Author Topic: The Reliability of the SHT-31 Humidity Sensor & What Psychrometer Should I Buy? (Read 113671 times)

ValentineWeather · « **Reply #1100 on:** August 28, 2018, 09:36:27 AM »

Did a chest freezer test on sht31 vs 75 this morning for anyone considering change over living in colder climate. Results unscientific with sensors tied together. Let freezer go through 3 cooling cycles so took over an hour. Temperature difference was either .5 or .6 with the SHT75 reading warmer.
As freezer warmed up the 75 would really leave the 31 behind (+3) before compressor started again. Then on cool down it rapidly made the 3 degree difference up catching the 31 within couple minutes of compressor kicking on. The 75 does adjust rapidly to temperature change.
Below zero
SHT75 -8.2 SHT31 -8.7
SHT75 -8.1 SHT31 -8.6
SHT75 -7.9 SHT31 -8.4

jgentry · « **Reply #1101 on:** August 28, 2018, 10:14:52 AM »

Quote from: ValentineWeather on August 28, 2018, 09:36:27 AM

Did a chest freezer test on sht31 vs 75 this morning for anyone considering change over living in colder climate. Results unscientific with sensors tied together. Let freezer go through 3 cooling cycles so took over an hour. Temperature difference was either .5 or .6 with the SHT75 reading warmer.
As freezer warmed up the 75 would really leave the 31 behind (+3) before compressor started again. Then on cool down it rapidly made the 3 degree difference up catching the 31 within couple minutes of compressor kicking on. The 75 does adjust rapidly to temperature change.
Below zero
SHT75 -8.2 SHT31 -8.7
SHT75 -8.1 SHT31 -8.6
SHT75 -7.9 SHT31 -8.4

That’s not bad at all. Rarely if Alabama gets below zero but looks like the 75 isn’t a bad performer in that range. Wonder what your calibrated thermometer would’ve read?

dendrite · « **Reply #1102 on:** August 28, 2018, 10:19:58 AM »

Quote from: jgentry on August 28, 2018, 10:14:52 AM

Quote from: ValentineWeather on August 28, 2018, 09:36:27 AM
Did a chest freezer test on sht31 vs 75 this morning for anyone considering change over living in colder climate. Results unscientific with sensors tied together. Let freezer go through 3 cooling cycles so took over an hour. Temperature difference was either .5 or .6 with the SHT75 reading warmer.
As freezer warmed up the 75 would really leave the 31 behind (+3) before compressor started again. Then on cool down it rapidly made the 3 degree difference up catching the 31 within couple minutes of compressor kicking on. The 75 does adjust rapidly to temperature change.
Below zero
SHT75 -8.2 SHT31 -8.7
SHT75 -8.1 SHT31 -8.6
SHT75 -7.9 SHT31 -8.4

That’s not bad at all. Rarely if Alabama gets below zero but looks like the 75 isn’t a bad performer in that range. Wonder what your calibrated thermometer would’ve read?

The specs in that range are around +/- 1.2C for the 75 and +/- 0.25 for the 31. So yeah, only a consistent 0.5F difference isn't too bad. Heck, maybe the 31 is -0.2C and the 75 is +0.3C.

ValentineWeather · « **Reply #1103 on:** August 28, 2018, 11:49:34 AM »

I'll know more this winter with 31 as temperature backup. The freezer may not be that reliable but sure looked decent after 3 cycles. The 75's for sure are within specs both humidity and temperature unlike the 31's.

Update: I don't have a certified in that range or surely would of used it. I tested the certified not this range it read -9.7 so about 1° lower than the 31.. Guess its possible the SHT31 is reading on high side and why not that big a difference between the 75...If anyone has a certified below zero I would like to see results.

ValentineWeather · « **Reply #1104 on:** August 28, 2018, 11:55:23 AM »

I went back and looked at the certified and it was only -9.7. Response time was so fast I thought it was more.

Very possible if the certified at 0c is more accurate than the 31 means the 31's aren't even accurate on the low end. It is little odd the 75 was within .5F.

Here are specs on certified:±0.05°C within –2.00 to 2.00°C (28.4° to 35.6°F),

jerryg · « **Reply #1105 on:** August 28, 2018, 12:00:40 PM »

I may have to run that test, my calibrated meter has a calibration point of -10 degrees and is right on calibration wise. I will have to try it when i get the new sensor out of the standard shield.

ValentineWeather · « **Reply #1106 on:** August 28, 2018, 12:08:35 PM »

Quote from: jerryg on August 28, 2018, 12:00:40 PM

I may have to run that test, my calibrated meter has a calibration point of -10 degrees and is right on calibration wise. I will have to try it when i get the new sensor out of the standard shield.

Good deal Jerry.
I'm wondering about the 31 accuracy down low now.
Added the specs to certified thermometer I used above post.

dendrite · « **Reply #1107 on:** August 28, 2018, 12:23:37 PM »

Quote from: ValentineWeather on August 28, 2018, 12:08:35 PM

Quote from: jerryg on August 28, 2018, 12:00:40 PM
I may have to run that test, my calibrated meter has a calibration point of -10 degrees and is right on calibration wise. I will have to try it when i get the new sensor out of the standard shield.

Good deal Jerry.
I'm wondering about the 31 accuracy down low now.
Added the specs to certified thermometer I used above post.

How well do they all match up at a warmer temp in the fridge?

