Author Topic: The Reliability of the SHT-31 Humidity Sensor & What Psychrometer Should I Buy?  (Read 107438 times)

0 Members and 1 Guest are viewing this topic.

Offline galfert

  • Global Moderator
  • Forecaster
  • *****
  • Posts: 6822
Hey everyone,
It is this guy's fault... Richard Anderson design engineer at Davis. Anyone know how to reach him? Well in all seriousness it probably isn't truly his fault alone. But we have a name at Davis. Maybe he has some answers.

https://www.davisinstruments.com/blog/meet-davis-design-engineer-richard-anderson/

From the link above it says,
Quote
     he has applied over 40,000 hours toward perfecting our weather products and components. We believe that it is one of the many reasons our weather stations have stood the test of time   

Well I think we all have something to say about that.

Also in the article it says under the picture where he is using a micrometer. (what is he measuring anyway?)
Quote
Precision matters, says Richard. We design our products to the tightest tolerances for consistent manufacturing and assembly, which results in dependable, quality products.   

Really now? Is that so?

Someone should copy that blog page. Davis will probably be taking it down soon. Sorry Richard. We love you. We just want some answers.

« Last Edit: July 28, 2018, 11:19:49 PM by galfert »
Ecowitt GW1000 | Meteobridge on Raspberry Pi
WU: KFLWINTE111  |  PWSweather: KFLWINTE111
CWOP: FW3708  |  AWEKAS: 14814
Windy: pws-f075acbe
Weather Underground Issue Tracking
Tele-Pole

Offline openvista

  • Forecaster
  • *****
  • Posts: 481
    • marquetteweather.com
Hey everyone,
It is this guy's fault... Richard Anderson design engineer at Davis. Anyone know how to reach him? Well in all seriousness it probably isn't truly his fault alone. But we have a name at Davis. Maybe he has some answers.

https://www.davisinstruments.com/blog/meet-davis-design-engineer-richard-anderson/

From the link above it says,
Quote
     he has applied over 40,000 hours toward perfecting our weather products and components. We believe that it is one of the many reasons our weather stations have stood the test of time   

Well I think we all have something to say about that.

I know that his name is on the 1999 patent application for the Davis FARS design. See: https://patents.google.com/patent/US6247360B1/en.

However, he is just a foot soldier. Contacting him isn't going to move the needle as he will have to get in line behind his superiors. 

But if I were him, I wouldn't want my name anywhere near this mess.
« Last Edit: July 28, 2018, 11:30:43 PM by openvista »
Davis Vantage Pro2 AC FARS | https://marquetteweather.com | EW7933

Offline galfert

  • Global Moderator
  • Forecaster
  • *****
  • Posts: 6822
However, he is just a foot soldier. Contacting him isn't going to move the needle as he will have to get in line behind his superiors. 

But if I were him, I wouldn't want my name anywhere near this mess.

I know. I agree. He is just a foot soldier as you said. He is probably as upset as we are. He is most undoubtedly a very smart person, and probably corporate beurocracy has held him back from giving us the system we all really want.

But it was kinda cool to put a face to the person whom probably does have some intimate knowledge of what we are dealing with.
« Last Edit: July 28, 2018, 11:46:53 PM by galfert »
Ecowitt GW1000 | Meteobridge on Raspberry Pi
WU: KFLWINTE111  |  PWSweather: KFLWINTE111
CWOP: FW3708  |  AWEKAS: 14814
Windy: pws-f075acbe
Weather Underground Issue Tracking
Tele-Pole

Offline galfert

  • Global Moderator
  • Forecaster
  • *****
  • Posts: 6822

I know that his name is on the 1999 patent application for the Davis FARS design. See: https://patents.google.com/patent/US6247360B1/en.
.

Maybe I should file for my own FARS design patent. You know one where the temperature sensor is separate from the humidity sensor and only the temperature sensor receives air flow because they are in two separate compartments. And add a heater to the humidity sensor for periodic reconditioning. And add some insulation and distance between compartments so that the heater doesn't affect the temperature sensor. What do you think of that Richard Anderson? Hey don't steal my idea.

