Topic: SHT30 vs SHT31

[idahowx]

I was curious about a measured performance difference between the SHT-30 and the SHT-31 (used in a more expensive weather station).   I would presume from the specs that it is probably a test/calibration expense difference.  I procured a SHT-31 eval board and I have several WH31Es  (SHT-30 inside).   At room temp, the temp readings were within 1/2 degree F difference and the humidity was around 1% difference (given the 1% resolution on the WH31E).  The next step will be a temp sweep comparison in a calibrated temperature environment.  The nice thing about the eval board is that it has a bluetooth interface to dump out the stored readings.  I will need to find out about a controlled humidity setup.  I have 6 SHT-30 sensors and they were all within +/- 0.5 F at room temp.  Of course this is a very limited sample comparison but you might still find it interesting. [ You are not allowed to view attachments ]  [ You are not allowed to view attachments ]

[galfert]

Great. Thank you for doing this. I'm interested in the results. For comparison here is spec test data from Sensirion comparing the SHT30 vs SHT31.

 [ You are not allowed to view attachments ]
 [ You are not allowed to view attachments ]

[galfert]

By the way I have a WH32B and it doesn't have whatever this is I've highlighted in your WH31E. Anyone know what that is? I realize my WH32B has a barometer and the WH31E doesn't.

 [ You are not allowed to view attachments ]

Here is what my WH32B looks like:
 [ You are not allowed to view attachments ]

[kbellis]

I will need to find out about a controlled humidity setup.

Back when I was doing archivist work, I used a large plastic storage tub with about 1" of water and some cheap plastic stackable closet organizers that had an open mesh to rest the documents on. About 30 hours in that sealed container and the dry brittle 200-year old documents relaxed enough to open and flatten while drying pressed between chip board. Just one idea.

[nincehelser]

I will need to find out about a controlled humidity setup.

You can do the salt test.  Mix table salt with water so that it is saturated... still crystals, yet no liquid.  Put in small enclosure with device. 
  Let stabilize.  Environment should be 75% RH regardless of temperature.

Or you could buy a Boveda bag.  It's the same principle, but you don't have mix your own salt and water (which can be tricky).

https://www.amazon.com/Boveda-One-Step-Hygrometer-Calibration-Kit/dp/B000A3UBLA/ref=sr_1_3_s_it?s=hpc&ie=UTF8&qid=1548458712&sr=1-3-spons&keywords=boveda&psc=1

[HankR]

By the way I have a WH32B and it doesn't have whatever this is I've highlighted in your WH31E. Anyone know what that is?

Idahowx, could you please do another pic with the mystery object(s) in perfect focus?

This would help a lot for chances of identifying.

[nincehelser]

By the way I have a WH32B and it doesn't have whatever this is I've highlighted in your WH31E. Anyone know what that is?

Idahowx, could you please do another pic with the mystery object(s) in perfect focus?

This would help a lot for chances of identifying.

I'm going to take a wild guess that it is a ferrite core antenna.  Not sure what it would be for, though.  I looked at the manual and it seems it might have a WWV receiver in the sensor?

[kbellis]

Just curious: How would a sling psychrometer compare ?

You can do the salt test.  Mix table salt with water so that it is saturated... still crystals, yet no liquid.  Put in small enclosure with device. 
  Let stabilize.  Environment should be 75% RH regardless of temperature.

Or you could buy a Boveda bag.  It's the same principle, but you don't have mix your own salt and water (which can be tricky).

https://www.amazon.com/Boveda-One-Step-Hygrometer-Calibration-Kit/dp/B000A3UBLA/ref=sr_1_3_s_it?s=hpc&ie=UTF8&qid=1548458712&sr=1-3-spons&keywords=boveda&psc=1

[kbellis]

By the way I have a WH32B and it doesn't have whatever this is I've highlighted in your WH31E. Anyone know what that is? I realize my WH32B has a barometer and the WH31E doesn't.

Could it pertain to the dip switches for the various channel settings?

