HTML-код
- Опубликовано: 13 окт 2024
- A simple Arduino-compatible device that can detect the water level - no moving parts and totally sealed.
PCB $5 for 10 pieces www.pcbway.com
First order free! 2-sided PCB just $5 for 10 pieces. Assembly for 20-pieces just $30.
PCB Prototyping the easy way - see my video #129 for Hands-On details.
You can read all this, with more information, and pictures in my GitHub:
github.com/Ral...
An ideal beginners' project to detect when liquid has reached a certain level. This module can be mounted at just about any angle,as I show in the video.
The best aspects of this device is that it is *totally sealed* and has no moving parts. The output is best detected by an analog GPIO pin but the value jumps very suddenly from the high 900s to pretty much zero.
The device fits through a hole in a water tank, water butt, pond... or anything else that contains water or other liquid. It could even be fitted to a bracket and totally immersed. It can detect the liquid at various angles, top down, bottom up, sideways; it's really very simple.
The video contains much more detail and also discusses progress on a couple of other PCB-based projects, such as the ATMEGA328PB processor board shown here. With a huge list of improvements over the standard 328P microcontroller it's probably worth thinking about a new design for this altogether. Probably.
Two I2C channels
Two SPI channels
Extra GPIO pins (at least two, up to six)
An extra UART (Serial) interface
Unique ID for each chip
Supported by Arduino IDE with a simple update
List of all my videos
(Special thanks to Michael Kurt Vogel for compiling this)
bit.ly/VideoLis...
You can read all this, with more information, and pictures in my GitHub:
github.com/Ral...
LINKS
Banggood For around $4, includes free 7-20 day shipping
www.banggood.c...
Various other types of water level detectors from Banggood
www.banggood.c...
eBay (UK) but doubtless available everywhere, for around $2
www.ebay.co.uk...
Amazon link for the Polulu 328PB board (if you find it elsewhere let me know)
amzn.to/2QgbHWN
Programmer for above (although others will [probably] work too, I haven't yet tested
amzn.to/34RulYi
If you like this video please give it a thumbs up, share it and if you're not already subscribed please consider doing so and joining me on my Arduinite journey
My channel and blog are here:
------------------------------------------------------------------
/ ralphbacon
ralphbacon.blog
------------------------------------------------------------------
You can read all this, with more information, and pictures in my GitHub:
github.com/Ral...
Just started watching and must say as one who keeps track as you say..
Your videos are amazing and got better if I may say.
The sound, camera resolution, your background.. everything!
In my behalf and my students (who I send to your channel frequently).
Thank you!
And thanks Benny for inspiring you once in a while..
You're very kind, Amos. I hope your students find them useful and interesting too. thanks for posting.
I always like it when I learn something from a video and then learn even more after reading some of the comments.
Yes, we can all learn from the comments too, George, it's amazing what information and experience there is out there.
This is brilliant, I needed a precise and reliable liquid level detector for my aquarium. As I am new to the Arduino world I thought it was out of my capabilities but thank to you no I have done it (and it works) and now understand more about programming and basic wiring required. Can not thank you enough, now a dedicated subscriber.
And welcome to my channel as a fully fledged Arduinite, Ian! So glad you tried and succeeded with this, that's exactly what I'm hoping I will achieve. Now the Arduino world is your oyster, what's the next project? A nice digital temperature gauge for that aquarium, perhaps? With notification to your phone when it goes out of band? All so very possible, although as always, one step at a time!
Thank you so much, the Arduino world is indeed my oyster now thanks to your brilliant, informative videos. Yes temperature measurement and remote monitoring is next along with LED lighting simulating sunrise and sunset. You bring this technology into the real word and everyday use.
Like your videos at lot, keep them coming! You did a video about the SR04T a while back and that got me to build a water tank measurement solution based on its waterproof sibling: JSN-SR04T. The advantage is that you can measure a range for the water level (not just a 'binary' value).
A few years, Eddy, ago I was looking for a waterproof ultrasonic detector but the only ones I could find were those designed for automobile use and had a very high price. Now you point me to this very cheap product. Drat! If only I could remember what I wanted it for in the first place I'd be really happy now!
Very interesting video, Ralph.
The transistor in the sensor is like an emitter follower. I highly recommend putting a resistor of a few hundred ohms in series with the collector (white wire). That won't affect the normal operation but if you accidentally short the emitter (yellow wire) to ground, it will protect the transistor.
