Author Topic: THGR810 as an Outdoor Sensor  (Read 8303 times)

0 Members and 1 Guest are viewing this topic.

Offline aweatherguy

  • Senior Contributor
  • ****
  • Posts: 288
    • Weather Station Data Logger
THGR810 as an Outdoor Sensor
« on: June 07, 2011, 03:29:10 AM »
I've been using a THGR810 in a custom-built outdoor fan-aspirated radiation shield. The shield has a healthy amount of airflow and the THGR810 was mounted with the sensing holes (at the bottom of the sensor) pointed into the airflow.

Where I live there is frequent fog and the last couple of years we've had rainy periods where the RH seldom dropped below 90% for many days at a time.

After about 18 months the THGR810 is toast. Both temperature and humidity readings are way off. Upon disassembly, I found corrosion on the battery terminals and metallic shield on the transmitter PC board. I also found small amounts of dirt/dust throughout the interior of the unit. This is not surprising perhaps as this unit is not really intended for outdor use. Attempts to clean the unit and improve the accuracy have failed.

I've replaced it with a similar unit that has been calibrated by comparison to a Sensirion SHT15 sensor and the readings are looking very good (CWOP/MADIS analysis). This time the sensing holes are pointed AWAY from the airflow direction in the FARS. I might try to find some thin, dense synthetic fabric to cover the sensing holes for more protection from dirt/contamination. I wonder if GoreTex would be a good idea...?

Some info on just how far you can push these sensors...

Offline DanS

  • Chiang Mai weather
  • Forecaster
  • *****
  • Posts: 5434
    • ThaiWx
Re: THGR810 as an Outdoor Sensor
« Reply #1 on: June 07, 2011, 06:03:33 AM »
I would imagine any type of cloth type material over the sensors vents would alter your humidity readings (although Goretex would be an interesting test). Perhaps some very fine brass screen taped over the holes. Like the type of screen you find from a sink faucet to filter particles in the water. This along with pointing your sensor away from the fan like you said may be enough.
« Last Edit: June 07, 2011, 06:26:26 AM by DanS »

Offline SlowModem

  • Weather at the speed of dialup!
  • Forecaster
  • *****
  • Posts: 6640
  • WX @ 26.4 kbs
    • Watts Bar Weather
Re: THGR810 as an Outdoor Sensor
« Reply #2 on: June 07, 2011, 06:05:37 AM »
I would imagine any type of cloth type material over the sensors vents would alter your humidity readings. Perhaps some very fine brass screen taped over the holes. Like the type of screen you find from a sink faucet to filter particles in the water. This along with pointing your sensor away from the fan like you said may be enough.

Maybe a piece of fiberglass cut out from a cheap a/c filter, taped onto the inlet?
Greg Whitehead
Ten Mile, TN USA

Offline FrozenPenguin

  • Senior Member
  • **
  • Posts: 93
    • FrozenPenguin
Re: THGR810 as an Outdoor Sensor
« Reply #3 on: June 07, 2011, 12:07:12 PM »

As far as I can tell, that sensor wasn't designed to be weatherproof in the first place.


Oregon Scientific WMR200A + UV

Weather Site | | CWOP:7791

Offline aweatherguy

  • Senior Contributor
  • ****
  • Posts: 288
    • Weather Station Data Logger
Re: THGR810 as an Outdoor Sensor
« Reply #4 on: June 07, 2011, 06:31:18 PM »
Yeah, I did the whole experiment knowing that it wasn't designed for that kind of use. I was interested in seeing how it would hold up to the abuse.

It also occurred to me that I could run a thin strip of duct tape around the seam between the case halves and the battery compartment cover. That way the only way in or out for dirt would be limited to the sensing holes. Changing batteries will be a bit more involved.

I like the brass screen idea. I'll search around and see what the screen opening size is supposed to be on fine mesh. Is it small enough to keep out dust/dirt? I guess I'd better find out just how big dust particles are.

Does anyone know where I could get a very small sample of GoreTex?

