Solar Radiation Shield (Stevenson Screen) for Outdoor Thermometer

by clough42 Aug 14, 2016
Download All Files

Thing Apps Enabled

Please Login to Comment

Hi, clough42. Awesome project did you make. I will print it soon.
I have some doubts: How and what kind of material did you involve the wemos, battery and charger?
It makes them waterprof? Is the DHT sensor suffering with humidity? I will use a BMP280 sensor instead the DHT, so I am concern about water/rain/humidity.
Thanks for your attention.

Hi, I have a doubt. Do insects or bugs enter the shield? Thank you

Great design Will soon have one of these built in Kuna. We will have to get together some time.

Thanks for the great design! I 3D printed a holder and attached it to the side of my house (near the roof).

Awesome! Thanks for sharing.

James how did you made power supply without 3.7 to 5v booster?
If i power my wemos from 3.7 v it wont start at all!

I looked through all my photos, and I'm pretty sure I'm feeding the voltage in through the 5V pin. The RT9013 regulator on mine has a very low dropout voltage (250mv) so it works fine. If I recall correctly, it doesn't work as well if you feed the power through the USB port, since there's a diode in series with the USB connector.

hi James

u made gr8 project

but seems i got some troubles. could u help?

how did you managed power supply? I tryes data send every 10 min for 30 secs, but i loose about 0.1 v per day.
even if i connect solar pover ( 1 pcs 6v like you did? but my area isnt really sunny(() it's same problem. and it lasts only 2 days or less(((

i'm using easyesp firmware + deepsleep mode. seems its ok.

what i'm doing wrong? could you share some ideas?

Sorry...the information in my previous reply was incorrect. I just looked back at my notes. I can get an ESP-01 board to sleep at 17uA, but it looks like the WeMos is sleeping at 260uA. It pulls 90mA for 2 seconds while reporting the temperature, and then sleeps at 0.26mA for 300 seconds. This works out to 21mAh per day, from a 2500mAh battery. If the battery capacity is accurate, this should run for more than 100 days on just the battery. The solar cell easily tops it off. It spent a month or so under snow, and continued working just fine.

thnx James!

esp-01 not enouph but got one and try some play with it))
i'm digging ito deep charging-discharging cicles... doubled my solar power and made 15mins deep sleep and less 5 secs uptime. Seems its stable (about 3.65v) with mt wemos +espeasy))

Your's thingspeak graphs looks great - thats all my questions came from. big tnx!

PS i think its my charging board. ordered another to check!

pps snow! wow! thought its russian tradition only)))) lol

It's all about power consumption. Is the ESP8266 really going to sleep? Is there something else in your circuit drawing power? I verified with a multimeter in series that mine is only drawing 0.017mA when it's in deep sleep. The solar cell and charge controller easily top off the battery to 4.14V every day. I think, based on some calculations I did, it will run for about 90 days with no sunlight at all.

thnx again

espeasy does work/ it does deedp sleep.
will search my drawing power anв make more comments!

Does anyone have data showing effect of direct sunlight on this? Any ideas how it compares to molded ones?

I printed mine in white ABS and then spray-painted it white and mounted it above grass, away from pavement. I do not have hard data, but I don't see anything that makes me believe it's picking up heat from direct radiation. The temperatures from my sensor are always at or a degree or so below the official national weather service temperatures for my location, so I think it's working well.

Great info, thanks.


first: you made an awesome model! Looks great... wanted to print it as soon as I saw it.

One thing though... on the middle rings... how did you get the steep overhangs right at the beginning to print properly?

I tried supports in S3D, which didn't work as they somehow don't get close enough horizontally and i can't get them to touch the line that prints into air. I played around with various settings but didn't see any useful change.

Then I reduced the layer height down to 0.2 and even 0.1, which improved things a bit but it still prints the outer outline of the new layer into air for about 7 layers or so, which creates a mess and also causes the edge to curl up and that messes up some of the upper layers as well, even if those outlines are printed down on previous lines.

Thanks for any inputs and for sharing this model!



I didn't have any trouble printing it at all. I looked for the GCODE files, but I don't still have them. Here are my usual settings:

eSun white ABS
Layer Height 0.2mm
First layer height 90%
Top/bottom solid layers 3
Perimeter shells 2
Support: 30%, dense infill 70%, dense layers 3
Print speed 6000mm/min

could you link a tutorial about this? im stuck with wiring, about battery/charger and solar panel thanks,, it's awesome ;D

Do you have your arduino code online? I am working on a similar project right now and have a problem with deep sleep and the dht22 sensor not awaking after deep sleep :(
Also i want to add at least a wind sensor in the next step(s) ...

My code is based on the Homie convention and the Homie-ESP8266 framework. My current development code, based on Homie-ESP8266 2.0 (unreleased) is here:


The actual code that's currently running on my weather station right now is older, based on Homie-ESP8266 1.5: