DIY WIFI Rain Gauge, powered by Home Assistant
ฝัง
- เผยแพร่เมื่อ 23 พ.ค. 2023
- The video was created with the cooperation of:
JLCPCB 1-8 Layer PCB at $2. PCBA from $0 (Free Setup, Free Stencil)
Sign Up Here to Get $54 New User Coupons at - jlcpcb.com/SMA
In this video, I'll show you how I built my very own self-made and integrated with Home Assistant a Rain Gauge. The whole project is open-source, so you can download all the files and build it yourself. You can find more details on my website.
Project details: smartsolutions4home.com/ss4h-...
My store: store.smartsolutions4home.com/
If you like what I'm doing, consider supporting me:
PayPal.Me/ss4h - วิทยาศาสตร์และเทคโนโลยี
Wow, this DIY rain sensor project is truly impressive! You have showcased incredible skills in PCB design, CAD modeling, and 3D printing. The attention to detail in each aspect of the build, from the efficient layout of the PCB to the robust and sleek 3D-printed enclosure is really inspiring. The integration with a 'water-tight' :-) ESP32 for real-time rainfall data monitoring is a fantastic touch. Kudos to you for sharing this detailed and well-executed project. This is a prime example of what dedication and technical expertise can achieve in the DIY community. Great work!
I love these kind of DIY products! far away from easy things you find most of the time in youtube. Congratulations!
Thanks, I'm glad you like it!
you leave the market smarter than you go to the market !
An improvement would be a separate circuit to only turn on the ESP if it is raining. This could be completely analog for example it could use a 555 timer to see if the bucket has tipped.
Clever project! I liked the idea of the buckets that can be calibrated.
Thank you! It wasn't my first idea. It actually came up after some long shower testing :)
I make commercial water, soil and rain monitoring equipment for one of my companies (mostly the PCB design, and CAD design). Anyway, the commercial rain gauges work in a similar way, except they use something like a "spoon" that just tips out when full, and activates a reed sensor switch. This then wakes an MCU to update a counter. Then a realtime clock wakes it every few hours to upload the readings via LoRa/GSM. The LoRa versions can last literally years on an 18650 battery, as it's only using a few uA when sleeping :)
I agree, LoRa is absolutely perfect for such applications. Not only does it have a long-range capability, but it also has minimal power consumption. It works perfectly in devices where there isn't a lot of data to transmit. I should definitely consider working on a LoRa project someday :)
Thanks for the comment!
@@SmartSolutionsForHome LoRa Gateway first then you set yourself up for more LoRa projects that feed HA through the gateway ;)
Valuable comment thanks. Do you use a separate rtc chip in your board designs? I suppose it comes down to whatever has the least current draw?
@@bsmithril yes, separate RTC both due to lower power consumption, and also because it can run directly from a LiPo battery with no voltage regulation. Same with the MCU and GSM modules in these - all run from 3.4V to 4.2V, so well within the range of standard batteries. Reduces the BOM/assembly cost and power losses of using regulation. The LoRa modules (when used) do need 3.3V but it’s only on for a few seconds during a normal upload cycle.
Just taken delivery of the Rain sensor from your store. The quality of the 3D printing is superb. The attention to detail in the design makes this a prosumer level product. Many congratulations. I would thoroughly recommend anyone thinking about this to buy it.
(The only suggestion I would add would be a note in the delivered package with a QR code to link you to your website documentation and more specific instructions on the site to how to add this to espHome. Its not obvious because this is battery powered and you need initially to configure an esp32 with a USB cable)
nice to hear a review... I might buy one, too!
Nice work! Two suggestions, both related to how the device is installed:
IMO working with similar devices, levelling the device is quite important. Leaning a bit will throw off any calibration you've done. Alternatively, if you don't have a level handy, check & adjust your calibration after installing it.