I suppose all of these ideas will also slow down the response time to changes in temp/RH at least a little. I've got another sensor in a similar FARS right next to the first one so I should be able to see any additional time lag when temp or RH changes quickly.

Today I took apart a THGN801 -- the unit that is designed for outdoor use and mates with the OS raditaion shield. Comparing the construction with the THGR810, I cannot see much difference -- there is no conformal coating on the PC board to protect against moisture and the rest of the case is not any more tightly sealed. I would guess that the only saving grace on this unit is that it is not in a forced air stream as supplied by OS. I would not be surprised if after a number of years in foggy weather this sensor starts to have accuracy issues too.

Offline DanS

  • Chiang Mai weather
  • Forecaster
  • *****
  • Posts: 5434
    • ThaiWx
Re: THGR810 as an Outdoor Sensor
« Reply #5 on: June 07, 2011, 07:19:46 PM »
The water faucet filter type of brass screen has a tight mesh and I don't believe it would slow or alter temp/humidity readings. The finer dust could still get through though.
« Last Edit: June 07, 2011, 07:21:53 PM by DanS »

Offline aweatherguy

  • Senior Contributor
  • ****
  • Posts: 288
    • Weather Station Data Logger
Re: THGR810 as an Outdoor Sensor
« Reply #6 on: June 09, 2011, 01:26:52 AM »
DanS, thanks for posting the photograph. I've found something similar at my local hardware store. I may try two or three of them on top of each other for a little more filtering effect.

There's a local business that manufactures stuff (sleeping bags, jackets, etc) from Gore-Tex and they were kind enough to give me a couple of small samples for free. I talked to someone in the warranty department to get it. Interesting stuff. If you put it up to your mouth it is hard to blow much air through it and of course water won't go through at all. I stretched the Gore-Tex sample over a cup of boiling water from the microwave oven. A room-temperature soup spoon held over the fabric readily condensed moisture. Wierd.

Because the Gore-Tex will clearly keep all of the dust/dirt out I've decided to try that experiment first. I suspect I won't like the results but am going to try it anyway. I've duct-taped over all the seams in the sensor case and taped a piece of the Gore-Tex over the sensor holes. I've mounted the sensor with the Gore-Tex facing INTO the airflow since I don't think dust will be a problem and it probably needs all the help it can get to circulate air/moisture inside the unit.

This same sensor was tracking well against another one in a second nearby FARS so I should be able to see what changes...

If I'm not happy with the Gore-Tex, then the brass mesh is next...

Offline DanS

  • Chiang Mai weather
  • Forecaster
  • *****
  • Posts: 5434
    • ThaiWx
Re: THGR810 as an Outdoor Sensor
« Reply #7 on: June 09, 2011, 01:38:21 AM »
Interesting experiment and thanks for the info you're providing. The spoon test was useful as it showed that the boiling water's heat passed through to form condensation on the spoon. As long as moisture doesn't collect and remain on/in the Gore-Tex you may have found the answer. Let us know after a while how your sensor does. :-)

Offline DanS

  • Chiang Mai weather
  • Forecaster
  • *****
  • Posts: 5434
    • ThaiWx
Re: THGR810 as an Outdoor Sensor
« Reply #8 on: June 09, 2011, 06:44:26 PM »
Another thing that may help with deciding what to use for a filter. Davis has some fine mesh material over their thermo/hygro sensor chip seen here in the white cover - . I don't own one to physically check and maybe someone that knows can verify for you the material used with it.

Offline Skywatch

  • !!Warning!! I'm a Skywarn certified storm spotter and insane like one!
  • Forecaster
  • *****
  • Posts: 1844
  • Hakuna Matata (Timon and Pumba)
Re: THGR810 as an Outdoor Sensor
« Reply #9 on: June 09, 2011, 08:15:57 PM »
Where I live the humidity is always low. However I do recall having the same issue in the main OS sensor (temo/hum/wind) before I bought 2 of those THGR810's. They work pretty well (so far). Lets hope. Though I'm learning really fast how much better Davis units are though I still like OS. At least better than La-crosse.

