Topic: The Reliability of the SHT-31 Humidity Sensor & What Psychrometer Should I Buy?

[jgentry]

This is the size i cut the sht31 down to compared to a nickel [ You are not allowed to view attachments ]

Gotcha. Thanks!

[CW7491]

I’m interested to hear over time what people have resolved with the SHT31 issues. I know some have said they’ve seen an improvement in the newer Davis 31s. For those folks, do you still see improved performance as it relates to a wet bias and being able to reach 100% humidity in saturation? Have folks just accepted that the issues with the Davis 31 are what they are? Is anyone still using the SHT75 instead? I’m still using the SHT75 and periodically compare my Davis 31s (after baking the sensors etc) and still ultimately run into the same problems. I guess I’m wondering if the lack of activity in this thread over time is more a function of acceptance or a function of sensor improvement that jgentry and jerryg have noted.

[jgentry]

I’m interested to hear over time what people have resolved with the SHT31 issues. I know some have said they’ve seen an improvement in the newer Davis 31s. For those folks, do you still see improved performance as it relates to a wet bias and being able to reach 100% humidity in saturation? Have folks just accepted that the issues with the Davis 31 are what they are? Is anyone still using the SHT75 instead? I’m still using the SHT75 and periodically compare my Davis 31s (after baking the sensors etc) and still ultimately run into the same problems. I guess I’m wondering if the lack of activity in this thread over time is more a function of acceptance or a function of sensor improvement that jgentry and jerryg have noted.

The new Davis SHT-31s finally showed the same ole wet bias issues. I thought they were resolved but I ended up being wrong. My SHT-31s did not reach to 100%. It only reached to 99%. I’m just forced to accept the errors. The WeatherFlow Tempest SHT-31 doesn’t have this problem except for one (no wet bias but started topping off at 95% after a few months of use). 

The way Davis handles and mounts the sensor, it doesn’t take long for it to be out of calibration.

[CW7491]

I’m interested to hear over time what people have resolved with the SHT31 issues. I know some have said they’ve seen an improvement in the newer Davis 31s. For those folks, do you still see improved performance as it relates to a wet bias and being able to reach 100% humidity in saturation? Have folks just accepted that the issues with the Davis 31 are what they are? Is anyone still using the SHT75 instead? I’m still using the SHT75 and periodically compare my Davis 31s (after baking the sensors etc) and still ultimately run into the same problems. I guess I’m wondering if the lack of activity in this thread over time is more a function of acceptance or a function of sensor improvement that jgentry and jerryg have noted.

The new Davis SHT-31s finally showed the same ole wet bias issues. I thought they were resolved but I ended up being wrong. My SHT-31s did not reach to 100%. It only reached to 99%. I’m just forced to accept the errors. The WeatherFlow Tempest SHT-31 doesn’t have this problem except for one (no wet bias but started topping off at 95% after a few months of use). 

The way Davis handles and mounts the sensor, it doesn’t take long for it to be out of calibration.

Thanks jgentry. I was afraid of that …

[jgentry]

I’m interested to hear over time what people have resolved with the SHT31 issues. I know some have said they’ve seen an improvement in the newer Davis 31s. For those folks, do you still see improved performance as it relates to a wet bias and being able to reach 100% humidity in saturation? Have folks just accepted that the issues with the Davis 31 are what they are? Is anyone still using the SHT75 instead? I’m still using the SHT75 and periodically compare my Davis 31s (after baking the sensors etc) and still ultimately run into the same problems. I guess I’m wondering if the lack of activity in this thread over time is more a function of acceptance or a function of sensor improvement that jgentry and jerryg have noted.

The new Davis SHT-31s finally showed the same ole wet bias issues. I thought they were resolved but I ended up being wrong. My SHT-31s did not reach to 100%. It only reached to 99%. I’m just forced to accept the errors. The WeatherFlow Tempest SHT-31 doesn’t have this problem except for one (no wet bias but started topping off at 95% after a few months of use). 

The way Davis handles and mounts the sensor, it doesn’t take long for it to be out of calibration.

Thanks jgentry. I was afraid of that …

No problem.

[ValentineWeather]

Well I have psychro-dyne now which is pretty accurate and a new SHT-31 I always purchase the SI 25 foot 31's due to my custom station setup and I'm finding the dew point is either right on or reading low 1 degree not high. I only use dew point chasing humidity around is a waste of time.  Anyway, the highest I've seen in DP is 71 and the lowest is around 48 so limited where I've checked but I no longer see those crazy high heat indices of old. Yes if you heavy aspirate it moves around a couple of degrees once the ambient gets above about 84 or so but if you don't heavily aspirate with a higher-powered AC fan the spiking movement is less.