ValentineWeather · « **Reply #1108 on:** August 28, 2018, 01:07:23 PM »

Quote from: dendrite on August 28, 2018, 12:23:37 PM

Quote from: ValentineWeather on August 28, 2018, 12:08:35 PM
Quote from: jerryg on August 28, 2018, 12:00:40 PM
I may have to run that test, my calibrated meter has a calibration point of -10 degrees and is right on calibration wise. I will have to try it when i get the new sensor out of the standard shield.

Good deal Jerry.
I'm wondering about the 31 accuracy down low now.
Added the specs to certified thermometer I used above post.
How well do they all match up at a warmer temp in the fridge?

Not sure they match well in meteo shield, within a tenth in 50's.

jerryg · « **Reply #1109 on:** August 28, 2018, 02:26:41 PM »

Interesting results, calibrated meter accuracy +/- .3F meter -3.4 31-3.7 75-2.4 15-1.4 so figuring the meter between -3.1 to -3.7 it looks like the 31 is the winner which i would have expected based on specs. the 75 is pretty close to specs for it, not good on low end. The 15 is much better in the south than the north lol.

dendrite · « **Reply #1110 on:** August 28, 2018, 02:43:02 PM »

Quote from: jerryg on August 28, 2018, 02:26:41 PM

Interesting results, calibrated meter accuracy +/- .3F meter -3.4 31-3.7 75-2.4 15-1.4 so figuring the meter between -3.1 to -3.7 it looks like the 31 is the winner which i would have expected based on specs. the 75 is pretty close to specs for it, not good on low end. The 15 is much better in the south than the north lol.

Nice work from you two guys. Now I anxiously await the pin 35 and an easy way to implement it with a VP2.

jerryg · « **Reply #1111 on:** August 28, 2018, 03:08:26 PM »

Put another 75 in the freezer and first check was meter -2.4 and 75 -2.1 will let them cool a while longer but looks like reference temp is important on the 75 at low temps. If 31 sensors are basically close then comparing to 31 should work out. I will check other 31 later to see if it is close like the first one.

ValentineWeather · « **Reply #1112 on:** August 28, 2018, 03:15:35 PM »

I was letting them cycle all the way to bottom and record where they turn. Can't trust between because some sensors drop faster then others.

CW2274 · « **Reply #1113 on:** August 28, 2018, 03:27:51 PM »

Quote from: ValentineWeather on August 28, 2018, 03:15:35 PM

I was letting them cycle all the way to bottom and record where they turn. Can't trust between because some sensors drop faster then others.

Yes. Even monitoring just one sensor I always wait for it to reverse course and go from there, then you're sure.

ValentineWeather · « **Reply #1114 on:** August 28, 2018, 03:38:12 PM »

I'm data logging both the 75 and 31 so there is no question.

The certified will just record lowest temperature.

jerryg · « **Reply #1115 on:** August 28, 2018, 03:43:49 PM »

Well the meter has been at -3.7 for a while and the 75 is at -2.9 still looking like the 75 is short on temp. But the specs at this temp tells it like it is.

ValentineWeather · « **Reply #1116 on:** August 28, 2018, 03:55:50 PM »

It's all being data logged but here is a look at zero on the SHT31 the 75 was + (.5). I cranked freezer to colder setting so test still going. I'll let it cycle a few times to make sure it repeats.
[ You are not allowed to view attachments ]

ValentineWeather · « **Reply #1117 on:** August 28, 2018, 03:59:25 PM »

At -4 we are at (+1) on 75.
[ You are not allowed to view attachments ]

ValentineWeather · « **Reply #1118 on:** August 28, 2018, 04:07:31 PM »

8 below
[ You are not allowed to view attachments ]

ValentineWeather · « **Reply #1119 on:** August 28, 2018, 04:09:36 PM »

9 below
[ You are not allowed to view attachments ]

ValentineWeather · « **Reply #1120 on:** August 28, 2018, 04:13:19 PM »

10 below. No way I can use the 75 in winter.
[ You are not allowed to view attachments ]

ValentineWeather · « **Reply #1121 on:** August 28, 2018, 04:17:43 PM »

11 below Certified is running with the SHT31 also.
[ You are not allowed to view attachments ]
[ You are not allowed to view attachments ]

ValentineWeather · « **Reply #1122 on:** August 28, 2018, 04:33:12 PM »

-12.1 below and it's going to turn the certified recorded -12.29, Sht 31 -12.1, SHT75 -10.0 F
I've seen enough. ending my test. Once it got to -2 degrees difference seemed to stay. Not sure what it would do even colder but I can't use in winter when its been -39F here.
[ You are not allowed to view attachments ]
[ You are not allowed to view attachments ]

jgentry · « **Reply #1123 on:** August 28, 2018, 04:35:49 PM »

Quote from: ValentineWeather on August 28, 2018, 04:17:43 PM

11 below Certified is running with the SHT31 also.
[ You are not allowed to view attachments ]
[ You are not allowed to view attachments ]

That’s a bummer but then again, the spec sheet did say that it should start drifting below 0°F

dendrite · « **Reply #1124 on:** August 28, 2018, 04:43:41 PM »

Nice obs Randy. I'll probably get a new 31 and swap it out from the 75 before we start getting well below freezing. I still like the idea of the low thermal mass pin model, but I'll have to wait for the 35. rdsman is doing good work over in that other thread with reading the data from the 35.

News:

Author Topic: The Reliability of the SHT-31 Humidity Sensor & What Psychrometer Should I Buy? (Read 113671 times)