'm at a station and see the tech using the sling incorrectly. I've seen them sling with the sling in direct sunlight, only 6 inches away from their body, using tap water on the wick, etc.  :lol:

As far as using another electronic humidity sensor to check another one? I wouldn't.

Thank you for your input!  I’m planning on getting one of these since I don’t have a humidity chamber. Lol

http://www.iisusa.com/mobile/Psychro-Dyne-Battery-Operated-Psychrometer-Fahrenheit-BLACK-ORGANIC/productinfo/22014/

The NWS in Alabama tells me that they change out their sensor every 18 months. Then again, it’s dog-gone humid here.

Since we know that the SHT-31 has a wet bias, how does it performs in terms of measuring air temperature compared to a calibrated standard?
Davis Vantage Pro2. SHT-75. WU: KXALJEMI2 & KALTHORS2. CWOP/APRS: C6353 & E6358

  

Offline jgentry

  • Forecaster
  • *****
  • Posts: 440
Unfortunately the Sensirion SHT-31 is (currently) the ONLY sensor compatible with our Davis units...unless you WANT to go BACK to the older SHT-11 or SHT-15 sensors.

Because of the communication change inherent with newer sensors, the SHT-31 is an "end-of-the-line" device, the last of its kind until/if DAVIS comes up with a new ISS.

I’m afraid you’re correct
Davis Vantage Pro2. SHT-75. WU: KXALJEMI2 & KALTHORS2. CWOP/APRS: C6353 & E6358

  

Offline WheatonRon

  • Forecaster
  • *****
  • Posts: 1018
    • WUnderground
So my email has been getting dinged that there's been replies to the SHT-31 thread.  :lol: My thoughts on the SHT-31 have already been expressed. I have not read this entire thread but I want to address the topic of humidity sensor replacement. In the professional world, it is standard practice to replace electronic humidity sensors at a minimum of every 24 months. My standard practice has been every 18 months. For the SHT-31, I would recommend in humid environments, every 9-12 months minimum. Dry environments, every 12-15 months. Extreme dry environments every 15-18 months.

I only approve of two ways for checking humidity sensor accuracy. Using a humidity sensor calibration chamber is my #1 choice. My #2 choice is a sling psychrometer out in the field. It's very important to have a matched set of thermometers in the psychrometer. It's also important to have the psychrometer certification tested so you know precisely what the temperature correction is on the dry bulb and wet bulb thermometers. If you break one of the thermometers, you need to replace both with another matched set. Also, you need to use the psychrometer correctly. I'm amazed when I'm at a station and see the tech using the sling incorrectly. I've seen them sling with the sling in direct sunlight, only 6 inches away from their body, using tap water on the wick, etc.  :lol:

As far as using another electronic humidity sensor to check another one? I wouldn't.

Appreciate you comment kcidwx and yes your old review of the SHT31 and thoughts came up as some are realizing  how fast they are going high. I'm taking your advice on change out and do highly recommend others do the same being you are the expert on the subject.

Now we have Davis Inst. involved not sure this will go anywhere. They are rather defensive of their products as they should be.

So it appears the life span of a SHT31 is at most 24 months to achieve the desired accuracy. I just reread the User Guide to my Integrated Sensor Suite with the 24 hour fan I just bought from Ryan Wilhour, and nowhere in that document (Rev B, 5/6/15) does it recommend any maintenance associated with Temperature-Humidity sensor. The User Guide talks about cleaning the ISS, the rain collector, changing the 24 hour fan batteries, maintaining the anemometer, etc., but nothing about maintaining or changing this sensor. This is where the problem lies. Davis should have stated in this manual to change the temperature-humidity sensor periodically based on the guidance provided by kcidwx based on where you live, and everybody would be happy, and this thread as it relates to the SHT31, would not exist.
« Last Edit: July 11, 2018, 03:15:47 PM by WheatonRon »
Davis VP2 with SHT31 (3 complete VP2 systems—2 with a daytime fan and 1 that has a 24 hour fan); CWOP--CW5020, FW3075 and FW4350; WU--KILWHEAT17, KILWHEAT36 and KILWHEAT39; WeatherCloud.net; CoCoRaHS--IL-DP-132; and Weatherlink 2.0