By changing the two resistors (R1 and R2) you might be able to make the output voltage vary over a wider range for applications where you want an analog output.
A good tip indeed, Bob. In much the same way that I always recommend putting a low value (180R) resistor from a GPIO to the Gate of a MOSFET - if the MOSFET goes short circuit it won't take your Arduino with it! Thanks for the reminder.
This works because of the change in refraction index. Even if nothing moved, it's not that reliable in the long run. If your water is salty or quite hard, there can be deposits on the surface of the prism and that would create a layer preventing a direct contact between the plastic and the water so the system is not able to detect if water is present
Yes, others have mentioned this, not to also mention algae in fish ponds. Depends on how easy it is to clean now and again. In my water tank in the attic the water is clear, clean and no sign of any deposits so this would potentially be an ideal replacement for the float type that I installed a while back.
Good day, Ralph.
I'm glad to see that you made it safely back from your Mom's in Germany. This video is great for almost every tenants and/or house owner. I'll sure keep this part in mind when and if I have need for this type of detection. I just might add this feature to my weather station .. but I want to put a lid on this one, so I'll behave for once ! Hahaha!
Have a nice day and again, welcome back in the U.K.
Thank for the warm welcome back, Daniel! But is a project ever really finished? I've successfully installed the second Home Alone unit into my mum's apartment and it is working well, talking to the other unit via ThingSpeak. But I could still make improvements, of course! Good to hear from you.
For the black Sensor, just put a small peace of styrofoam with a zip tie on in 😉, then it never touches the water. Small trick that I made on a commercial dishwasher, works since two years now without any proplem.
So the moving arm bit is attached to the styrofoam thus raising it (and the hinge bit) well above water level?
@@RalphBacon yup, just underneath the swing arm. In my case, it was a bit try and error with the dimensions, I had to adjust the water level a centimeter lower 😕 to prevent overfill.
Interesting, indeed. I see how that would work, and keeping all the moving parts out of the water. You could even grease the hinge without the grease contaminating the water. If it were deemed to be necessary. Cool, thanks for the suggestion, not just for me but for others reading this.
@@RalphBacon you're welcome 😉
Your videos remind me of ice cream. No matter how big a helping I scoop out I find I always want more.
Very kind of you to say, Dirk, just don't over-indulge! Less is more (but not when it comes to ice-cream)!
Hi Ralph.
Welcome back. Hope your trip to Germany was enjoyable. Yet another helpful/useful video Ralph, "knowledge for the back burner" as I have other projects on the go just now. BTW, you converted me a while ago from the old techniques to make PCBs to a fully fledged Autodesk Eagle user with four different board designs now having been produced by PCBWay. The time saving now was worth the two full days it took me to get a reasonable understanding of Eagle. The PCBWay service is (as you said) really excellent, both quality, value and speed of delivery. From order to delivery in a little more than a week. Awesome. Thank you for your support and information.
I'm very pleased that, like me, you've taken the PCB plunge! Unlike me, you've chosen the ('knobbled') Eagle software (max 80mm square board) whereas I've gone the EasyEDA / KiCAD route. How do you find Eagle? Easy to use or just a matter of learning about its foibles? I find KiCAD more powerful (but far less intuitive, and seems to be written by techies for techies) but EasyEDA is perfect for the noob to PCBs. I'd be interested in your take, Paul.
@@RalphBacon Hi Ralph. I found some helpful RUclips tutorials which got me through Eagle's foibles. The main issue was setting the net classes in the edit menu. I would never have intuitively found it without a tutorial. The edit menu has many wondrous things in it!!! I feel reasonably fluent with it though and after the second board I no longer get warnings or errors which is rather satisfying. I would agree that there are many menu options that I've not explored and terms I've no idea the meaning of. They are probably for the nerdy techies, I'm just a techie. No offence to nerdy techies implied. I've never tried EasyEDA or KICAD. Not sure why. My largest board so far is 50 mm by 90 mm or 450 mm2. I have no projects, current or planned that need anything larger at the moment. If I do in the future and the board size restriction becomes an issue, then I'll consider switching to EasyEDA. So now that you're fully rested Ralph, can I tempt you to sort the NRF24L01 to work with an ESP VROOM 32 ??
It's still on my (ever growing) list of things to look at, Paul. Oh hang on, you said "fully rested". Flying to Germany, via Heathrow down the M25/M4, finding the car park full, down to level 2, also full, down to level 1, got stopped by some official who said better use T3 car park (in your dreams) and then finally finding a spot. Then flying off in a rather dilapidated plane (but which arrived on time), getting my hire car, not understanding where the handbrake is (there is just a push button), driving on the wrong side of the road, spending a week with a 91-year old who can't hear that well and who mixes both German and English in the same sentence these days, and then the repeat journey coming home. Yes, I'm really rested. I'm going for a lie down now.
Hey Ralph. Good news. With the aid of the guys at this webs site (see link) I have cracked my problems with the NRF24L01 and the ESP32 . It was beneficial to turn the BT and WiFi off too. www.iotsharing.com/2018/03/esp-and-raspberry-connect-with-nrf24l01.html. Basically replace the RF24.h with a modified version provided. Using the modified library, you can nominate any of the assignable GPIO lines for the SPI. E.g. RF24 radio(12, 14, 26, 25, 27); // RF24 radio(CE, CSN, SCK, MISO, MOSI); works fine. All the best and keep up the good work. Your tutorials have armed me with the confidence to work on these issues.
I very very interesting sensor, Ralph and welcome back. The 'lack' of surface area visa vi its use as a rain sensor: The surface or sensor area can be 'increased' by using a funnel. This is how rain gauges were designed pre the electronic maker heaven in which we now live. Imagine a measuring cylinder with a funnel at the top. The wide diameter of the funnel's top increases the catchment area of the device while the narrow diameter of the cylinder increases the sensitivity of the whole thing. The total rainfall would be calculate by multiplying the measurement in the cylinder by (diameter of cylinder/diameter of the top of the funnel).
In the case of a rain detector your sensor would be as near the bottom of the cylinder as is practicable. Does this mean you have to empty the cylinder each time it stopped raining? No. one places a syphon in the receiver such that the inside end of the tube was at the bottom of the receiver (at least, lower than the sensor) and the outside end being lower than the bottom of the receiver. Thus, the syphon will empty the receiver each time the receiver is filled. In your case, the receiver only needs to be a couple of centimeters tall. I apologise for this rambling description. A drawing would have been much simpler and easier to understand. When oh when will youTube allow attachments in their comments section.
C-Types such as uint16_t are a good habit to get into in the maker world. In the world of desktops most processors are 64 bit with a small number of legacy 32 bits machines still in use. In such environments, the signed and unsigned int types are 32 bits long. The wide variety of processors in the maker's world are far more varied ranging from 8 bit to 64 bit data busses. Many of the compilers for these devices adjust the size of the types correspondingly.. The Arduino is a case in point. Its int types are 16 bits long. This can lead to confusion . Hence the inclusion of the C-Type. All uint8_t and int8_t variables are 1 byte long and so on through int16_t, int32_t and int64_t types. This is not processor dependant and will remain the case no matter how many bits there in the data bus. So, code written for a STM will compile just as well on a 8050. The learning curve is really not that steep and porting code to other devices will be so much easier.
An interesting suggestion there, Michael, using a funnel. I've thought about what I said and I mabe over-egged it. I think four of these sensors might be enough to reliable detect the onset of rain (drizzle). How well they would do that in the Real World I don't know. I could test my single one out with a hosepipe I suppose...
You've hit the nail on the head regarding C-types. That's why I started suggesting this. I should really use these types all the time, like it did in the Home Alone project. And avoid the Arduino String type like the plague (causes Heap fragmentation and kaboom!). I'll see.
When you first started talking about your timer board, I thought you were going to use JCLPCB's fabrication service to build it - are you still going to use them to build the new daughter board? I was really looking forward to seeing how one goes about using this capability. I have watched several videos on the subject, but everybody just glosses over the the details about the actual ordering of the assembly and I figured that with your attention to detail, you would actually cover that.
Yes, Edward, I've been invited to use their STM assembly service. As you say, now for a 'daughter board' - well, almost. Ordering the assembled board is quite easy. As you develop the circuit/PCB choose SMT devices that are on the LCSC assembled 'Basic' list, or you will incur a $3 charge for each 'extended' component (sometimes cannot be avoided as it might be a chip).
Then, generate the Gerber files a normal, plus the BOM (Bill of Materials, aka component list) plus the Pick and Place list, al of which are submitted as part of the order. I'll be going through this in some details once I get a (working) PCB up and running. It's easiest to use EasyEDA as it's all tied into their stock and ordering system, but I guess as long as you can get the list of parts etc then any design tool will work.