Now yes it reads higher than the ASOS but it's correct where the ASOS is actually reading low humidity and actually is a POS itself they are force-feeding this warming by cranking the rural ASOS temperature up so it's hard to trust anything about it. 
This crap started around March of 2017 the local official coop station happens to be run by the radio station owner in Valentine too so everyone pretty much knows what's up with the Valentine KVTN ASOS. Just a short example without getting into full detail about being above freezing during icing events, the ASOS inability to reach full saturation so now the max humidity is 94% the AP ASOS has had 9-100+ days while all the other local Wx stations including the COOP-3 this summer and it goes on and on. Really rock-solid case they are tampering with the temperatures on these FAA weather stations likely inserting a multiplier because the hotter it gets the wider the spread from other thermometers.

[Tensor]

I have been using SHT85 for over a year now. Its accuracy is much better than that of a SHT21 (which has been outdoors for 5 years now). SHT21 has always had a small but not bad wet bias, and the bias has become gradually worse. I also tested HYT221 some time ago and the sensor turned out to be really poor. It lost its calibration in rather mild conditions and has a dry bias of about 10-15%.

Since SHT4x series is coming from Sensirion, I will report here if I happen to buy and test those.

[CW7491]

Well I have psychro-dyne now which is pretty accurate and a new SHT-31 I always purchase the SI 25 foot 31's due to my custom station setup and I'm finding the dew point is either right on or reading low 1 degree not high. I only use dew point chasing humidity around is a waste of time.  Anyway, the highest I've seen in DP is 71 and the lowest is around 48 so limited where I've checked but I no longer see those crazy high heat indices of old. Yes if you heavy aspirate it moves around a couple of degrees once the ambient gets above about 84 or so but if you don't heavily aspirate with a higher-powered AC fan the spiking movement is less.

Now yes it reads higher than the ASOS but it's correct where the ASOS is actually reading low humidity and actually is a POS itself they are force-feeding this warming by cranking the rural ASOS temperature up so it's hard to trust anything about it. 
This crap started around March of 2017 the local official coop station happens to be run by the radio station owner in Valentine too so everyone pretty much knows what's up with the Valentine KVTN ASOS. Just a short example without getting into full detail about being above freezing during icing events, the ASOS inability to reach full saturation so now the max humidity is 94% the AP ASOS has had 9-100+ days while all the other local Wx stations including the COOP-3 this summer and it goes on and on. Really rock-solid case they are tampering with the temperatures on these FAA weather stations likely inserting a multiplier because the hotter it gets the wider the spread from other thermometers.

That’s great you’re getting that kind of performance. I can’t think of a time I’ve had a Sensirion sensor run low on dewpoint compared to the psychro-dyne, let alone a Davis Sensirion sensor. Is that a raw reading from the sensor or are you running a software calibration?

[ValentineWeather]

Raw and not just on 1 sensor I have two running.  Like I said I do the dewpoint calculation only. Make sure you get the exact temp as close to possible too. If it's 66.3 don't put 66.5 in the calculator.

[jgentry]

Well I have psychro-dyne now which is pretty accurate and a new SHT-31 I always purchase the SI 25 foot 31's due to my custom station setup and I'm finding the dew point is either right on or reading low 1 degree not high. I only use dew point chasing humidity around is a waste of time.  Anyway, the highest I've seen in DP is 71 and the lowest is around 48 so limited where I've checked but I no longer see those crazy high heat indices of old. Yes if you heavy aspirate it moves around a couple of degrees once the ambient gets above about 84 or so but if you don't heavily aspirate with a higher-powered AC fan the spiking movement is less.

Now yes it reads higher than the ASOS but it's correct where the ASOS is actually reading low humidity and actually is a POS itself they are force-feeding this warming by cranking the rural ASOS temperature up so it's hard to trust anything about it. 
This crap started around March of 2017 the local official coop station happens to be run by the radio station owner in Valentine too so everyone pretty much knows what's up with the Valentine KVTN ASOS. Just a short example without getting into full detail about being above freezing during icing events, the ASOS inability to reach full saturation so now the max humidity is 94% the AP ASOS has had 9-100+ days while all the other local Wx stations including the COOP-3 this summer and it goes on and on. Really rock-solid case they are tampering with the temperatures on these FAA weather stations likely inserting a multiplier because the hotter it gets the wider the spread from other thermometers.