[HankR]

I think that your ferrite antenna guess is a very good one.  Downloading the recent WH31 and WH32 manuals shows that the WWVB receivers for both are mounted in a sensor housing, not in the console.

Maybe the WH32B version is older and had the antenna jammed somewhere else; maybe even indoors in the console.

[galfert]

I think several of you have touched in different key points regarding he mystery component. The WH31E is a device that can be one of many talking to the same console. So it has dip switches to change broadcast channels. So perhaps this mystery component is used in the setting transmission of those different channels. Where the WH32 just transmits on a set frequency as it has been designed as the stations primary temp/humidity and barometer.

Thank you all.

[HankR]

I can't believe this.  Just using a different browser (Firefox) to view wxforum and I've found it makes a huge improvement.  Other browser does not render pictures properly.  In particular many pictures are clipped on the right side.

Only now have I a proper full view of the mystery component, and I can say without hesitation that it is indeed the WWVB ferrite core antenna for the sensor head.  So this coil resonates with a capacitor at 60 kHz.

Here's further confirmation.  A picture of a DCF module; DCF being the German equivalent of WWVB.  The same antennas are probably sold for both stations, with just a change of the resonating capacitor's value to adjust for the slightly higher freq. of DCF.

Gosh, I can't upload this little jpg of about 14 kByte size.  Have no idea why and I've uploaded many pics to a forum almost identical to this one.  I'll keep working on the upload and it may be due to almost running out of HD working memory on this PC, which is the reason for going to Firefox; the other browser refuses to run without more memory.

It's a solid state drive and isn't big enough.  It was an experiment and I already have another twice-as-big SSD of 100 GB for this PC.

[HankR]

Okay, I see that it did come in as an attachment.  Now just have to see how to do embedded images, which I sort of tried first but that didn't seem to work either.

[galfert]

Okay, I see that it did come in as an attachment.  Now just have to see how to do embedded images, which I sort of tried first but that didn't seem to work either.

To embed the picture attachment you need to click on the link next to where you "Choose file" and that will insert HTML code into your post that will then be used as a placeholder of where within the body of the message to show the picture. You won't see the actual image till you post or preview.

[galfert]

I can't believe this.  Just using a different browser (Firefox) to view wxforum and I've found it makes a huge improvement.  Other browser does not render pictures properly.  In particular many pictures are clipped on the right side.

I'm using Chrome and the embed pictures are full resolution and in order to see the full image I just pan left to right and I can see the overhang. If Firefox shows the full image without panning then it is reducing the resolution in order to fit it. You can also right click on image and choose show on new tab to seem more of image without any website text.

Only now have I a proper full view of the mystery component, and I can say without hesitation that it is indeed the WWVB ferrite core antenna for the sensor head.  So this coil resonates with a capacitor at 60 kHz.

Here's further confirmation.  A picture of a DCF module; DCF being the German equivalent of WWVB.  The same antennas are probably sold for both stations, with just a change of the resonating capacitor's value to adjust for the slightly higher freq. of DCF.

So is your conclusion that this component WWVB can be used to set transmit on variable channels?

[kbellis]

I thought that the wwv reference was in the context of the national institute of standards and technology time, but don't understand how that relates to the 8 channel dip switches. Please let me know what parts I've misunderstood.

[galfert]

I thought that the wwv reference was in the context of the national institute of standards and technology time, but don't understand how that relates to the 8 channel dip switches. Please let me know what parts I've misunderstood.

Duh me. I totally forgot what WWVB was. Now it all makes sense to me. WWVB is to pick up the atomic time from Fort Collins, CO. The description of the WH31E is that it includes an RTC RCC. The RTC RCC is not needed for some consoles. But the WS-3000 does require the sensors to have their own RTC RCC. So that explains why my WH32B does not have this component.

[kbellis]

RTC? or RCC?

[galfert]

RTC? or RCC?

Yes RCC....fixed.