It's also important that the installation to be secure. Vibration (eg, from wind) will induce slightly lower static friction in the hinge mechanism, which causes the bucket to tip at a slightly lower threshold, resulting in overestimates of precipitation relative to a vibration-free bench test.
Thanks for the kind comment. Regarding your suggestion about leveling, there was actually a scene with me checking if the Rain Gauge is level, but I ended up cutting it in post-production :)
As for the wind, you're absolutely right. It's important, and I didn't think about that. Maybe I should consider installing it in a slightly more sheltered location. Thank you!
Got really happy to see you uploaded a new video. Love the humor in the video to. Epic! 👏🏻
I'm really glad! I'm happy that someone is looking forward to my videos :)
Hello, I like the calibration - that is a smart solution :)
In fact you can build this thing without any soldering or ESPHome - simply buy a ZIGBEE DOOR sensor - they already have the hall sensor and everything you need - and they cost about 5$ and will work on battery for years :) Then just let Home assistant do the calculation ... I'm actually in the middle of doing that. :)
Great project, well engineered & designed. Also love the quote from a fantastic movie “Like tears in rain”
Thanks, I'm glad you enjoyed my reference to a true classic :)
This is exactly what I was hoping for, great idea.
I'm glad I met your expectations :)
wow. I really like the project. Also Your sense of humor. I been searching for something similar for ages. Finally a good project, clear instructions. And most funny - You are located quite close to my old home :)
I love your sense of humor. :)
This is excellent. I've just got it all working. I am currently looking at how to integrate it into Home Assistant - since the unit only transmits when there is a bucket tip, out of the box HA thinks the rain rate is whatever was last reported. I guess some kind of automation that sets the rate to zero in home assistant if there has been no update for X seconds is needed. Very cool project.
Might it be possible to modify the ESP code to report a zero X seconds after the last rainfall, before going into deep sleep - so that HA knows there is now no rain, as opposed to the rain rate last reported...?
I built a rain gauge like this last year. Used an existing funnel type rain gauge and retrofitted an esp32 and Hall effect sensor. I programmed the esp32 with micropython that makes a http call to a weather server I wrote in Python and Django. The http call tells the server to write an indicator to a MySQL database. The weather application then polls the database and calculates the amount of rainfall for a specific time period, 5 min, 1 hour etc.. it publishes that data to MQTT and Home Assistant is set to take those MQTT readings and update the gauges. One thing I did is used a lipo battery and charging circuit. I have it connected to a small solar panel so no battery charge out. :) Great build!
I love projects like this. This is a reason to follow you. So you've got a new follower.
Thank you and welcome aboard! :)
Great video, the design is so clever. I also loved the Blade Runner reference! ❤
I was wondering if anyone would recognize it. +10 points to you for knowing classic movies! ;) P.S. I'm glad you liked the video and the project.
I bought a stand alone rain gauge years ago with the idea to make something like this.
But I think I prefer this one. We often have a drizzling zero point something mm/hour rain fall and this one is easy to modify to use a bigger funnel. Definitely something I like to add to my HomeAssistant installation
Oooh... I like the way you present your video's. Nice tut. The tip for the calibration screws is a very good idea. Thanks for sharing this. 😀
Very nice design! Excellent solution! Well done Mate!
Amazing project, thanks!
Excellent!
The design is very nice. You could change the top grill to be a upper cone or something like that so a leaf has less chance of be stucked on it.
Thanks for the interesting idea. I considered it, but I think it could negatively affect the measurements. Some droplets could bounce off that grill and not enter the interior. That's why I placed it slightly below.
great job man!! i am impressed with the richness of details you had figured out. as a bonus it is a funny video.congratulations
if you don't bother i suggest to incorporate some spikes on the grid to avoid birds to make nests on the device, belive me, by practice i have discoverd that such things can happen
Great video, thank you.
Glad you liked it!
Kolejny - SUPER PROJEKT....
Dzięki :)
Ok. This was funny and informative. You earned my subscribe with ease.