I don’t have the psychrodyne right now but I was basically getting your results with the new SHT-31s until the latter part of July when finally the upper 70s to lower 80s DP airmass settled in. Then, the bias showed back up.

[CW7491]

I have both a psychro-dyne and an old Belfort 566 and read them to the nearest tenth of a degree as suggested. I’ve had Sensirion sensor raw dewpoint readings match the psychrometers, but I’ve never had them read low. I’ve also noticed that the psychrometers either match ASOS for dewpoint or read a bit high on dewpoint compared to ASOS when I’ve tested them co-located

[ValentineWeather]

I'm just saying what my experience is. Havn't had any high DP's above 71 and don't want any.   I did a check with the psychro-dyne but got distracted these were current snapshots at the time where everyone was at. . I'm not seeing anything but accuracy at the current DP, if anything it came in slightly low again. I attached a snapshot of the station I also run .67 mi east of here. The dewpoint is running lower but here in town over grass it matters. 

 [ You are not allowed to view attachments ]

[CW7491]

I'm just saying what my experience is. Havn't had any high DP's above 71 and don't want any.   I did a check with the psychro-dyne but got distracted these were current snapshots at the time where everyone was at. . I'm not seeing anything but accuracy at the current DP, if anything it came in slightly low again. I attached a snapshot of the station I also run .67 mi east of here. The dewpoint is running lower but here in town over grass it matters. 

 [ You are not allowed to view attachments ]

That’s great news. Do you know how old the sensors are and how long you’ve been using them? Please let us know if you continue to get good performance. Maybe I’ll finally come around to spending another $50 trying to give the Davis 31 another chance …

[ValentineWeather]

I'm just saying what my experience is. Havn't had any high DP's above 71 and don't want any.   I did a check with the psychro-dyne but got distracted these were current snapshots at the time where everyone was at. . I'm not seeing anything but accuracy at the current DP, if anything it came in slightly low again. I attached a snapshot of the station I also run .67 mi east of here. The dewpoint is running lower but here in town over grass it matters. 

 [ You are not allowed to view attachments ]

That’s great news. Do you know how old the sensors are and how long you’ve been using them? Please let us know if you continue to get good performance. Maybe I’ll finally come around to spending another $50 trying to give the Davis 31 another chance …

They are new units only about a month in use, Ryan at Scaled Instruments may have them specially made, only because I haven't seen the 25' cable sold anywhere else.  I couldn't even find them on DI own site. All my stations are custom so I can't use the short cables that are getting shorter lately. The last one I tried I couldn't reach anywhere so had to do a splice.

I've purchased what he had in stock but sure he'll be getting more in. I've got 4, two are backups still in the bags. All the units I've got have the black rubber material I don't know if that matters some come in white. I'll send a pic with stock code number. 

 [ You are not allowed to view attachments ]

[CW7491]

I’m pretty sure the 210609QN is a manufacture/inspection date too, which means they are very recently coming from Davis.

[ValentineWeather]

I don't know if the molding process has changed or maybe the material used or maybe nothing at all has changed other than I know Ryan uses low VOC packaging now. How Davis ships to him no clue.  I ordered two for replacements and noticed dew points were pretty dang good even at 70 degrees so I went ahead and ordered 2 more and that was what was left in stock.

[johnd]

They are new units only about a month in use, Ryan at Scaled Instruments may have them specially made, only because I haven't seen the 25' cable sold anywhere else.  I couldn't even find them on DI own site. All my stations are custom so I can't use the short cables that are getting shorter lately. The last one I tried I couldn't reach anywhere so had to do a splice.

The T/H sensors with a 25ft cable are a standard Davis part (7346.221) and have been on the Davis spares list for a few years now. This is the sensor used in the 6830** T/H sensor with shield IIRC. Of course some online outlets only stock/sell the faster-moving Davis items, but any specialist Davis dealer worldwide should be able to source them. We certainly carry them for UK/EU customers.

** 6830 is the sensor introduced roughly along with Weatherlink Live and with the demise of the 6372/6382 packages. WLL can obviously receive data from any combination of up to 8 fully or partially populated ISS units and so rather than sell 6332 bundled with various other sensors as specific products, Davis just suggest that you buy 6332 (or 6331) plus whatever sensor you want. For T/H sensors a shield is obviously needed, so the relevant sensor parts are 6830 and 6832 (the FARS variant) .

