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

[Old Tele man]

MY THOUGHT, Part Deux:  How about a small nylon tie-wrap?

[drew1021]

Tie wrap might actually work. They make some very thin ones.

[ValentineWeather]

Can't be just any silk has to be breathable. Pillow cases aren't the same as pantyhose.   

[Bobvelle]

Can't be just any silk has to be breathable. Pillow cases aren't the same as pantyhose.   

Yeah, I guess so. I figured the thin, light weight or mm weight would be plenty breathable.

Anyway...I just looked at a 31 that I had cooking in the 'Ole Crock pot, and I realized the mounting holes for the Davis cap are not being used. Therefore... A double wrap of thin stainless wire through the old mount holes, each lying on either edge of the filter cap, twist wires together on the back and Done.
I don't know why I didn't just do that now. I have the wire in my shop. Got it from Lowe's ... or Amazon.

[Bobvelle]

MY THOUGHT, Part Deux:  How about a small nylon tie-wrap?

I was just about to add " or monofilament line" when the stainless wire hit me. Also after realizing that the old cap mounting holes are not used, your idea about the cotton string is now looking like a possibility too. Just no way to garrantee it maintain pressure against the cap.

[dendrite]

Installed my new SHT75 yesterday evening with no filter and 9V running on the fan. I'll give it some time before analyzing the data too much. It hit 96% RH overnight without prime radiational cooling conditions.

[openvista]

Dendrite, would you be able to break down the steps involved in hooking up an SHT75 to the SIM?

As I understand it, you can snip the wire off an old SHT sensor, but which wires go to which leads?

You also mentioned "conformal coating". Could you describe that process? Looks like just slathering on some silicone-like substance over the leads to insulate them.

There appears to be some kind of "jack" (for lack of a better term) that the sensor leads slip into which then attaches to the cable going to the SIM. I assume you can also procure that part from mouser.com (along with the sensor), but what is it called?

If you are busy or unavailable, no sweat.

[Bobvelle]

Interesting. SHT31 (turquoise) vs SHT11 (black) Plot shows a RH crossover point at ~86%. Between 60%-86% the SHT31 shows 3-5% higher RH. Above 86% SHT11 starts showing higher RH and continues to till 10:30 AM this morning as it's till above 95% RH here.
Disregard prior to 5pm. SHT11 Sensor just came on-line.

Note: Both Temps were tracking within .5 deg and the Temp at crossover was 78 deg F.
SHT11 is in a Passive Rad Shield, 6+ year old and just baked and re-hydrated Sunday.
SHT31 is in a FARS & is about a week old.
 [ You are not allowed to view attachments ]


Just initial observations.. doesn't mean much..But I'll see if this continues over the next few days.

[openvista]

Interesting. SHT31 (turquoise) vs SHT11 (black) Plot shows a RH crossover point at ~86%. Between 60%-86% the SHT31 shows 3-5% higher RH. Above 86% SHT11 starts showing higher RH and continues to till 10:30 AM this morning as it's till above 95% RH here.
Disregard prior to 5pm. SHT11 Sensor just came on-line.

Note: Both Temps were tracking within .5 deg and the Temp at crossover was 78 deg F.
SHT11 is in a Passive Rad Shield, 6+ year old and just baked and re-hydrated Sunday.
SHT31 is in a FARS & is about a week old.

Looks like the lines cross not long after sunset.

You might, at some point, think to reverse the shield arrangement - placing the SHT11 in the FARS and the 31 in the passive shield. Reason I say that is because a FARS is going to produce a higher temperature on many nights (except overcast and windy) because it better protects against radiational cooling. The dew point should always be the same in both shields, theoretically, assuming the sensors have no bias or an identical bias. So, all things being equal, the FARS will generally have the lower humidity reading.

[Bobvelle]

Interesting. SHT31 (turquoise) vs SHT11 (black) Plot shows a RH crossover point at ~86%. Between 60%-86% the SHT31 shows 3-5% higher RH. Above 86% SHT11 starts showing higher RH and continues to till 10:30 AM this morning as it's till above 95% RH here.
Disregard prior to 5pm. SHT11 Sensor just came on-line.

Note: Both Temps were tracking within .5 deg and the Temp at crossover was 78 deg F.
SHT11 is in a Passive Rad Shield, 6+ year old and just baked and re-hydrated Sunday.
SHT31 is in a FARS & is about a week old.

Looks like the lines cross not long after sunset.

You might, at some point, think to reverse the shield arrangement - placing the SHT11 in the FARS and the 31 in the passive shield. Reason I say that is because a FARS is going to produce a higher temperature on many nights (except overcast and windy) because it better protects against radiational cooling. The dew point should always be the same in both shields, theoretically, assuming the sensors have no bias or an identical bias. So, all things being equal, the FARS will generally have the lower humidity reading.

Good point. I didn't consider that. However, on this particular evening/afternoon there was no sun as a storm blew in around noon and it was dark overcast the rest of the day. Not sure how that affects the results either.

[dendrite]

Dendrite, would you be able to break down the steps involved in hooking up an SHT75 to the SIM?

As I understand it, you can snip the wire off an old SHT sensor, but which wires go to which leads?

You also mentioned "conformal coating". Could you describe that process? Looks like just slathering on some silicone-like substance over the leads to insulate them.

There appears to be some kind of "jack" (for lack of a better term) that the sensor leads slip into which then attaches to the cable going to the SIM. I assume you can also procure that part from mouser.com (along with the sensor), but what is it called?

If you are busy or unavailable, no sweat.

I used this for reference.

http://www.wxforum.net/index.php?topic=14175.msg189484#msg189484

White (clock) - Yellow (vdd) - Green (gnd) - Blue (data) using the Davis cord. Red and black are unused.

And yeah...I'm just using a silicone based product.

https://www.amazon.com/MG-Chemicals-Silicone-Modified-Conformal/dp/B07BSK5S7B

It says "modified" now...not sure what that means. I got mine a few years ago and it was just regular silicone conformal coating.

Sensirion does not recommend soldering directly to the pins, but rather to use a connector (1.27mm pitch). I used this to eliminate soldering altogether...

http://www.newark.com/harting/14310410301000/receptacle-4way-1row-1-27mm-har/dp/53X2803

This is direct from the datasheet...

1.2 Sockets and Soldering
For maintain high accuracy specifications the sensor shall
not be soldered. Sockets may be used such as “Preci-dip /
Mill-Max R851-83-004-20-001” or similar.
Standard wave soldering ovens may be used at maximum
235°C for 20 seconds. For manual soldering contact time
must be limited to 5 seconds at up to 350°C7.
After wave soldering the devices should be stored at
>75%RH for at least 12h to allow the polymer to rehydrate.
Alternatively the re-hydration process may be
performed at ambient conditions (>40%RH) during more
than 5 days.
In no case, neither after manual nor wave soldering, a
board wash shall be applied. In case of application with
exposure of the sensor to corrosive gases the soldering
pads of pins and PCB shall be sealed to prevent loose
contacts or short cuts.

[openvista]

Thanks, dendrite! I think that will help a lot of readers, myself included.

[ValentineWeather]

FYI not sure about the SHT11 but I'm finding the SHT15 runs about .5° cooler constantly. I remember some mentioned this when 31's first came out. I checked last night without sun both in FARS shields and the 31's are right on with NIST certified, temperature was upper 50's at time.  Even today the 15 is running cooler at 80°.

Wind is calm currently and look how well the passive at top #1 is doing. 

 [ You are not allowed to view attachments ]

[Bobvelle]

I'm glad you brought that up. I honestly can't tell what the exact temp diff is between my SHT31 and 11 because I can't get "2nd Temp" to report in tenths of deg. In other words, it only reports as whole numbers.
This is, I'm sure, because I am using stock software and I'm in no way setup to do precise measurements of different sensors. Not yet anyway  .

However, The graphs appear to tracking each other by less that 1 deg since yesterday. But again it could be .5 difference or .7 or... 

[CW2274]

Can't be just any silk has to be breathable. Pillow cases aren't the same as pantyhose.   

