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

[CW2274]

Oh, BTW, I unplugged my fan for a length of time to see if that would make a difference...none, the 0.5F spread remained unchanged.

[fkapp]

How is humidity variance past few days after sensor went above normal range limit of 90%?
did 5% drift effect start?

Curious as this will indicate how fast humid climate sensors accuracy impacts will start.

[CW2274]

How is humidity variance past few days after sensor went above normal range limit of 90%?
did 5% drift effect start?

I'll assume your question is for me. There is no drift, the sensor is just as "in-line" before as it was after.
I think you're over borrowing some of the hype here.

[fkapp]

Okay. I would say there are enough field data tests on this 60 pager to say hype may be a bit bold.
anyways. good to hear. the SF2 seems to help and the dry climate seems to help.  so could be a frequency of sensor getting exposed above normal range eventually starts it, or maybe  Davis has found way to correct (assumption: as none of us unless inside the walls can know for sure)

keep us posted on how it goes.

[CW7491]

So I decided to do a little holiday baking ... with my temp/hum sensors. My hope was to remove the wet bias in drier conditions but also found surprisingly better performance at saturation after the bake.

I have a new Davis SHT31 (which I believe was manufactured in June 2019) that I've reported on in this forum with disappointing results. I also have a few SHT75s. I baked them all per Sensirion's datasheet for 10hrs at 220°F. I did not bother with the rehydration procedure because it's the Puget Sound in December ... they will get plenty of hydration.

The SHT75 I used for the test had been outside for about 9mos. When I brought it in it read about 4% high at room temperature compared to my Belfort pyschrometer @ 40% RH. The new Davis SHT31 has been inside aside from a week of testing outside a few weeks ago. It was 6% high and +3°F dew point compared to the Belfort even having been acclimated to the dry conditions of a heated house.

After baking I tested them against a Belfort psychrometer at room temperature. Temperature results were so accurate at 72°F among all sensors that there is nothing to note.

The Belfort results were:
72.3°F Dry Bulb/59.0°F Wet Bulb giving: Humidity 45% Dew Point 50°F

Davis SHT31:
72.3°F  43%  DP: 49°F

SHT75:
72.3°F  42%  DP: 48°F

Not necessarily surprising after 10hrs at 220°F. I then put the sensors outside in what has been near saturation conditions. I expected the sensors to carry a dry bias at high humidity. Surprisingly, I found the opposite.

Before the bake, I could not get the SHT31 above 98%. Now it goes to 100% and actually stays there. Not the "Sensirion jump" to 100% for one update cycle. The SHT75 before the bake would get to 98% and stay there with a jump periodically to 99 or 100%. Now it goes to 99% and stays there. It will jump to 100% but not sustain it like the 31.

Obviously the real test will be as conditions dry out. Unfortunately, that happens infrequently here this time of year. We've only gotten as low as the 80%s and there are signs of the SHT75 drying out better, but only by a percent or two and when you consider it being 1% low at 100%, I would say that's negligible.

Just thought I'd pass along as I was surprised by the high humidity results. I have a feeling that if it does dry out, I will be right back to where I was before the bake with all this high humidity, but if we get a dry day, I'll post the results.

[ValentineWeather]

 

[fkapp]

Thanks
So you bake the whole circuit board in the oven?
Or do you unsolder the sensor and bake?

what do you place on, cookie sheet or something?

[CW7491]

I had tested with an older sensor without negative effects, so I baked the whole thing with circuit board, wire, connector etc on a baking sheet.

[CW2274]

Not the "Sensirion jump" to 100% for one update cycle.

Hadn't heard that one before, but it's spot on. My last 31 would indeed hit 100%, but if you blinked, you missed it. This current one won't even do that, but it's still a better overall performer.
Thanks for the test!

[jgentry]

So I decided to do a little holiday baking ... with my temp/hum sensors. My hope was to remove the wet bias in drier conditions but also found surprisingly better performance at saturation after the bake.

I have a new Davis SHT31 (which I believe was manufactured in June 2019) that I've reported on in this forum with disappointing results. I also have a few SHT75s. I baked them all per Sensirion's datasheet for 10hrs at 220°F. I did not bother with the rehydration procedure because it's the Puget Sound in December ... they will get plenty of hydration.