[ValentineWeather]

The search doesn't show the part. I can't find any spares list either. The Davis website since the change a few years back is not user-friendly.  Can't even find a 25 foot anything using search. As far as I'm concerned SI was the only retailer in the US that carried the 25-foot sensor and still is.  If you know of another link it.

[ValentineWeather]

 [ You are not allowed to view attachments ]

[johnd]

TBH Davis's own online listings for spare parts (which is what we're talking about with the .221 T/H sensor) is rather hit and miss and its Search isn't great either. Their Replacement Parts PDF is the definitive listing for readily available parts - not sure that Davis make this public but try a search online for PR67 Replacement Parts 2021 - someone may have uploaded it.  And also try a Google search for eg Davis 7346.221 and you'll see plenty of hits.

[ValentineWeather]

TBH Davis's own online listings for spare parts (which is what we're talking about with the .221 T/H sensor) is rather hit and miss and its Search isn't great either. Their Replacement Parts PDF is the definitive listing for readily available parts - not sure that Davis make this public but try a search online for PR67 Replacement Parts 2021 - someone may have uploaded it.  And also try a Google search for eg Davis 7346.221 and you'll see plenty of hits.

I wondered about that. So they have a parts list they don't publish. The old website had every part they carried.

[johnd]

I wondered about that. So they have a parts list they don't publish. The old website had every part they carried.

I don't know why they don't publish the Replacement Parts PDF - there's nothing obviously confidential about it - just a listing of all the standard spares and US list prices. But there are, give or take, around 200 parts on the list and to an extent you do need to know your way around the list to identify the right part. In the past, the Davis product line-up was much simpler so it was easier to make the website comprehensive. But now, especially with all the EM gateways, nodes and sensors the number of spares has grown considerably. I'm not sure how many casual users would welcome needing to plough through 200 or so options in an online shop to find the right part. But that's just IMO of course.

[ValentineWeather]

  Agree, what's the big secret it can't be helping with sales. The only thing I can think of is like you said they keep adding parts and don't have an in-house website handler anymore.

[johnd]

Actually, I see Ryan does have a downloadable PR67 copy on his website. There is a newer one available now with, I'm afraid, some price rises just announced, but which I haven't analysed in detail as yet, but eyeballing in the +5 to 7% band on average. So if you're considering a Davis station or part, now might be a good time to order before online prices increase. (Guessing September, but obviously dealer-dependent.)

[drew1021]

Well I have psychro-dyne now which is pretty accurate and a new SHT-31 I always purchase the SI 25 foot 31's due to my custom station setup and I'm finding the dew point is either right on or reading low 1 degree not high. I only use dew point chasing humidity around is a waste of time.  Anyway, the highest I've seen in DP is 71 and the lowest is around 48 so limited where I've checked but I no longer see those crazy high heat indices of old. Yes if you heavy aspirate it moves around a couple of degrees once the ambient gets above about 84 or so but if you don't heavily aspirate with a higher-powered AC fan the spiking movement is less.

Now yes it reads higher than the ASOS but it's correct where the ASOS is actually reading low humidity and actually is a POS itself they are force-feeding this warming by cranking the rural ASOS temperature up so it's hard to trust anything about it. 
This crap started around March of 2017 the local official coop station happens to be run by the radio station owner in Valentine too so everyone pretty much knows what's up with the Valentine KVTN ASOS. Just a short example without getting into full detail about being above freezing during icing events, the ASOS inability to reach full saturation so now the max humidity is 94% the AP ASOS has had 9-100+ days while all the other local Wx stations including the COOP-3 this summer and it goes on and on. Really rock-solid case they are tampering with the temperatures on these FAA weather stations likely inserting a multiplier because the hotter it gets the wider the spread from other thermometers.

Totally agree with you regarding ASOS issue. The local site near me,KINT, has been reading higher temperatures than all surrounding ASOS and private stations going on 4 months now. Also their dew point is in my opinion way off, reading low. Max humidity I have seen over the past 4 months is 94%
I have contacted the NWS in Raleigh where their technicians are stationed and was told that they do a field spot check every so often and as long as it falls within 5 degrees it’s considered good. Bullshit I say!! Trust my VP2, you bet.