@@RalphBacon I am glad to hear that you are still going to do the STM assembly video. I am currently designing a board for a Bally Miss Pacman game and it has a 5x5 array of lights that I want to replace with leds. Since the board has 25 identical circuits, I used Kicad with MitjaNemec's replicate_layout action plugin to duplicates schematics and layouts automatically - I could not find a capability like that on EasyEDA. There are 300 leds and 150 resistors on the board and that would be a lot to hand place. Kidad produces placement and bom files, but they are a little different than the samples up on JLCPCB, but they should not be hard to adjust since they contain all the important information. By the way, that Kicad plugin would be great if you wanted to panelize your own board.
Elektor produced a 328PB Arduino board with associated board files to allow use of the extra features. That was maybe 5 years sgo.
Yes, I Googled it and they did, Bernd. But it seems to have died a death as it is no longer available. I suspect that they overpriced it (€26 a board) when you can get the _standard_ Uno or Nano for just €3 from China! I'm wondering how cheaply I can make one, having now almost completed the Nano-sized 328PB PCB. I suspect it will not be less than €10 for small, hobbyist quantities. And I have currently ditched the USB socket in order to accommodate the extra pins whilst keeping it pin-compatible, just keeping the ICSP socket. Not sure if that is a wise move.
Congratulations on your project, Sir. Allow me to suggest that one of the best sensors for measuring water level, especially if the water is concentrated in a reservoir in a residence, are externally mounted sensors, as they do not come into contact with the water, avoiding any type of contamination of the water. Before using a sensor that comes into contact with water, it is extremely important to check with the manufacturer whether the raw materials used to produce the sensor contain any type of toxic material.
What external sensors would you recommend to accomplish this?
XKC-Y25-V
Thanks!@@carlossimoescampos1
Look interesting - I need to have an alarm for my coffee machine running out of water so I might try one of these!
Mr. Ralph, before buy this external sensor see its datasheet. I am not sure, but this external sensor can not be used on metallic surfaces.
Another great video, thanks Ralph. Just wondering why the values on the serial monitor changed in the latter part of the video (from 900+ to around 300 or so).
I'm going to have to watch that bit of the video now, aren't I, Ward, hang on... OK, found it, just before my project updates. I suspect there was _some_ water still on the sensor from my dunking experiments and that has changed the value - it is an analog value we are reading after all. If I did the same sort of test really, really, slowly I could get it to change much more slowly from the 900 range down to 0 but not smoothly at all, a gnats cough at 100m would make it jump suddenly. It probably could do with a (very) small capacitor on A0 to smooth out those jumps. That's my take, anyway!
Welcome back from Germany. I hope everyone is doing well and I was not aware I could find one of these on Banggood. This might get me back on the ball of hacking and improving my Biopod. I have a few of the hinged detectors and I can say that I am not a fan. Might be time to pull out the silicone and place an order at Banggood.
And I prefer using uintXX_t so that I know exactly what size the variable I'm using is as uint16_t should be the same regardless of your platform where as unsigned int might get you something different. But I will say, I use byte vs uint8_t because it is shorter and regardless of the platform, a byte is always an unsigned 8 bit value. A shame that the universal timer is moving to a daughter board, but it is equally exciting to see what new board you may come up with based on the 328pb. Nice to see that AVR might not be quite dead yet, but I'm curious to see the actual pricing and capabilities of the EPS32S2 as that might make for a hard pitch to go with any AVR based design.
I had to Google Biopod, now I see it's a type of vivarium. Yes this could work well in that, although I'm not sure if dirt/algae could affect it's longer term reliability.
Regarding the C++ method of defining variable types, I fully understand why Arduino.cc went the way they did and if users only ever use the Arduino platform that will be fine. But as soon as they move away from that, they might be surprised at how many bytes each variable type takes, hence my nudge to them!
I don't plan on using it directly in the Biopod, it has a horrible reservoir system that requires filling far too often so I want to improve my quality of life by attaching a larger one with features that let me know when the water level is low. Right now all I get is a dry running pump to tell me when I need to change things. Not the best system for such an expensive 'living' terrarium.
Well, if all you need is a simple switch type sensor (when the water drops too low, so you can refill automatically) this is worth a punt. You could connect this up to a simple (bright) flashing LED just as a proof of concept to see whether it would work OK over a month or so. I've also got an ultrasonic waterproof one on order for a future video, keep tuned!
Nice little sensor. Alas the water where I live is so hard that it would be caked in a few days. Some sort of sensing that never touches the water is what I need.
You say that, but the inside of my central heating expansion tank has no deposits in it. Nor the main water tanks, it's as clean as glass inside. And I live in a hard water area too. Perhaps it is only a problem if the water evaporates? But on to your desire for a non-touching sensor, an infra-red sensor could detect the level of water in a tank. Or a magnet floating in the tank (eg on an aquarium float) could be detected by a sensor on the outside of the (plastic) tank. It all depends on the environment you want to measure in.
I've pretty much stopped using anything but (u)intX_t. I always had to look up if "byte" or "char" are signed and how many bytes an "int" is on a particular platform.
Excellent, I'm glad to hear it, and for the reasons you mention here that's why I have mentioned it in the video. Glad we're all singing off the same hymn sheet.
Ralph, you could put that analog input on to a digital input and it would work fine.
I agree. This is obviously a boolean/binary output: Is water above or below a critical level.
It's a waste of an analog pin and code lines to read the analog value.
Hmm... You say " _waste_ of an analog pin", Martin, but remember the analog _capable_ pins are also standard digital pins (except A6/A7 on the ATMega328P, anyway). Unless you have a project requiring several analog pins I don't see how this is a 'waste'. I've yet to find a project that requires more than a couple of analog inputs, but then again I'm not aware of all the projects in the world either! But, I agree, given the sudden switch over in voltage levels, a traditional digital pin would work too , which means it would also work with pin extenders too.
Pardon me if I come across as a bit blunt when I say "waste". My line of thought is: All the analog pins can function as digital as well (you also state that), but not all of the digital pins can take analog input, so analog input pins are the scarce resource here. That means if you use an analog pin on a sensor that only ever gets digital HIGH/LOW signals, I see it as waste of the more capable analog/digital pin.
You are entirely right on the rest of your points too - it's rarely the case that you need a lot of analog pins, so hooking up sensors to them while prototyping is often not really a problem - especially if it's like your case, when the sensor is not really known to be one or the other beforehand.
In my current project I need at least 6 analog pins + I2C, which means I'm immediately out of any spare analogs even if I use a Nano, leaving me no room for adding more analog sensors. So I'm kind of used to not spending analog pins unless it's absolutely necessary.
Edit: On AliExpress this type of sensor is documented as being digital 0/1-logic. www.aliexpress.com/item/32946341765.html?spm=a2g0s.8937460.0.0.7a502e0elqqI4M
No problem, Martin, I didn't take offense! Interesting that it is described as digital. When you look at the internal circuitry you would think it more analog but I guess the water drop allows (all the) light to escape and thus _suddenly_ cause the transistor to stop conducting. So whilst not exactly a Schmitt trigger it probably _should_ be used as a digital device. Wow, your project does need lots of analog inputs, hence your concern. Thanks for clarifying, always good to hear from you.
Welcome home.
Is the "Home alone# project installed?
If so, has your mother got use to it?
Nice sensor, first time i see that, thank you.
I'm looking forward for you updates, will be interesting to see which way you go with the timer.
Indeed, Flemming, my Home Alone v2 device was successfully installed (after a quick code edit as I found an undocumented feature). Now I have two graphs on my phone, one for the Hallway, one for the Lounge. They both talk to each other via ThingSpeak and download the last 4 data entries to check whether that email is really required. The Lounge device also checks whether the Hallway has been set to "Going Out" in which case it just quietly ignores any non-activity.
On the timer front, I've submitted my pseudo-daughter PCB design, both bare boards and partially assembled SMT ones. Fingers crossed!
Hey Ralph, can you please make a video about your background and career? I've been watching your videos for a long time now and i'm curious to know more about the guy behind them. I think everyone would be delighted to know how you became an expert on electronics and programming.
_Expert_ is such an emotive term, Shakeel, and I do not consider myself to be one, neither in electronics nor even programming (which was my day job for a few decades). In fact, if you read the comments under each video you would probably get a good picture of my life - I'm pretty sure Google does that already and knows more about me than I do about myself! Hence the targeted ads I get served up on most web pages, just to ensure maximum enjoyment. Yeah, really.
5:21 but i think the optical prism could get dirty from water, too? 🤔
Depends on what what you are measuring, I guess. Pond water could definitely make it dirty. Household water (in a storage tank) pretty unlikely, I would have thought.
a quick wipe with vinegar soaked tissue will remove any scale that might get on the prism. however that scape accumulation is very rare unless you're using this in very brackish water
If the transition is so quick, you really only need to return 1 or 0 as a flag to beep or not beep.
So to fit in an INT you only need to define what low is, use bitwise operation to return a 1 or 0 so you only need an int.
Yes, that is true, although I had expected more of an analog signal when I fired it up. Just on the point of using a two-byte INT, that is somewhat overkill for a simple 1 or 0 as you suggest; better to use a one byte bool (or even a char, but the bool more accurately documents what you've proposed). But, unless you are _really_ short of run time memory I guess it's not that much of an issue. Good to hear from you, Mark, thanks for posting.
@@RalphBacon You know you said something about this not being suitable for a rain sensor as in the method you referred to.
Got me thinking that if you were to mount the sensor so that it had a 45 degree angle with the senor presented with a flat surface angled a bit, like 5 degrees to encourage roll off.
Because of the sensitivity of the meter you demonstrated, I'd say that if it is that sensitive that it might just work, your reading were in the 95x range at the leas, so you could safely say that anything over 800 is not rain.
Log anything between that and a definite of less than 100
The theory I have is this, if you have a water sprayer, try spraying with a fine mist and see what readings you get. and try with larger droplets too...
I am guessing that droplet size may have some factor in this, if it goes off with any water on it, then you might want to try one as a rain sensor just as a POC.
I take it arduino or the library used is what is responsible for converting the binary data to a decimal value for the user.
If you use bit shifts and masks, you can easily get a boolean and much more faster than using conditionals or ternary operators until you need them... if you need them because if all you need is an alarm to sound, you'd only need the boolean to turn a pin high or low which you could possibly check, but I reckon that the Pin High value is a binary 1 which outputs a logic level 1 (on)
I also thought that the option for a user chosen indicator like a flashing light, ring a door bell or other options, IDK, chair vibrator, not talking about that ridiculous thing called a butt licker, erm, kicker, so that someone who is deaf and blind can know if its raining when their chair vibrates.
Then... people don't need to ask Alexa if its raining outside.
Till I saw the date I thought you were being extremely lucky getting quick shipment from Banggood! You see I ordered 3 ESP32's from Banggood on March 20'th, I got a note today, May 3rd that it is still en-route "but may be arriving latter then expected!" Damn COVID crap!
So true, Jerry. I've been waiting weeks for stuff now, so long that I've forgotten what I ordered it for. Sigh.
I use an HC-SR04 ultrasonic distance measuring sensor to read the level of my swimming pool . No contact with the water. Precision I've found is +- .5 cm. Good enough for me but it could be better if I corrected the measurement with the speed of sound as a function of temperature.
The only problem I have with using HC-SR04 modules is that they will get damp from the moisture (evaporation) and can then corrode. Waterproof version are available (every car has them) but they are not cheap. You don't get this issue, I take it?
@@RalphBacon Hello Ralph! True, I've only been using the HC-SR04 for a few months. I haven't noticed any corrosion or strange behavior for now. The problem with your device for my appli is that it's going to detect absence of water every few days with the evaporation. I'm only interested in losses of about 5 cm which occurs about every few weeks or more. I saw a sealed distance measuring device (Tof10120) but it uses a laser which will only indicate the distance to my pool floor. If you have the reference of the waterproof ultrasonic version, I would like to inform myself. Thanx!
Bye and keep up with your interesting tutorials!
hey Ralph, big fan of your work... I was wondering if this would work on high temperatures and more importantly with liquids other than water.
It should - it's totally waterproof, so oil should be fine too, as long as it runs off the lens. Regarding high temperature, that depends on how high. I don't think boiling water is a good idea, but maybe up to 60° is _probably_ OK. See if the datasheet has more information before trying!
@@RalphBacon Thanks for the heads up!
Handy little device, I think I will have to get a couple.
Read the comments here about algae, salt, hard water first though! If I had not already got a working float in my tank I would definitely try this, though. Even if I had to wipe the surface now and again.
Hi Ralph, interesting sensor indeed, I wonder how it behaves with other fluids, I like the lean towards C++ despite being a devout Arduinite (as you no!) I have a 328 PB on order to play with to and still working with my battery (not sure if you got Email, probably lost in either!!) and I swear down that weird 1.5v AA is still fine ! good to see other projects moving on and I think great decision with your beeper board. I do look forward to your thoughts on MiniCore (its my goto now) but havent done anything with "LTO" I hope you can explain it with you usual style...glad you are well....cheers.