[idahowx]

Here are some quick data points.  The sensors were placed in a sealed bag and placed in a liquid fluoro-inert temperature chamber with an accuracy of 2% of a degree (C).  I chose 3 temperature points (-10C, 30C, 50C) that were within the operating temperature range of the eval sensor board (-10C to 60C).

            WH31E    Eval_Brd
            SHT30      SHT31
-10C     -8.5C       -9.93C
30C      29.9C       30.00C
50C      49.4C       49.87C

I was surprised to see the SHT-31 with a stable 30.00C reading but the SHT-30 was very close at this temperature.

Relative Humidity:  Since both sensors were in the same sealed bag, we could assume both sensors were exposed to the same relative humidity level.  The SHT-30 was within 2% to 3% of the SHT-31.    One humidity check at an endpoint was obtained by placing the sensors in a 100% dry nitrogen environment.  The SHT-31 sensor displayed 0.00%.  Unfortunately the WH-31E only goes down to 10.0%. 

Summary:  Both sensors are fairly accurate for the normal temperature ranges.   The SHT-31 was more accurate at the end point temperatures.  The SHT-31 seems worth the additional expense for extended ranges.

[kbellis]

Great report! Thanks for sharing.

[HankR]

I chose 3 temperature points (-10C, 30C, 50C) that were within the operating temperature range of the eval sensor board (-10C to 60C).

That's only 14 F, so not very cold.  I'd like to see a nice precise measurement taken down to at least 0 F; -10 F even better.

This extended measurement would be easy to do with a variety of small uP boards.  If I sent you an Arduino board so programmed, would you be able to make a few more cold accuracy checks?

[galfert]

Idahowx,
Good test. Seems to confirm spec data sheet from Sensirion. I'm a bit concerned only with better acuracy above 90% humidity living here in Florida. I'm sure some people in dry climates may be concerned with the low humidity below 10%. And those in cold climate could be concerned with cold temperature accuracy. So it seems everyone could stand to gain from a SHT31 upgrade. I'll be planning such an upgrade when my warranty is up in June. But why stop at the SHT31? I'm thinking SHT35 or SHT85.

[idahowx]

I chose 3 temperature points (-10C, 30C, 50C) that were within the operating temperature range of the eval sensor board (-10C to 60C).

That's only 14 F, so not very cold.  I'd like to see a nice precise measurement taken down to at least 0 F; -10 F even better.

This extended measurement would be easy to do with a variety of small uP boards.  If I sent you an Arduino board so programmed, would you be able to make a few more cold accuracy checks?

I am immersing the entire module or unit in a transparent sealed bag into the fluoro-inert.  However, the liquid is not real clear at -10C  or colder to easily see the display.  I am not sure I can trust getting precise results outside of the operating range and if the LCD display will function.   Temp sweeps take forever with this setup but I can try -20C and the RF interface.  I can also try the saturated sodium chloride solution as suggested to see what sort of difference there is at 75% RH.  This was just a curiosity project with what I have ... but it appears with a good precision soldering setup, an upgrade might be worthwhile.  The eval board if you are interested is available from vendors such as Digi-key. 

[idahowx]

Update / Summary  :

               WH31E       Eval_Brd
Temp        SHT30        SHT31
-10C          -8.5C       -9.93C
 30C          29.9C       30.00C
 50C          49.4C       49.87C

Humidity
N2 purge     10%         0.0%
NaCl sol.     74%       75.26% @18.31C
 
Notes:
Display Resolutions
WH31E               Temp: +/-0.1C,  RH: +/- 1%
Sensirion Board   Temp: +/-0.01C, RH: +/- 0.01%

Temp control: Calibrated liquid fluorinert chamber (+/- 0.2C)
Humidity: 100% dry Nitrogen & NaCl saturated solution
WH31E is limited to a low reading of 10% humidity[/color]
 
NaCl saturated solution RH (NBS publication 1976)
75.61%+/-0.18% @15C
75.47%+/-0.14% @20C
75.29%+/-0.12% @25C