I realized the mounting holes for the Davis cap are not being used. Therefore... A double wrap of thin stainless wire through the old mount holes, each lying on either edge of the filter cap, twist wires together on the back and Done.

You saved me a trip to the hardware store. I found some braided wire for hanging pictures in the house and tried it on my spare 31 and it works just fine, only one or two strands is plenty. I would say you don't really need to do the side thing, over top will be fine, especially for older eyes.
Thanks!

[Bobvelle]

Can't be just any silk has to be breathable. Pillow cases aren't the same as pantyhose.   

I realized the mounting holes for the Davis cap are not being used. Therefore... A double wrap of thin stainless wire through the old mount holes, each lying on either edge of the filter cap, twist wires together on the back and Done.

You saved me a trip to the hardware store. I found some braided wire for hanging pictures in the house and tried it on my spare 31 and it works just fine, only one or two strands is plenty. I would say you don't really need to do the side thing, over top will be fine, especially for older eyes.
Thanks!

Outstanding!
Good to know. I'll likely do this to my other 31 as a backup.

Speaking of (sort of) You think this SF2 filter would have any advantages or disadvantages in passive shield setup?

[CW2274]

Can't be just any silk has to be breathable. Pillow cases aren't the same as pantyhose.   

I realized the mounting holes for the Davis cap are not being used. Therefore... A double wrap of thin stainless wire through the old mount holes, each lying on either edge of the filter cap, twist wires together on the back and Done.

You saved me a trip to the hardware store. I found some braided wire for hanging pictures in the house and tried it on my spare 31 and it works just fine, only one or two strands is plenty. I would say you don't really need to do the side thing, over top will be fine, especially for older eyes.
Thanks!

Speaking of (sort of) You think this SF2 filter would have any advantages or disadvantages in passive shield setup?

What's good for the goose, is good for the gander. If I understand this filter correctly, it allows air to pass through, but also makes it waterproof when sealed correctly. I believe it is suppose to help the sensor from becoming saturated for you folks in the humid parts of the country. I don't think I really need this filter, as I not really experiencing problems like the others, but what the hell, can't hurt.

[Bobvelle]

Can't be just any silk has to be breathable. Pillow cases aren't the same as pantyhose.   

I realized the mounting holes for the Davis cap are not being used. Therefore... A double wrap of thin stainless wire through the old mount holes, each lying on either edge of the filter cap, twist wires together on the back and Done.

You saved me a trip to the hardware store. I found some braided wire for hanging pictures in the house and tried it on my spare 31 and it works just fine, only one or two strands is plenty. I would say you don't really need to do the side thing, over top will be fine, especially for older eyes.
Thanks!

Speaking of (sort of) You think this SF2 filter would have any advantages or disadvantages in passive shield setup?

... If I understand this filter correctly, it allows air to pass through, but also makes it waterproof when sealed correctly. I believe it is suppose to help the sensor from becoming saturated for you folks in the humid parts of the country....

WHAT!  Waterproof
I totally missed that. I thought it was just less "obstructive" and allowed moving air closer contact to the sensor than the Big Davis Filter. Also the other reason it was looked at was to add a dust filter in addition to the Davis filter to lessen the amount of contaminates causing premature aging of the Humidity sensor.

Well either way, I am noticing a big lag time between my SF2 filtered ,aspirated 31 and my passive shielded 11. But only on the humidity side and its the FARS sensor that's lagging. Humidity changes on my passive is way faster.  But there are way to many variables involved here to make any judgment about that, just not what I was expecting.

[dendrite]

Thanks, dendrite! I think that will help a lot of readers, myself included.

Like we were talking about before though...

If you don't like almost instantaneous response (like raw, non-averaged ASOS data) this may not be for you. Especially if you are using a custom FARS with much more CFM passing through it than the Davis fans. If that isn't a detriment to someone, then I think that person will love the 75. Of course if you find a way to put a filter over it or use the Davis stock fan, that will knock the response down. I would think a passive shield would knock it down even further, but given the recent passive vs FARS tests, who knows?

[openvista]

Thanks, dendrite! I think that will help a lot of readers, myself included.