« Last Edit: July 29, 2018, 12:03:12 AM by galfert »
Ecowitt GW1000 | Meteobridge on Raspberry Pi
WU: KFLWINTE111  |  PWSweather: KFLWINTE111
CWOP: FW3708  |  AWEKAS: 14814
Windy: pws-f075acbe
Weather Underground Issue Tracking
Tele-Pole

Offline galfert

  • Global Moderator
  • Forecaster
  • *****
  • Posts: 6822
Okay I just came up with neat upgrade idea.

What if someone like Belfryboy or someone that knows how to work with microcontrollers builds a custom sensor replacement.

Key design aspects:
- Uses two SHT31 connected to a microcontroller.

- Purpose of the microcontroller is to switch sensors every 24 hours. The microcontroller intercepts temperature and humidity from one sensor at a time. The microcontroller will then pass along this data down the Davis bus.

- Both sensors are placed in the radiation shield using a new mounting board that creates some separation and insulation from the other. Or maybe you add a second radiation shield and keep them separate that way.

- Both sensors get heaters. But only the inactive sensor bakes for 10 hours and then gets adapted back to ambient humidity for reconditioning completion before becoming active again.

- Sensors alternate between reconditioning and being active every 24 hours.

- Possible extras are to correct for wet bias slope at different temperatures with different degrees.

- Another possible extra is that the heating and humidity reconditioning does not occur unless recent humidity was high.
« Last Edit: July 29, 2018, 12:57:13 AM by galfert »
Ecowitt GW1000 | Meteobridge on Raspberry Pi
WU: KFLWINTE111  |  PWSweather: KFLWINTE111
CWOP: FW3708  |  AWEKAS: 14814
Windy: pws-f075acbe
Weather Underground Issue Tracking
Tele-Pole

Offline SnowHiker

  • Forecaster
  • *****
  • Posts: 352
As Johnd, our reseller friend from the UK, has noted the demand for PWS is not huge so we don’t have a lot of pent-up demand for a better product so unless the Atlas device takes off, is reliable and well built, we don’t have a lot of choices. Frustrating yes, but....
Could be a double-edged sword for Davis.  If the market is small, Davis may have a hard time affording the resources necessary for quick research, testing, development, production, and rectifying the problems with the units already sold and produced.

On the other hand, if the market is that small, Davis can't afford to lose customers, and their reputation, by not correcting the problem.

But considering how doomsday predictions usually go, I'm not going to dump my Davis stock just yet.  (As if I had any.  :lol:)

Offline CW7491

  • Senior Contributor
  • ****
  • Posts: 202
From the side by sides I’ve done, the unmodified 75 runs consistently drier than the Davis-mounted 31 in the midrange as low as humidity in the 20s. (And I stress it is unmodified). That said, it’s not perfect and is still a Sensirion. It struggles to hit 100%, but gets there on occasion. It hits 98 and 99 regularly. It will still exhibit a wet bias if in high humidity for a prolonged period and I’d say it may have a slight wet bias generally, but it compares really closely to my psycho-dyne. I’d love to find a way to get a 31 or 35 that Davis hasn’t touched to work with my VP2, but until then, my experience is having the 75 is markedly superior to the Davis 31. Temp specs are not as good obviously and not exactly a rigorous study, but I’ve tested side by side with an ASOS at -8F and it was right on. Not trying to convince anyone, but just my experience ...

How much of the difference is there between the 75 & psych-dyne?

Usually pretty spot on. Within 1F in dew point. Sometimes 2F. The nice thing about the 75 is you can put it right in the intake of the psychro dyne to verify temp and dew point. I know some others are using the 75 so they may be able to provide more input in terms of temp/hum accuracy of the 75 across the extreme ranges

That’s good!  Have you ever done a comparison between the 75 & 31 in regards to air temp?

How did your 31 compared to the psychro-dyne?

The psychro dyne is dead on next to ASOS for temp and humidity. I’ve never found a difference in temp between the 75 and 31 when I’ve run them side by side. I could maybe do some more tests to get you some numeric answers. The Davis 31 is dead on for temp with the psychrodyne, but high on dewpoint usually by up to 4F, which is consistent with the comparison to ASOS. I have had the 75 next to an ASOS from -8F up to about 90F and it’s always been exactly on. I know the specs aren’t as good, but practically it seems to be a really good performer...

Thanks to some of these guys for going to Davis so they at least know the issue with their sensor. . I purchased the Acurite with the 31 after reading about it in this thread. It is really very good. I wish Davis’ 31 performed this way on the humidity side

« Last Edit: July 29, 2018, 01:30:53 AM by CW7491 »

Offline ValentineWeather

  • Forecaster
  • *****
  • Posts: 6364
    • Valentine Nebraska's Real-Time Weather
Psychro dyne is made by Belfort one of the most trusted names in meteorological instruments.
 
The very best is Assmann psychrometer which is used as reference instrument tool. A little out of my price range however.
Randy

Offline dendrite

  • Forecaster
  • *****
  • Posts: 367
The very best is Assmann psychrometer which is used as reference instrument tool. A little out of my price range however.
I'm more of a leg guy anyway.

Had a decent storm go through yesterday while I wasn't home. You can pick up some pretty cool microscale effects in temp/dew with the FARS, 75, and 1-min intervals. For those unfamiliar with the Davis text files, it's Temp-MaxTemp-MinTemp-RH-Dew

 [ You are not allowed to view attachments ]

RH on the old 75 only maxed out at 96% (31 had only been hitting 94%) last night. I put another coat of conformal coating on the new one this morning so it may not be until later tomorrow that I'll get to install it.

Offline jgentry

  • Forecaster
  • *****
  • Posts: 533
Psychro dyne is made by Belfort one of the most trusted names in meteorological instruments.
 
The very best is Assmann psychrometer which is used as reference instrument tool. A little out of my price range however.

If I’m not mistaken, I’m the NWS ASOS training manual, they say to use the psychro-dyne to verify the ASOS’ DP readings.

“There is a procedure used by the ASOS Technician to check the validity of the ASOS ambient/dew point temperatures by comparing the values with the Psychron Model 566-2 psychrometer.  A measurement is considered valid if the temperature difference is within ±5 F.  dew point validity is determined by values within tolerances based on the temperature/dew point depression and if the temperature falls above or below 32 F.”

https://training.weather.gov/nwstc/DATAACQ/d.ASOShuman/ASOSTemp.htm
« Last Edit: July 29, 2018, 08:06:38 AM by jgentry »
Davis Vantage Pro2 & WeatherFlow Tempest. WU: KXALJEMI2, KALJEMIS7, KFLPANAM363 & KALTHORS2. CWOP/APRS: C6353 & E6358

Offline ValentineWeather

  • Forecaster
  • *****
  • Posts: 6364
    • Valentine Nebraska's Real-Time Weather
Psychro dyne is made by Belfort one of the most trusted names in meteorological instruments.
 
The very best is Assmann psychrometer which is used as reference instrument tool. A little out of my price range however.

If I’m not mistaken, I’m the NWS ASOS training manual, they say to use the psychro-dyne to verify the ASOS’ DP readings.

https://training.weather.gov/nwstc/DATAACQ/d.ASOShuman/ASOSTemp.htm

Could be...WMO according to one of the humidity studies uses the Assmann as standard...They come with a NIST certificate of 3 point calibration and corrections for thermometers. Accuracy 1% humidity .1C temperature.
Randy

Offline CW7491

  • Senior Contributor
  • ****
  • Posts: 202
Not sure if this has been sent before, but here is the ASOS tech manual I found online. Chapter 5 discusses verification procedures for temperature and dewpoint (5.5.2.4)...

https://faaco.faa.gov/index.cfm/attachment/dload/?filename=/Appendix_C_-_ASOS_Tech_Manual.pdf
« Last Edit: July 29, 2018, 08:30:31 AM by CW7491 »

Offline openvista

  • Forecaster
  • *****
  • Posts: 481
    • marquetteweather.com
Had a decent storm go through yesterday while I wasn't home. You can pick up some pretty cool microscale effects in temp/dew with the FARS, 75, and 1-min intervals. For those unfamiliar with the Davis text files, it's Temp-MaxTemp-MinTemp-RH-Dew

Hmmm... a 12% swing down in humidity in 2 minutes (3:46-3:48) while it's raining? And the temp only changed 0.3F? That would drive me nuts.  I already had to write a routine to compute the 5 minute rolling average for humidity & dew point read by the 31 in the FARS so that I wasn't getting wild fluctuations in data sent to the website and CWOP (NWS prefers smoothed data).

To the extent it's possible, I don't want to be reporting microscale effects. People come to my site or check one of my data feeds to see what the weather is like in the city center, not to learn when a butterfly has flapped its wings.

As always, your mileage may vary, but maybe the 75 isn't for me.
Davis Vantage Pro2 AC FARS | https://marquetteweather.com | EW7933

Offline dendrite

  • Forecaster
  • *****
  • Posts: 367
Had a decent storm go through yesterday while I wasn't home. You can pick up some pretty cool microscale effects in temp/dew with the FARS, 75, and 1-min intervals. For those unfamiliar with the Davis text files, it's Temp-MaxTemp-MinTemp-RH-Dew

Hmmm... a 12% swing down in humidity in 2 minutes (3:46-3:48) while it's raining? And the temp only changed 0.3F? That would drive me nuts.  I already had to write a routine to compute the 5 minute rolling average for humidity & dew point read by the 31 in the FARS so that I wasn't getting wild fluctuations in data sent to the website and CWOP (NWS prefers smoothed data).

To the extent it's possible, I don't want to be reporting microscale effects. People come to my site or check one of my data feeds to see what the weather is like in the city center, not to learn when a butterfly has flapped its wings.

As always, your mileage may vary, but maybe the 75 isn't for me.
You can see it was more stable before and after the storm passed. When that gust front first hits it becomes a pretty dynamic situation. I come from the meteorological research side of things so I find all of these brief fluctuations fascinating. I don't think someone coming to my site is worried about a few % RH during a storm when they see a rain rate of 7"/hr and 1.25" through the tipper in a half hour, but that's just me.

Offline openvista

  • Forecaster
  • *****
  • Posts: 481
    • marquetteweather.com
Had a decent storm go through yesterday while I wasn't home. You can pick up some pretty cool microscale effects in temp/dew with the FARS, 75, and 1-min intervals. For those unfamiliar with the Davis text files, it's Temp-MaxTemp-MinTemp-RH-Dew

Hmmm... a 12% swing down in humidity in 2 minutes (3:46-3:48) while it's raining? And the temp only changed 0.3F? That would drive me nuts.  I already had to write a routine to compute the 5 minute rolling average for humidity & dew point read by the 31 in the FARS so that I wasn't getting wild fluctuations in data sent to the website and CWOP (NWS prefers smoothed data).

To the extent it's possible, I don't want to be reporting microscale effects. People come to my site or check one of my data feeds to see what the weather is like in the city center, not to learn when a butterfly has flapped its wings.

As always, your mileage may vary, but maybe the 75 isn't for me.
You can see it was more stable before and after the storm passed. When that gust front first hits it becomes a pretty dynamic situation. I come from the meteorological research side of things so I find all of these brief fluctuations fascinating. I don't think someone coming to my site is worried about a few % RH during a storm when they see a rain rate of 7"/hr and 1.25" through the tipper in a half hour, but that's just me.

Ah, gotcha. I didn't know how to interpret all the columns. Missed the gust front.  :oops: Yes, before and after the dp seems pretty stable.

Where I really get fluctuations is in the summer when there's a synoptic, upper level pattern bringing warm/moist air in from the S or SW but not enough boundary layer wind flow to prevent a competing lake breeze from Lake Superior. Tug of war ensues where you get cool/drier air from the Lake vs maritime tropical air from the Gulf. If you walk around near the ISS, you can feel the cool pockets and warm pockets as they drift about. It's wild! The local NWS seems to really hate those days because it's damn near impossible to nail the timing or even the development of the lake breeze. So either they undershoot or overshoot the high (usually undershoot because the breeze loves to show up late). Oh and I forgot to throw in downsloping wind effects as air descends from about 1200 to 600 feet in just a few miles before hitting the lake. Trying to interpret or predict temps, dew points and winds in the U.P. can be a sadomasochistic practice sometimes.
Davis Vantage Pro2 AC FARS | https://marquetteweather.com | EW7933

Offline ValentineWeather

  • Forecaster
  • *****
  • Posts: 6364
    • Valentine Nebraska's Real-Time Weather
I can't remember if we confirmed the SHT15 also had high bias. May have I remember thinking it was high and using cumulus to remove some %, so just to make sure I've started testing the reconditioned sht15. Been in saturated high humidities since 9pm yesterday 13 hours now 90% or greater.

Problem I'm getting into here in Nebraska is climate changes so high humidity and high heat bias may almost be gone going into August. Seems it does dry out some we've had over 17" rain last 3 months with 23 events over .10" so it doesn't all come at once like Houston where much of family lives they get 10" rain events and then long spells of nothing.
 [ You are not allowed to view attachments ]

Randy

Offline WheatonRon

  • Forecaster
  • *****
  • Posts: 1237
    • WUnderground
“Davis Instruments is an established company with an outstanding product line and an excellent team.  Davis weather stations are sold around the world and have a leading position in the markets we serve, with foreign sales accounting for more than a third of our revenue.  Our ever-growing weather product line includes products that serve direct consumers, government, agriculture, construction, and education markets.  Our 100 employees design, engineer, manufacture, and distribute our products from our Hayward, California headquarters.”

Davis is looking for a sales manager per its website. A copy of the introduction to that position is quoted above. Folks that are looking for a position reading this Forum thread may wish to throw his or her hat in the ring to help Davis address its SHT31 issues. Maybe this new sales manager can motivate the design team to address this issue as well as pushing the company to launch a superior product to the VP2!

However, it also shows the reality of the problem. Davis has only 100 employees, so improvements to its product line are really limited.

Fellow posters--please don't make my posting a topic for discussion--I only posted it to show the reality of the situation.
« Last Edit: July 29, 2018, 12:54:53 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 openvista

  • Forecaster
  • *****
  • Posts: 481
    • marquetteweather.com
Found some info on RAWS stations with regard to temp/humidity.

First, and most helpful, is this page: https://raws.nifc.gov/standards-guidelines. If you click on the "Noncompliance Report" you'll get an active list of stations that are overdue for their yearly maintenance. Be sure to check your local RAWS on that if you use it for calibration purposes. For example, I found one that is about 35 miles from me hasn't had a visit in over 3 years!

Also, the current sensor is made by FTS and it's the THS-03 (https://ftsinc.com/meteorology/products/sensors/temperature-and-humidity-sensor/). Unfortunately, there's a lack of info on the specs for humidity. In a 2015 product brochure, they list the specs as +/-2% 0-60C (https://www.awra.org/memberservices/brochures/FTS_ProductCatalog_2015.pdf. No further info available.

Also, according to this page - https://raws.nifc.gov/stationassets/relative-humidityair-temperature/relative-humidityair-temperature-wbracket - there's mention of Vaisala, but those sensors seem to be no longer supported.
Davis Vantage Pro2 AC FARS | https://marquetteweather.com | EW7933

Offline galfert

  • Global Moderator
  • Forecaster
  • *****
  • Posts: 6822

This is exactly what Davis hates to hear.  If you think about it this thread will live on forever and it's not great for Davis advertising. You aren't the only person I'm sure with same thoughts about holding off purchase.
Best thing Davis can do is address and fix problem quickly or thread will be 100 pages long eventually and come up on every Google search for years.

I suppose you are right. I just noticed this thread is the second longest ever in number of replies on wxforum.net.

767 replies and counting.

I'm sure it will soon surpass the 1448 replies of the number one thread, "One Thing About New Mexico."
https://www.wxforum.net/index.php?topic=8240.0

This thread will forever be infamous.

Davis,
The ball is in your court. You have the opportunity to write the last chapter.
Ecowitt GW1000 | Meteobridge on Raspberry Pi
WU: KFLWINTE111  |  PWSweather: KFLWINTE111
CWOP: FW3708  |  AWEKAS: 14814
Windy: pws-f075acbe
Weather Underground Issue Tracking
Tele-Pole

Offline openvista

  • Forecaster
  • *****
  • Posts: 481
    • marquetteweather.com
I can't remember if we confirmed the SHT15 also had high bias.

If you're referring to the Davis labelled 15, yes. Based on my tests and others' reports in this thread, it seems to develop a wet bias with time, at least, although that could be said of most capacitive humidity sensors. I don't know if any 15s start with a wet bias. I didn't test mine when it was new.

I would note that my 15's bias isn't as extreme as my defective 31 and seems to fluctuate more. Occasionally, it runs close to specs while at other times it's off the mark (5+%). In fairness, since I've been testing it, the temp hasn't risen above 80 so I don't have a full picture, but at lower temps it seems more accurate, generally, when measuring humidity. It's also about twice as old as my bad 31 so, by rights, it should be worse, not better.

Is it possible that Davis changed the way they package or attach the sensor in the last few years making the 31 especially susceptible? Maybe the 31 is more sensitive to chemical contamination for some reason?
Davis Vantage Pro2 AC FARS | https://marquetteweather.com | EW7933

Offline WxLover16

  • Forecaster
  • *****
  • Posts: 479
The very best is Assmann psychrometer which is used as reference instrument tool. A little out of my price range however.
I'm more of a leg guy anyway.

Had a decent storm go through yesterday while I wasn't home. You can pick up some pretty cool microscale effects in temp/dew with the FARS, 75, and 1-min intervals. For those unfamiliar with the Davis text files, it's Temp-MaxTemp-MinTemp-RH-Dew

 [ You are not allowed to view attachments ]

RH on the old 75 only maxed out at 96% (31 had only been hitting 94%) last night. I put another coat of conformal coating on the new one this morning so it may not be until later tomorrow that I'll get to install it.

Those are some wild swings. I actually went in and took the filter off my 31. I love the faster response time, tho my only main concern is the fan blowing all that air straight to the humidity sensor part. I wonder how long before it starts to dry out or just go bonkers, anyone know? I was very reluctant to for that reason, but what's done is done. I will enjoy my fast response time  \:D/
Davis Wireless VP2 SHT31 24hr 24CFM FARS

Offline WxLover16

  • Forecaster
  • *****
  • Posts: 479
The very best is Assmann psychrometer which is used as reference instrument tool. A little out of my price range however.
I'm more of a leg guy anyway.

Had a decent storm go through yesterday while I wasn't home. You can pick up some pretty cool microscale effects in temp/dew with the FARS, 75, and 1-min intervals. For those unfamiliar with the Davis text files, it's Temp-MaxTemp-MinTemp-RH-Dew

 [ You are not allowed to view attachments ]

RH on the old 75 only maxed out at 96% (31 had only been hitting 94%) last night. I put another coat of conformal coating on the new one this morning so it may not be until later tomorrow that I'll get to install it.

Those are some wild swings. I actually went in and took the filter off my 31 yesterday evening. I love the faster response time, tho my only main concern is the fan blowing all that air straight to the humidity sensor part. I wonder how long before it starts to dry out or just go bonkers, anyone know? I was very reluctant to for that reason, but what's done is done. I will enjoy my fast response time  \:D/
Davis Wireless VP2 SHT31 24hr 24CFM FARS

Offline dendrite

  • Forecaster
  • *****
  • Posts: 367
The very best is Assmann psychrometer which is used as reference instrument tool. A little out of my price range however.
I'm more of a leg guy anyway.

Had a decent storm go through yesterday while I wasn't home. You can pick up some pretty cool microscale effects in temp/dew with the FARS, 75, and 1-min intervals. For those unfamiliar with the Davis text files, it's Temp-MaxTemp-MinTemp-RH-Dew

 [ You are not allowed to view attachments ]

RH on the old 75 only maxed out at 96% (31 had only been hitting 94%) last night. I put another coat of conformal coating on the new one this morning so it may not be until later tomorrow that I'll get to install it.