I got your email, Andy (and replied yesterday evening) and I think it's all smoke and mirrors with that 1.5v battery of yours. Also known as a buck boost giving you a full 3 volts! I await your details. I have read up quite a bit on MiniCore and LTO, very interesting. Amazing what you can learn on a couple of flights to Germany and back!
@@RalphBacon Hi, Oh dear! it would appear that the photos did not make it in my Email, as I was describing two projects and I can see that my written description has caused confusion I will send the details of just the Atmega 328_AU mounted on a breakout board, this time just one photo of the set up, and all will become clear with no other circuitry involved, thanks for reading it sorry it appears too not make sense, its my fault for using the photos with the other PCB in it.
As I said, smoke and mirrors (possibly unintentionally, I'm not so sure). 🤣
@@RalphBacon 🤣 All will become clear....Taken proper photos and I promise you no smoke and mirrors!!!...will get my setup and pics off to your inbox this weekend for your thoughts.
Welcome back. Mum and Germany all fine I trust? FYI It's a Male receptacle and they are resistors, not transistors. lol
Mum's just fine and dandy (as well as a 91-year old can be, anyway), thanks for asking, John. Ah yes, the plug on the end of this wire is female so I would definitely need a male *receptacle* (nice word, so much better than socket or whatever I mumbled). Is there a difference between a resistor and transistor then? What on earth did I say? I'm losing it.
Good stuff. As always.
I'm surprised you have the time still watch my videos, Jeff, have you sorted out that project of yours yet or are you waiting on a response from me? Since my Germany trip I've lost the thread (yes, and my grip, I know). Glad you liked it though!
Thank You Ralph!
Glad you liked it, Evgeny. As you might say, Очень хорошо!
Ralph S Bacon, да)), отлично! Your projects are very interesting for me. Some of our video bloggers use these ideas, but I like to get to know them in the original more.
So when it gets dirty it will reflect back to the detector and be on all the time? I would not use it in any critical application I think it works the same as a car rain detector but you don't have the wipers to keep the surface clean so will fail in detection when it gets dirty it took a long time to get my rain sensor working. reliably all the best Bob
If it gets dirty (eg in a fish pond) then I'm not sure what will happen - either it will reflect the light back or let it escape. But that's not a failure of the device, it's the implementation (and maintenance). In a clean water tank (eg the one in my attic) it would not get dirty (I hope!). But using it just for signalling (rather than controlling a pump, for example) is probably wise.
uint16_t is not C++, these fixed width types were added to stdint.h with C99 and the counterpart std::uint16_t was added with C++11.
Yup, twenty years now part of the standard and you still see "int" and "char".
Indeed, that's the problem, Rudolph, once users learn a particular construct it's most difficult to persuade them to try something new. I guess 'int' is OK in a loop but if you really need a particular size then the C/C++ types are the way to go. I wonder how many people use 'auto' as a type (as in, 'I don't care', which I've slowly introduced into my sketches and not one person has commented on it). Sort of proves what I say!
Incidentally, this refraction method is how rain sensing wipers on some cars work.
Which is what my previous video #179, was inspired by, Pete. But I guess car manufacturers have better resources than me (and better engineering skills too) to make it work reliably, like it does in my car. I'm still wondering whether just 4 of these in an array might not be a good replacement for Benny's cat run rain detector.
I use "unsigned int", "unsigned long" and the like in C++ all the time. Granted it's sometimes better to be more specific, but saying "it's Arduino speak, not C speak" is plain wrong. Arduino is merely a subset of C++. It doesn't contain anything that plain C++ doesn't have.
Well, I was over-simplifying it a little but it's still true to say that the Arduino environment hides the underlying C types for simplicity - and then users get confused when they read libraries and see such types mentioned.
Using the generic 'int' is fine when we don't care about the underlying data type and size (it's a system word size) but I do think that if developers (in this case, hobbyists for the most part) were to make a decision about what size they need their variable to be then they would be more aware of what is going on generally.
The same applies to the Arduino String (rather than std::string) which can cause heap fragmentation and an eventual crash. Which Arduino (hobbyist) developers use std::string? Do they even understand what the std:: library is all about? Or stdio or... you get the picture. Most don't even understand the use of flash memory to store string variable and stop running out of runtime memory!
I think it's great that Arduino.cc have abstracted all the complexity away but sometimes if a window is opened onto the Real (C/C++) World they might understand what is going on better. A bit like driving a car, if you know what a clutch is, how it works you might be a better driver. It doesn't mean you have to change your own clutch though!