Offline Jim's Weather

  • Forecaster
  • *****
  • Posts: 329
jgentry, I never tested the temperature side of the SHT-31. So, I can't speak to its accuracy. The one thing that bothers me with the way Davis uses the SHT-31 is the aspiration. We don't aspirate humidity sensors. The ASOS DTS1 is not aspirated. You want to keep that sensor clean or it can go out of calibration. So putting the SHT-31 in the FARS and drawing the dirty outside air across it bothers me. I don't know how good the filtration is on that FARS or how easily dust and dirt can get to it by other means.
Meteorological Technician
NWS Certified Aviation Weather Observer
B.S. Meteorology

Offline openvista

  • Forecaster
  • *****
  • Posts: 453
    • marquetteweather.com
Please recall earlier in the thread where I quoted from The Weather Observer's Handbook in which Stephen Burt wrote in 2012 that the cheapest capacitive humidity sensors had 5% or more movement per year. Is that what Davis puts in the VP2? The cheapest sensor? I would think such a sensor would be relegated to the <$200 weather stations sold by their competitors.  Here we are all these years later and we should just accept that the humidity will be, what, 8-10% off in 2 years? How does that square with the vendor claiming 0.25% drift per year?
 
Something isn't right here. I'm not saying it's entirely Davis' fault. But the SHT31s have been in production since early 2016 and this is the first anyone's heard of the sensor package being a disposable maintenance item.

Until specs are changed, this is a defective part and should be subject to warranty claims.

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

Offline openvista

  • Forecaster
  • *****
  • Posts: 453
    • marquetteweather.com
jgentry, I never tested the temperature side of the SHT-31. So, I can't speak to its accuracy. The one thing that bothers me with the way Davis uses the SHT-31 is the aspiration. We don't aspirate humidity sensors. The ASOS DTS1 is not aspirated. You want to keep that sensor clean or it can go out of calibration. So putting the SHT-31 in the FARS and drawing the dirty outside air across it bothers me. I don't know how good the filtration is on that FARS or how easily dust and dirt can get to it by other means.

The filter cap on the sensor only filters out the largest particles. Fine dust and VOCs can easily get through that filter. That being said, such deposits should act to negatively bias the sensor, thereby lowering humidity values, should it not? The foreign particles block the adsorption process. Instead, what we're dealing with is a sensor that gets rapidly wetter over time.

When I disassembled my FARS a couple days ago, the filter was dirty. I wouldn't say it was plugged, but still plenty of black particulates on the screen. Regardless, that sensor was running wet by 8%.
Davis Vantage Pro2 AC FARS | https://marquetteweather.com | EW7933

Offline openvista

  • Forecaster
  • *****
  • Posts: 453
    • marquetteweather.com
I should point out that after the screen was cleaned and the offending sensor was re-installed in a passive shelter, the wet bias has not lessened.
Davis Vantage Pro2 AC FARS | https://marquetteweather.com | EW7933

Offline ValentineWeather

  • Forecaster
  • *****
  • Posts: 5854
    • LiVE WEATHER
Please recall earlier in the thread where I quoted from The Weather Observer's Handbook in which Stephen Burt wrote in 2012 that the cheapest capacitive humidity sensors had 5% or more movement per year. Is that what Davis puts in the VP2? The cheapest sensor? I would think such a sensor would be relegated to the <$200 weather stations sold by their competitors.  Here we are all these years later and we should just accept that the humidity will be, what, 8-10% off in 2 years? How does that square with the vendor claiming 0.25% drift per year?
 
Something isn't right here. I'm not saying it's entirely Davis' fault. But the SHT31s have been in production since early 2016 and this is the first anyone's heard of the sensor package being a disposable maintenance item.

Until specs are changed, this is a defective part and should be subject to warranty claims.

I agree, they surely are not within claimed specs of 2% and all should get free replacements every 6 months until a solution is found.
But getting Davis to agree is another thing.. :evil:

The sht31 is only a couple dollars if purchased separately without the hardware included so yes they are a bottom feeder when it comes to price.

Digikey sells them bulk 1000 chips at $3.46 each.
« Last Edit: July 11, 2018, 01:37:20 PM by ValentineWeather »
Randy