Those are some wild swings. I actually went in and took the filter off my 31. I love the faster response time, tho my only main concern is the fan blowing all that air straight to the humidity sensor part. I wonder how long before it starts to dry out or just go bonkers, anyone know? I was very reluctant to for that reason, but what's done is done. I will enjoy my fast response time  \:D/
My temp is still spot on. People with filters are already seeing humidity drift anyway so I don't really worry about it. We all do this for ourselves so whatever gives you the most enjoyment I say go for it. We all want the best possible accuracy, but there's so many sensors with different specs and different ways to come up with an "official" 2m ambient temp that there's always going to be some discrepancies...different shielding, different specs, different response, different smoothing/averaging algorithms, etc.

I may gobble up a few 75s on Newark or Mouser before they become too hard to find. For ~$30 each, if they croak or drift too much in a year I could just swap another one in. By then we'll have a pin 35 that maybe we can adapt to a VP2.

Offline SnowHiker

  • Forecaster
  • *****
  • Posts: 352
We all want the best possible accuracy, but there's so many sensors with different specs and different ways to come up with an "official" 2m ambient temp that there's always going to be some discrepancies...different shielding, different specs, different response, different smoothing/averaging algorithms, etc.
Yep, and as we strive for more accuracy we give up a little in consistency.  If I can get accuracy that at least approaches that of historical records taken with manual equipment in Stevenson screens without spending hundreds extra to get a fraction of a degree accuracy, I'm happy.  I just want to know what it feels like outside and keep records of highs and lows and such.  Of course my "official" highs and lows may be slightly different than a neighbor with high-end equipment, but at least I'm comparing apples to apples, blissfully ignorant of any sensor drift in my uncalibrated instruments anyway.  As long as it's in the ballpark with other stations in the region I don't worry.  I also have a spare ISS and other equipment I can compare it to.

But then, I'm kind of old fashioned and tend to like simple, cheap, reliable things.  The fancier things are, the more goes wrong with them, causing more inconsistencies, in my view.  Or maybe I just use that as an excuse to be cheap.  :-P

Offline jgentry

  • Forecaster
  • *****
  • Posts: 533
The very best is Assmann psychrometer which is used as reference instrument tool. A little out of my price range however.
I'm more of a leg guy anyway.

Had a decent storm go through yesterday while I wasn't home. You can pick up some pretty cool microscale effects in temp/dew with the FARS, 75, and 1-min intervals. For those unfamiliar with the Davis text files, it's Temp-MaxTemp-MinTemp-RH-Dew

 [ You are not allowed to view attachments ]

RH on the old 75 only maxed out at 96% (31 had only been hitting 94%) last night. I put another coat of conformal coating on the new one this morning so it may not be until later tomorrow that I'll get to install it.

Those are some wild swings. I actually went in and took the filter off my 31. I love the faster response time, tho my only main concern is the fan blowing all that air straight to the humidity sensor part. I wonder how long before it starts to dry out or just go bonkers, anyone know? I was very reluctant to for that reason, but what's done is done. I will enjoy my fast response time  \:D/
My temp is still spot on. People with filters are already seeing humidity drift anyway so I don't really worry about it. We all do this for ourselves so whatever gives you the most enjoyment I say go for it. We all want the best possible accuracy, but there's so many sensors with different specs and different ways to come up with an "official" 2m ambient temp that there's always going to be some discrepancies...different shielding, different specs, different response, different smoothing/averaging algorithms, etc.

I may gobble up a few 75s on Newark or Mouser before they become too hard to find. For ~$30 each, if they croak or drift too much in a year I could just swap another one in. By then we'll have a pin 35 that maybe we can adapt to a VP2.

You have your new 75 up and running?
Davis Vantage Pro2 & WeatherFlow Tempest. WU: KXALJEMI2, KALJEMIS7, KFLPANAM363 & KALTHORS2. CWOP/APRS: C6353 & E6358