Like we were talking about before though...

If you don't like almost instantaneous response (like raw, non-averaged ASOS data) this may not be for you. Especially if you are using a custom FARS with much more CFM passing through it than the Davis fans. If that isn't a detriment to someone, then I think that person will love the 75. Of course if you find a way to put a filter over it or use the Davis stock fan, that will knock the response down. I would think a passive shield would knock it down even further, but given the recent passive vs FARS tests, who knows?

For now, I will install it as a backup in a passive shield. If my 31 starts acting up, then, yes, I will put it in my primary FARS that has an AC fan (20-25 CFM). I've got averaging routines I can apply to the readings if necessary. That's the good thing about having a completely custom site and CWOP script. At least this way, I've got something accurate if I need it.

[openvista]

WHAT!  Waterproof
I totally missed that. I thought it was just less "obstructive" and allowed moving air closer contact to the sensor than the Big Davis Filter. Also the other reason it was looked at was to add a dust filter in addition to the Davis filter to lessen the amount of contaminates causing premature aging of the Humidity sensor.

It's waterproof but not water VAPOR proof. It just keeps large water drops from sitting against the sensor. That's assuming it's installed at the factory. If it's at all loose, water may find a way in underneath and, if it does, stay there longer than if you had no filter.

[CW2274]

Can't be just any silk has to be breathable. Pillow cases aren't the same as pantyhose.   

I realized the mounting holes for the Davis cap are not being used. Therefore... A double wrap of thin stainless wire through the old mount holes, each lying on either edge of the filter cap, twist wires together on the back and Done.

You saved me a trip to the hardware store. I found some braided wire for hanging pictures in the house and tried it on my spare 31 and it works just fine, only one or two strands is plenty. I would say you don't really need to do the side thing, over top will be fine, especially for older eyes.
Thanks!

Speaking of (sort of) You think this SF2 filter would have any advantages or disadvantages in passive shield setup?

... If I understand this filter correctly, it allows air to pass through, but also makes it waterproof when sealed correctly. I believe it is suppose to help the sensor from becoming saturated for you folks in the humid parts of the country....

WHAT!  Waterproof

You'd have to find another way, but doable I'd think.
https://www.sensirion.com/fileadmin/user_upload/customers/sensirion/Dokumente/0_Datasheets/Humidity/Sensirion_Humidity_Sensors_SF2_Filter_Cap_Datasheet.pdf

[CW2274]

I was noticing when putting on the filter that the rubbery substance around the sensor allows for what I think would be a good seal if the filter is smoothed off and the wire torqued down enough. At least you would avoid a sealant if it works.

[openvista]

My plan with the 75 is to use an old 11, cut the cable off, remove the cap and drill 2 small holes in the board.  I will then thread some wire around the plastic of the wire-to-board connector and through the holes in the sensor PCB, twisting and securing underneath. This should hold it tightly against the circuit board which I can then install in any Davis shield.

I'm uncertain how to secure a cap over this as it's too long for the Davis cap (which doesn't do much anyway) and the SF2 caps I have are not the right shape. I've read people suggesting SF1 caps, but, again, the issue is how would you secure them properly in this application? The cap isn't a deal breaker for me as I can stockpile and replace annually if this works out.

[Bobvelle]

I was noticing when putting on the filter that the rubbery substance around the sensor allows for what I think would be a good seal if the filter is smoothed off and the wire torqued down enough. At least you would avoid a sealant if it works.

Agreed.
 But I noticed 2 things while reading that link you provided.
1. touching the membrane can contaminate with oils and reduce the amount of humidity that can transfer to the sensor.  Yeah.. I touched it... like  a lot. I thought it was just a screen.

2. is ref to another link:
http://datasheets.closedcube.uk/Sensirion/Sensirion_Humidity_Sensors_SF2_Handling_Assembly_V1.pdf
 Where I Read the following:
 The following adhesives were tested for use with SF2 filter
caps and SHT2x: Electrolube SMA 10SL, DELO
MK096, DELO AD066, DELO 6093,EPO-TEK H70E/S,
EPO-TEK T6067, Lord MD-130