Sometimes a hat helps LOL!!
« Last Edit: June 09, 2011, 08:20:07 PM by mckTXaws »
I live in an apartment and for the moment am not a home weather watcher.

I am a storm chaser.

Offline aweatherguy

  • Senior Contributor
  • ****
  • Posts: 288
    • Weather Station Data Logger
Re: THGR810 as an Outdoor Sensor
« Reply #10 on: June 09, 2011, 10:08:44 PM »
I'll have more complete data this weekend, but right now it appears that the temperature is experiencing perhaps a 3-minute delay and the humidity is more like 5-minutes (compared to the other sensor). I might be able to live with that. So far the Gore-Tex is working better than I thought it would...

I remember a marketing photo for Gore-Tex where they had a clear glass cup with hot water in it. A piece of Gore-Tex was stretched over the top of the cup. The cup was tipped over on its side and you could see the water being held in by the fabric, but just above the water line steam was coming through the fabric. I've looked for this photo on the internet but can't find it. Does anyone else remember that?

It's nice to see the Davis photo. I recognize the sensor -- a Sensirion SHT11 unit. I'm not sure what the "346" numbers mean, but the "11" numbers just below "346" means SHT11 as opposed to SHT10 or SHT15. I'll poke around and see if I can find some info on what the material in the white cover might be.

Offline DanS

  • Chiang Mai weather
  • Forecaster
  • *****
  • Posts: 5434
    • ThaiWx
Re: THGR810 as an Outdoor Sensor
« Reply #11 on: June 09, 2011, 10:35:11 PM »
Yes, the SHTxx Sensirion sensor chip  is the "Swiss Accuracy" mentioned because of it's origin. Even LaCrosse is using the same or similar in their stations (LaCrosse thermo/hygro sensor board, upper right in photo).
« Last Edit: October 04, 2013, 05:48:58 PM by DanS »

Offline aweatherguy

  • Senior Contributor
  • ****
  • Posts: 288
    • Weather Station Data Logger
Re: THGR810 as an Outdoor Sensor
« Reply #12 on: June 13, 2011, 04:21:19 PM »
Here's about 4 days worth of comparison between a bare OS sensor and one sealed with GoreTex over the sense holes and duct tape covering all the other seams in the case.

The two FARS I'm using are not completely identical so some of what we're seeing here may be due to differences in FARS performance. Other than that, I see little difference in temperature response times. At least for temperature then this looks like a good solution.

For RH, one of the sensors (the sealed one) may have a compromised RH sensor so I'm going to get a new sensor and re-run the experiment. What is shown here is where I've zoomed into one area of the data where you can see the time lag. It looks to me to be around 8-11 minutes on average throughout the data. That's not huge, but it does make a difference. I'll wait to get a new sensor and re-run this before coming to any conclusions however.

Offline DanS

  • Chiang Mai weather
  • Forecaster
  • *****
  • Posts: 5434
    • ThaiWx
Re: THGR810 as an Outdoor Sensor
« Reply #13 on: June 13, 2011, 06:21:46 PM »
The Gore-Tex looks like a winner. There are so many things between the 2 FARS that can/will change the readings and for you to get that close between the 2 says a lot. Perhaps run the same test again but this time swapping positions between the 2 sensors to see if the RH lag follows or not?

Offline johntr13

  • Member
  • *
  • Posts: 4
Re: THGR810 as an Outdoor Sensor
« Reply #14 on: October 04, 2013, 02:05:20 PM »
Hello,I live in Greece West Greece to Epirus state and I own a WMR200A oregon weather station.After 2 years my THGN801 temp and hum sensor had accuracy themes,so after installing new capacitors and new humidity internal sensor component,I decided to buy the THGN810.
As far in comparison with my old one and with the internal WMR200 panel I hope the accuracy problem is solved.I have created a FARS so far and I hope these wierd "stacks" of the THGN801 hum sensor(for many hours to 90+% ) will be solved.

 

anything