I appreciate that, thanks! :)
YOu rock my friend!
there is another approach i would recommend you. If you use a reed switch instead of the hall sensor you could save some battery as they are passive components
Great rain sensor and super video
Thank you so much 😀
Thanks for great video! 🙂
I tried my best to make it as good as possible, and I'm thrilled that it turned out that way! :)
I will definitely try to create your device, will see how it will perform south of Poland.
@@KrzysztofStrug Powodzenia! Daj znać jak będziesz miał jakieś pytania :)
very nice
Well done! Might be a concern with the top grill getting blown off or even floating off if the cylinder fills with water. Could put a lip around the cylinder that the grill snaps into.
Superb video and great project. I want to try this! Is there a way to ensure it registers when it starts raining a very small amount? In the UK we call it "spitting", which is when there are only a few drops of rain. Your example was 8ml of water to tip the bucket. I guess this can be adjusted?
Can it be changed to 4ml or even 2 or 3ml? I am trying to think of disadvantages. Obviously the first one is battery life, but I will be using a power supply so that doesn't matter. (I plan to house all parts inside the house, we have a special waterproof hole that comes out of the wall, so I can pass cables through etc!)
Other disadvantages: If you make the funnel opening more narrow, then it will overflow in very heavy rain. But maybe it can be done by adjusting those screws? Also maybe the lifespan will be half (parts wearing out etc.) - what do you think?
I like these DIY projects and looking for a rain detection sensor. I wonder how robust your design is. I do not like to rely on a sensitive mechanism as dirt can accumulate over time and influence the output. What is you experience with this design until now?
Great video and project! What about using a solar panel to charge the batteries? Thanks for sharing this with us.
This is something I've been thinking about for quite some time. But I didn't want to prolong this project any further.
I plan to make a weather station at some point, with wind, rain, temperature, humidity, and pressure measurements. If possible, I might even add an air quality sensor - we'll see :) And when that happens, I'll definitely add a solar panel because, with such power consumption, batteries won't stand a chance.
That was very cool project! Haha, still waiting for your PCB making tutarials! Cheers from Trójmiasto! ;)
Dzięki! W takim razie pozdrawiam kolegę również z Trójmiasta! :)
Get a door/window zigbee sensor and you already have all the electronics you need
Stay tuned! An alternative version of the Rain Gauge is currently being created :)
Another great video! Congratulations!
Quick question, what is the estimated life time of the batteries? Thanks.
Thanks, I'm glad you like it! In my climate (Poland), it should hold up for about a year or so. We'll see how it goes in reality.
Great Project, what material did you print in, PETG or ABS ? Any specific infill required for the balance/tipping items ?
I used PETG, and as for specific requirements, there aren't any. However, to eliminate any slight differences, I added two screws that can be used to calibrate the device precisely during assembly. I talk about it in the video :)
Hi, could I replace the Esp32 by a reed switch, so every time the magnet moves it creates a pulse.
Then, I can read it with a 2 wire cable connected to another MCU.
I love this video including the humour 😂 Did you know ESP32 has a built in hall sensor?
Yeah, I actually wanted to use it in the first version. The problem is, it's not active in deep sleep mode. So, in order to minimize power consumption as much as possible, I had to use an external sensor and wake up the ESP32 using a regular GPIO.
@@SmartSolutionsForHome I see. I also agree about limitations with ESPHome. It's great and fast for many things, but when you want to do more advanced things I often find it difficult and you have to resort to crazy constructs, even with lambdas.
I've been playing with this concept to build a rain detector to warn me when it starts raining while I'm having my windows still open. Unfortunately it didn't work out as I was hoping for because many times the rain was already pouring for quite some time before the first notification came.