The SHT75 I used for the test had been outside for about 9mos. When I brought it in it read about 4% high at room temperature compared to my Belfort pyschrometer @ 40% RH. The new Davis SHT31 has been inside aside from a week of testing outside a few weeks ago. It was 6% high and +3°F dew point compared to the Belfort even having been acclimated to the dry conditions of a heated house.

After baking I tested them against a Belfort psychrometer at room temperature. Temperature results were so accurate at 72°F among all sensors that there is nothing to note.

The Belfort results were:
72.3°F Dry Bulb/59.0°F Wet Bulb giving: Humidity 45% Dew Point 50°F

Davis SHT31:
72.3°F  43%  DP: 49°F

SHT75:
72.3°F  42%  DP: 48°F

Not necessarily surprising after 10hrs at 220°F. I then put the sensors outside in what has been near saturation conditions. I expected the sensors to carry a dry bias at high humidity. Surprisingly, I found the opposite.

Before the bake, I could not get the SHT31 above 98%. Now it goes to 100% and actually stays there. Not the "Sensirion jump" to 100% for one update cycle. The SHT75 before the bake would get to 98% and stay there with a jump periodically to 99 or 100%. Now it goes to 99% and stays there. It will jump to 100% but not sustain it like the 31.

Obviously the real test will be as conditions dry out. Unfortunately, that happens infrequently here this time of year. We've only gotten as low as the 80%s and there are signs of the SHT75 drying out better, but only by a percent or two and when you consider it being 1% low at 100%, I would say that's negligible.

Just thought I'd pass along as I was surprised by the high humidity results. I have a feeling that if it does dry out, I will be right back to where I was before the bake with all this high humidity, but if we get a dry day, I'll post the results.

How are the results going so far with your baked sensors?

[jgentry]

So I decided to do a little holiday baking ... with my temp/hum sensors. My hope was to remove the wet bias in drier conditions but also found surprisingly better performance at saturation after the bake.

I have a new Davis SHT31 (which I believe was manufactured in June 2019) that I've reported on in this forum with disappointing results. I also have a few SHT75s. I baked them all per Sensirion's datasheet for 10hrs at 220°F. I did not bother with the rehydration procedure because it's the Puget Sound in December ... they will get plenty of hydration.

The SHT75 I used for the test had been outside for about 9mos. When I brought it in it read about 4% high at room temperature compared to my Belfort pyschrometer @ 40% RH. The new Davis SHT31 has been inside aside from a week of testing outside a few weeks ago. It was 6% high and +3°F dew point compared to the Belfort even having been acclimated to the dry conditions of a heated house.

After baking I tested them against a Belfort psychrometer at room temperature. Temperature results were so accurate at 72°F among all sensors that there is nothing to note.

The Belfort results were:
72.3°F Dry Bulb/59.0°F Wet Bulb giving: Humidity 45% Dew Point 50°F

Davis SHT31:
72.3°F  43%  DP: 49°F

SHT75:
72.3°F  42%  DP: 48°F

Not necessarily surprising after 10hrs at 220°F. I then put the sensors outside in what has been near saturation conditions. I expected the sensors to carry a dry bias at high humidity. Surprisingly, I found the opposite.

Before the bake, I could not get the SHT31 above 98%. Now it goes to 100% and actually stays there. Not the "Sensirion jump" to 100% for one update cycle. The SHT75 before the bake would get to 98% and stay there with a jump periodically to 99 or 100%. Now it goes to 99% and stays there. It will jump to 100% but not sustain it like the 31.

Obviously the real test will be as conditions dry out. Unfortunately, that happens infrequently here this time of year. We've only gotten as low as the 80%s and there are signs of the SHT75 drying out better, but only by a percent or two and when you consider it being 1% low at 100%, I would say that's negligible.

Just thought I'd pass along as I was surprised by the high humidity results. I have a feeling that if it does dry out, I will be right back to where I was before the bake with all this high humidity, but if we get a dry day, I'll post the results.

How are the results going so far with your baked sensors?

Glad to see that the forum is back online. I’m thinking about baking a old SHT-31 sensor. But I’m beginning to think that if I bought a new one, baking it before installing it would ensure me that any offset that the sensor has would be rid of.

[CW2274]

So I decided to do a little holiday baking ... with my temp/hum sensors. My hope was to remove the wet bias in drier conditions but also found surprisingly better performance at saturation after the bake.

I have a new Davis SHT31 (which I believe was manufactured in June 2019) that I've reported on in this forum with disappointing results. I also have a few SHT75s. I baked them all per Sensirion's datasheet for 10hrs at 220°F. I did not bother with the rehydration procedure because it's the Puget Sound in December ... they will get plenty of hydration.

The SHT75 I used for the test had been outside for about 9mos. When I brought it in it read about 4% high at room temperature compared to my Belfort pyschrometer @ 40% RH. The new Davis SHT31 has been inside aside from a week of testing outside a few weeks ago. It was 6% high and +3°F dew point compared to the Belfort even having been acclimated to the dry conditions of a heated house.

After baking I tested them against a Belfort psychrometer at room temperature. Temperature results were so accurate at 72°F among all sensors that there is nothing to note.

The Belfort results were:
72.3°F Dry Bulb/59.0°F Wet Bulb giving: Humidity 45% Dew Point 50°F

Davis SHT31:
72.3°F  43%  DP: 49°F

SHT75:
72.3°F  42%  DP: 48°F

Not necessarily surprising after 10hrs at 220°F. I then put the sensors outside in what has been near saturation conditions. I expected the sensors to carry a dry bias at high humidity. Surprisingly, I found the opposite.

Before the bake, I could not get the SHT31 above 98%. Now it goes to 100% and actually stays there. Not the "Sensirion jump" to 100% for one update cycle. The SHT75 before the bake would get to 98% and stay there with a jump periodically to 99 or 100%. Now it goes to 99% and stays there. It will jump to 100% but not sustain it like the 31.

Obviously the real test will be as conditions dry out. Unfortunately, that happens infrequently here this time of year. We've only gotten as low as the 80%s and there are signs of the SHT75 drying out better, but only by a percent or two and when you consider it being 1% low at 100%, I would say that's negligible.

Just thought I'd pass along as I was surprised by the high humidity results. I have a feeling that if it does dry out, I will be right back to where I was before the bake with all this high humidity, but if we get a dry day, I'll post the results.

How are the results going so far with your baked sensors?

I’m thinking about baking a old SHT-31 sensor. But I’m beginning to think that if I bought a new one, baking it before installing it would ensure me that any offset that the sensor has would be rid of.

I'd bake the old one, compare it to it's "pre-baked" old self and leave the new one untouched, at least for now to compare.

[CW7491]

How are the results going so far with your baked sensors?

Unfortunately I can’t go into detail at the moment due to not having the time and I will follow up with more info as I’m able, but I’ve baked (another) brand new Davis 31 with positive results. Sensor goes to 100% AND this one follows the SHT75 when humidity drops which I’ve never seen before. This one has a clear potting not the black potting which my other “new” sensor had. I don’t know if that is part of the reason for the difference. The clear potting WILL MELT at 220F. It won’t turn completely into a liquid, but you have to be careful. More to follow as I’m able, but I’m cautiously optimistic from what I’ve seen so far and as you know I’ve been pretty critical ...

[jgentry]

How are the results going so far with your baked sensors?

Unfortunately I can’t go into detail at the moment due to not having the time and I will follow up with more info as I’m able, but I’ve baked (another) brand new Davis 31 with positive results. Sensor goes to 100% AND this one follows the SHT75 when humidity drops which I’ve never seen before. This one has a clear potting not the black potting which my other “new” sensor had. I don’t know if that is part of the reason for the difference. The clear potting WILL MELT at 220F. It won’t turn completely into a liquid, but you have to be careful. More to follow as I’m able, but I’m cautiously optimistic from what I’ve seen so far and as you know I’ve been pretty critical ...

Please let us know when you’re able!

[galfert]

Oooh sounds like someone may be bound to an NDA and have some test hardware. How exciting.

[jgentry]

How are the results going so far with your baked sensors?