Quite recently I got the idea of building a acoustic rain sensor using a piezo electric buzzer as an analog input. I just got the parts a few days ago so I haven't been able to play with it yet. I actually have no idea how to make it work but I'm hoping for the best 😅
You've raised a very interesting point! :)
The device you're describing is something completely different from mine. The task of a Rain Gauge is to accurately measure the amount of rainfall in a given time. Due to the specific bucket capacity, you will indeed receive information about the rain with a certain delay. Therefore, it won't work well as a good warning system.
As for the acoustic sensor, I have mixed feelings. It seems to me that it will be very susceptible to interference from any other sound, but I could be wrong, of course.
I think that optics would be a better solution than acoustics.
Look into how such sensors work in cars. Just a few drops on the windshield, and the wipers already know to turn on :)
@@SmartSolutionsForHome when I started experimenting with my rain gauge, I didn't have a clue how long it might take for it to tip over first 😉 And having the ability to measure fainfall in a given time period felt like a bonus to me because I could estimate the soil moisture levels. But things turned out differently so now I'm measuring soil moisture levels instead ;-)
I read an article of a student from the Dutch university TUDelft, who started experimenting with quite a similar setup. His goal was to build an affordable acoustic rain gaug (disdrometer) and he uses the setup to count individual drops and measure the size of each droplet to calculate the amount of rainfall. Obviously, that is way out of scope for my project but it got me thinking. A €0,50 piezo electric speaker and a €0,05 resistor connected to an ESP32, what possibly go wrong? 😄
An optical rain sensor will be my next option. I allready scrapped 2 mice for that, but unfortunately they didn't have a commonly used sensor and I didn't have the time to play around with a signal analyzer yet.
💖💖💖💖
Can I ask - that plastic holder you put the pcb in when you were applying the solder mask - did you make this yourself? Or is it something you can get from JLCPCB?
I designed and 3D-printed it myself.
if i use domoticz, not HA, what solution i have to count/mesure ?
Out of curiosity, the ESP32 actually has a built in Hall effect sensor (though I don't know how good it is). Is the reason you didn't just use the built in one a matter of placement or is it not as easy to use the built in one to wake up the CPU from deep sleep?
Someone already asked the exact same thing earlier :) I replied to him: "Yeah, I actually wanted to use it in the first version. The problem is, it's not active in deep sleep mode. So, in order to minimize power consumption as much as possible, I had to use an external sensor and wake up the ESP32 using a regular GPIO."
I can see that you know ESP32 more than the average bread enthusiast. Fellow engineer? :) Thank you for the comment.
@@SmartSolutionsForHome Maybe I'm in the hardcore enthusiast category :-) I started from ESP8266 examples and have put together a unified firmware for my house that can be configured to drive various displays (ws2821, LCD, OLED, LED matrix) or various sensors (button, bme280, Lux, VOC, dust particles, IR motion) and provide REST, MQTT, Redis interfaces. Instead of using something like Home Assistant I'm running services I wrote in Java, Python, and Prolog. Later I made it portable to ESP32. I'm kind of roll your own personality. Haven't had a rooftop weather station since I took down the one I got from Dallas Semiconductor implemented with 1-wire devices, but always tempted. When I was first reading the specs and included devices for the ESP32 and got to the touch and Hall sensors I started laughing - of course you did...
awesome.. been wanting to do this for HA... Your site says the pcb assembled is unavailable. Are you doing per piece? Would like to buy so I can 3d print gauge as well... nice job!
I'm glad you like it! P.S. I soldered a few pieces, so for now, it's available: store.smartsolutions4home.com/product/ss4h-rg-rain-gauge-assembled-pcb/
@@SmartSolutionsForHome Hey thanks... was going to order but s/h says $65 to ship to US. Have to pass for now unless it's using wrong currency.
@@fins5150 Yeah, I know sending a courier from Poland to the States is ridiculously expensive. But if you don't mind waiting a bit for the delivery, the shipping only costs 8 bucks.
How about a fixed coil with a swinging magnet. The output of the coil could be (bridge) rectified (or just a diode) and the output "spike" clamped with zener diode. That output could fire an interrupt that wakes the ESP32 and so on... That would negate the need for running the hall effect sensor. This is just random thinking and I haven't really done this - let me know where my thinking is bad.
Thanks for the interesting comment. Something like that does seem like it could work, but I think it would be much more power-hungry than the Hall sensor.
@@SmartSolutionsForHome The idea is that a moving magnet generates the interrupt pulse; there should be no power used at all [except deep sleep quiescent current], power is generated by the magnet moving.
Unfortunately the "Notify me when available" is not working, would love to know when the assembled PCB is back in stock.
I'm sorry to hear that. I'm not sure why that would be... Thank you for letting me know! I'll look into it right away.
By the way, I've assembled a few boards if you're still interested :)
store.smartsolutions4home.com/product/ss4h-rg-rain-gauge-assembled-pcb
@SmartSolutionsForHome I figured it doesn't work from the product list (maybe because name or email hasn't been entered yet).
From the product itself it does appear to work.
Thanks for the info 👍
I might have missed it. What filament did you print with? Is it ASA?
You didn't miss it; I guess I just forgot to mention it. I used PETG.
@@SmartSolutionsForHome Maybe the most important thing is layer height, on the funnel? Because it needs to be smooth? Any other tips?
Very similar to Netatmo rain gauge module
The "tipping bucket" concept itself is a popular method for measuring rainfall. That's why many devices use it, like Netatmo, for example.
Cześć!
Jakiego lakieru użyłeś aby pokryć płytkę powłoką wodoodporną?
Planujesz może w przyszłości stworzyć całą stację pogodową? To było by na prawdę super. Obecnie nie ma żadnej na rynku która nie kosztowała by tyle co samochód albo była chińskim ... wynalazkiem.
Cześć! Użyłem dokładnie tego: "PVB16". Jest to dość popularny lakier do PCB do kupienia w większości tematycznych sklepów lub Allegro.
A co do stacji pogodowej to tak, planuję coś takiego zbudować. W tajemnicy Ci powiem, że już nawet zacząłem to robić :)
@@SmartSolutionsForHome Super!
A nie rozważałeś może otworzenia zbiórki kickstartera z tym lub innym projektem tak abyś mógł wziąć sobie trochę urlopu od pracy a żona nie marudziła, że pracujesz po pracy nad tymi projektami? Było już w historii sporo takich inicjatyw. np HackRF był wyceniony na 80.000 usd a udało się uzbierać 602.000 usd. Niestety linki będą usuwane przez YT więc nie mogę wstawić linka ale google pomoże.
Myślę, że poziom profesjonalizmu i zaangażowania w swoją pracę czyni Cię bardzo wiarygodnym na tym rynku więc szanse są bardzo duże. Z mojej strony mogę Cię tylko zachęcić. Chętnie wesprę ten projekt!
@@marcinu10 Szczerze to nie rozważałem nigdy tego typu zbiórek. O HackRF słyszałem oczywiście. To jest coś naprawdę użytecznego i innowacyjnego. Nie dziwię się, że aż taki sukces osiągnęło. Ja raczej nie zrobiłem nic super odkrywczego.. póki co :) Ale to miłe co napisałeś.
Ps. co do pracy po pracy i marudzenia drugiej połówki to nawet nie wiesz jak bardzo masz rację! ;)
Hi Sebastian,
I really like your content, I subscribed because of the rain sensor video. I have created a smart Irrigation setup for my garden last year and it relies heavily on Internet data sources for the amount of rain that had fallen. Like you stated in the video, this is not very accurate.
I love your creation, and there comes my question. I am quite proficient in 3D printing, but my soldering techniques are lacking.
I was wondering if you would consider selling the pcb and components soldered, in stead of the bare pcb? I can then print the housing and configure the ESP myself. This would also save some shipping costs I assume?
Please let me know what you think of this idea?
Keep up the work, I will follow your channel closely 😀.
I'm glad you're enjoying my content! I'm doing my best :)
As for gathering information from the weather station, I totally agree with you. It can't be relied upon when it comes to watering anything.
Regarding the ready/assembled PCBs, here's the link: store.smartsolutions4home.com/product/ss4h-rg-rain-gauge-assembled-pcb/.
I'm currently waiting for the components, but I should have a new batch available in the store soon. Sign up for the mailing list, and I'll notify you by email once it's ready.
@@SmartSolutionsForHome thanks for the delivery. I have received the assembled PCB. I now want to install the ESPHOMe part. On your site, you point to the chapter 'How to use OTA' , but I can't find this chapter? Can you help me? I do see an SSID but don't know how to connect?
Does anybody knows how to do the initial flash of ESPHome? I'm using an USB to serial adapter, but I do not now the wiring and next steps exactly... I see a Rain-Gauge-Store-Test access point, but I do not know the password...
Sorry that there's no information on the board I sent you. The password is: "1is9NCSaaThA."
If you're planning to do it via Wi-Fi, I recommend you this link: esphome.io/components/captive_portal.html
As for using the USB-UART adapter, I'll try to make a quick instruction guide as soon as possible.
did you know that you have a built-in hall sensor in esp32 :)?
What is the size of the magnet used in the meter?
It's 5mm in diameter and 3mm tall.
How is the paste jig created?
It's made by the PCB manufacturer. In this case, it's JLCPCB.
Can I change the program of the esp 32 and send the data to thingspeak?
Sure, you can change the software to whatever you want. I even added programming pins on the PCB so it's easier for you to do that :)
Another follow up, what's the estimated time the batteries run?
I mean how long can the batteries sustain the system before it drains
@@rainergeraldr.galinato7142 you can search for "mosfet as load switch", which will keep the microcontroller off until the bucket it tilted (then it sends the update to thingspeak and opens the mosfet again to power off).
While I like the video, I found that the Aqara Door sensor is a much cheaper Rain sensor, and lasts years on battery.
Maybe I'm not saying no. Actually, I quite like that idea! :)
Do you have it implemented yourself somehow? Could you tell me more about how you did it?
You can shave off the hall sensor, since this is build in to the ESP32... (choose the correct version😀)
I actually wanted to use the internal Hall sensor in the first version of my gauge. The problem is, it's not active in deep sleep mode. So, in order to minimize power consumption as much as possible, I had to use an external sensor and wake up the ESP32 using a regular GPIO.
@@SmartSolutionsForHome ahh, yes. That makes sense 😀
I think there is a typo error here (vlotage instead of voltage)
#Battery
max_battery_vlotage: '6'
min_battery_voltage: '3.6
Of course, you're right. Thanks for pointing it out - I'll fix it. P.S. You've got a really keen eye :)
haha I just spotted that today also
think you put the batteries in the wrong way but really good project to do one day
Thanks, I'm glad you like the project.
And about the batteries, I definitely put them in properly. In the next scene, you can see that the LED is lit. I didn't move them around between shots :)
@@SmartSolutionsForHome 7:09 it looks like you have there, looks like you got the + end against the spring with you meant to have the - against it, it's normally the - with the flat bottom of the battery goes onto the spring
@@AdamHinckley If you take a close look at the moment you mentioned, you can see that there's a "-" written on the battery with an orange tip. It might actually seem like the tip is not flat, but it's definitely flatter than the one on the opposite side :)
@@SmartSolutionsForHome normally the '-' has a flat bottom, them batteries look to me haven't, normally where I am, the '-' has the flat bottom
i need oneright now, can you sell me please one ? neighbor, I'm from Slovakia
Here :) store.smartsolutions4home.com/product-category/ss4h-rg-rain-gauge/
crack!!!
2 video is enough to subscribe your channel, macro button video editing and this video.
By the way Mr Bruce have an effort to make me subscribe to your channel on this video
I don't know Mr. Bruce, but I guess I should thank him then :)