Unfortunately I can’t go into detail at the moment due to not having the time and I will follow up with more info as I’m able, but I’ve baked (another) brand new Davis 31 with positive results. Sensor goes to 100% AND this one follows the SHT75 when humidity drops which I’ve never seen before. This one has a clear potting not the black potting which my other “new” sensor had. I don’t know if that is part of the reason for the difference. The clear potting WILL MELT at 220F. It won’t turn completely into a liquid, but you have to be careful. More to follow as I’m able, but I’m cautiously optimistic from what I’ve seen so far and as you know I’ve been pretty critical ...

I’m baking an older SHT 31 with the black coating and see how it does. Planning on installing it in my remote unit tomorrow morning as long it’s not raining. You can view data from that 31 here:

https://www.wunderground.com/dashboard/pws/KALTHORS2?cm_ven=localwx_pwsdash

I’ll let everyone know when the 31 goes live. Right now, I have the 75 installed.

[jgentry]

How are the results going so far with your baked sensors?

Unfortunately I can’t go into detail at the moment due to not having the time and I will follow up with more info as I’m able, but I’ve baked (another) brand new Davis 31 with positive results. Sensor goes to 100% AND this one follows the SHT75 when humidity drops which I’ve never seen before. This one has a clear potting not the black potting which my other “new” sensor had. I don’t know if that is part of the reason for the difference. The clear potting WILL MELT at 220F. It won’t turn completely into a liquid, but you have to be careful. More to follow as I’m able, but I’m cautiously optimistic from what I’ve seen so far and as you know I’ve been pretty critical ...

I’m baking an older SHT 31 with the black coating and see how it does. Planning on installing it in my remote unit tomorrow morning as long it’s not raining. You can view data from that 31 here:

https://www.wunderground.com/dashboard/pws/KALTHORS2?cm_ven=localwx_pwsdash

I’ll let everyone know when the 31 goes live. Right now, I have the 75 installed.

Aight. I have the baked 31 installed on my remote station. Shouldn’t left the filter cap on when I baked it. lol. It melted some. I took it off and installed the sensor without a filter. Going to see how long it lasts. Lol.

[ValentineWeather]

How are the results going so far with your baked sensors?

Unfortunately I can’t go into detail at the moment due to not having the time and I will follow up with more info as I’m able, but I’ve baked (another) brand new Davis 31 with positive results. Sensor goes to 100% AND this one follows the SHT75 when humidity drops which I’ve never seen before. This one has a clear potting not the black potting which my other “new” sensor had. I don’t know if that is part of the reason for the difference. The clear potting WILL MELT at 220F. It won’t turn completely into a liquid, but you have to be careful. More to follow as I’m able, but I’m cautiously optimistic from what I’ve seen so far and as you know I’ve been pretty critical ...

Curious where did you order the clear potting sensor from? Davis or Ryan at Scaled?

[jgentry]

BTW. How quick did your baked 31 go up to 100%?

[CW7491]

Oooh sounds like someone may be bound to an NDA and have some test hardware. How exciting.

I’m not sure if this was for me, but I wish that were the case. The truth is I apparently just enjoy wasting my own time and money ...

[CW7491]

How are the results going so far with your baked sensors?

Unfortunately I can’t go into detail at the moment due to not having the time and I will follow up with more info as I’m able, but I’ve baked (another) brand new Davis 31 with positive results. Sensor goes to 100% AND this one follows the SHT75 when humidity drops which I’ve never seen before. This one has a clear potting not the black potting which my other “new” sensor had. I don’t know if that is part of the reason for the difference. The clear potting WILL MELT at 220F. It won’t turn completely into a liquid, but you have to be careful. More to follow as I’m able, but I’m cautiously optimistic from what I’ve seen so far and as you know I’ve been pretty critical ...

Curious where did you order the clear potting sensor from? Davis or Ryan at Scaled?

From Ryan. I’ve ordered another and will let you know if it’s the clear kind.

[fkapp]

How are the results going so far with your baked sensors?

Unfortunately I can’t go into detail at the moment due to not having the time and I will follow up with more info as I’m able, but I’ve baked (another) brand new Davis 31 with positive results. Sensor goes to 100% AND this one follows the SHT75 when humidity drops which I’ve never seen before. This one has a clear potting not the black potting which my other “new” sensor had. I don’t know if that is part of the reason for the difference. The clear potting WILL MELT at 220F. It won’t turn completely into a liquid, but you have to be careful. More to follow as I’m able, but I’m cautiously optimistic from what I’ve seen so far and as you know I’ve been pretty critical ...

Curious where did you order the clear potting sensor from? Davis or Ryan at Scaled?

While not a direct answer to the replacement sensor question, my new VP2 arrived the day after Christmas from Scaled Instruments.  I modified the filter with the SF2 Cap.  The picture of the SHT that came with the new VP2 is shown below (appears to be the clear coating mentioned above). Posting to help with what a recent SHT from Davis looks like to help with the questions.  So far been very happy and am running the Davis 7714 passive shield. When station went online was a .85" of rain and 96% humidity day/night. Today is 42% Humidity and sunny day, and has been +/- 2% to the NWS ASOS 2.6 miles away since online, and currently am 2% below the NWS ASOS, so fingers crossed wet bias does not rear its head.

[CW7491]

BTW. How quick did your baked 31 go up to 100%?

Pretty quickly. Within a day or two of being put outside. Keep in mind though I’m in Western Washington. We haven’t been below 70% humidity in probably close to a month. The sensor reaches 100% and stays there. It stays right with my SHT75 as humidity drops. Unfortunately we’ve only been as low as 76% for the raw humidity reading since I’ve had it out there. I will prob pull it inside and check it at room temp and humidity. A bit busy with other things at the moment, so will post more details as I can. The high Sensirion bias is still there, but it doesn’t seem exacerbated by Davis after the bake with this one. Finally, Davis has gone back and forth with the clear and black potting. I had 2 SHT11s when they used that sensor and one was clear and one was black. It may just be luck of the draw and not necessarily some improvement.

[CW2274]

How are the results going so far with your baked sensors?

Unfortunately I can’t go into detail at the moment due to not having the time and I will follow up with more info as I’m able, but I’ve baked (another) brand new Davis 31 with positive results. Sensor goes to 100% AND this one follows the SHT75 when humidity drops which I’ve never seen before. This one has a clear potting not the black potting which my other “new” sensor had. I don’t know if that is part of the reason for the difference. The clear potting WILL MELT at 220F. It won’t turn completely into a liquid, but you have to be careful. More to follow as I’m able, but I’m cautiously optimistic from what I’ve seen so far and as you know I’ve been pretty critical ...

Curious where did you order the clear potting sensor from? Davis or Ryan at Scaled?

96% humidity day/night. Today is 42% Humidity and sunny day, and has been +/- 2% to the NWS ASOS 2.6 miles away since online, and currently am 2% below the NWS ASOS, so fingers crossed wet bias does not rear its head.

You should actually compare dew points. I've seen as much as a 4% difference between stations when they show the same temp and dew point due to the particular spread at that time. Humidity AFAIC is only worth watching when very high or very low.

[CW2274]

How are the results going so far with your baked sensors?

Unfortunately I can’t go into detail at the moment due to not having the time and I will follow up with more info as I’m able, but I’ve baked (another) brand new Davis 31 with positive results. Sensor goes to 100% AND this one follows the SHT75 when humidity drops which I’ve never seen before. This one has a clear potting not the black potting which my other “new” sensor had. I don’t know if that is part of the reason for the difference. The clear potting WILL MELT at 220F. It won’t turn completely into a liquid, but you have to be careful. More to follow as I’m able, but I’m cautiously optimistic from what I’ve seen so far and as you know I’ve been pretty critical ...

I’m baking an older SHT 31 with the black coating and see how it does. Planning on installing it in my remote unit tomorrow morning as long it’s not raining. You can view data from that 31 here:

https://www.wunderground.com/dashboard/pws/KALTHORS2?cm_ven=localwx_pwsdash

I’ll let everyone know when the 31 goes live. Right now, I have the 75 installed.

Shouldn’t left the filter cap on when I baked it. lol. It melted some.

Hopefully it didn't emit any of the dreaded VOC's....