Given it’s either high or low why not just feed it into a digital input pin instead? Or did you start out assuming it was a variable output..
I guessed it might be an analogue output, Eliot, and to some degree it is, but not really. You can treat this as a digital signal, works just fine.
Hello I have a question. My Sensor came with a small board connecting the sensor wires to the wires connected to the board. I have been trying to run your code, but my values remain at a flat 695-696
I suspect it's a wiring issue. Verify that the wires from which you are (not) receiving a signal are the ones actually connected to the sensor.
Wondering why the circuit is not more like say an opto switch where, rather than having a resistor in the emitter to ground, the circuit has the resistor between the collector and the supply, then the signal from the collector. Cannot see how it would be analog in either case. Perhaps it would work just as well as a digital input.
I would expect that you can do it either way, depending on what you want the OFF signal level to be. With the resistor in the emitter circuit, the output will be LOW if the transistor is NOT conducting, that is, the light from the LED has escaped via a water droplet. Putting the resistor in the collector leg means the exact opposite.
Strictly speaking this is an analog circuit. The transistor will partially conduct depending on the light level reaching it from the LED (which depends on how much has escaped via a water droplet). But even partial conductance ensures a big enough change in signal level as I showed in the demo. It's not a Schmitt trigger but acts pretty much like one as far as the µController is concerned, so it would work just fine on a digital input pin.
@@RalphBacon Point being, as the base is excited, would the transistor in current configuration not tend to turn off as it's emitter starts to rise.
So you need to buy the sensor and just add two resistors, is that right? If I understand it correctly, the optocoupler is inside the sensor.
That's right, all built in.
Thanks for the video! Ufortunately I saw after buying, now I realize it doesn't work for me, I thought it could told me the distance from the water mirror to the sensor.
Perhaps an combined transmitter/receiver IR module (or Ultrasonic module) would reflect off the surface of the water?
@@RalphBacon I saw that idea before, but there's a lot of steam inside the tank, I should get it water proof.
Those black plastic float switches are pure garbage. Almost fried a sump pump because they get stuck closed. Not physically mind you, rather something in the magnetic reed switch implementation. Stay away if you're using them for anything else than a notification. I've had two from different suppliers and both had the same issue.
How much power has the sensor to deal with?
The moving sensor types should only be dealing with tiny power levels - signal level. That is, 5v via a 10K resistor. Reed switches cannot handle any power at all.
Nice sensor, I want some (I have ordered some on Ebay for about $2). But I am not sure it is more reliable than a magnetic sensor. With optical you have problems with algae and sediments, neither will affect a magnetic sensor.
Your explanation of uint16_t was not very precise, for analogRead() a int is perfectly fine, the 0..1023 value will fit into both a signed and unsigned value and a int avoid some math hazards. Generally a int or unsigned int can be 16, 32 or even 64 bits wide, it is supposed to match the value with best performance on the actual processor. When you use xxx_t types you override this, this may be perfectly fine in some situation (Especially when communicating with sensors), but if you want to do some processing on the values it may not be optimal. A int16_t/uint16_t is not a good datatype on a 32 bit processor (ARM or ESP32), there int or unsigned int is much better. I frequently use the specific data types when wanting to save some RAM memory or communicating with digital sensor, but if I want performance I use int as data type.
Integer data types has a huge range, from 8 bit values to 64 bit (Maybe even larger) values, depending on processor (Try long long as datatype).
Let's hope you don't have problems with algae and dirt, Henrik. Not you personally, your yet-to-be-delivered sensor! I'm leaving mine outside for a month, pointing upwards, to see what happens to it.
I'm not sure I follow you with your argument of 'int16_t' and 'int' as the former will always be 16 bits but the latter will be whatever the designers of the platform decided it would be (probably a system word) so it will change depending on what you are compiling for. Using an 'int16_t' for 32-bit registers is probably not the most efficient code (so we can use int32_t, right, as we know the target is 32-bits) but using an unknown sized 'int' could be worse!
int is supposed to be the optimal integer size for the processor. For small processors I assume 16 bit or +/-32000 count, for larger processors at least 32 bit. Using (u)int16_t will not get your anything on a 8 or 16 bit processor (compared to (unsigned) int) and may slow your down on a 32 or 64 bit processor.
When I need something more specific I use the (u)int##_t types, but it is usually only when matching structures (for communication) or wanting to save memory that is necessary.
On 8 bit Arduino I often use int8_t/uint8_t(or byte), because it is a 8 bit processor and 8 bit variables is much faster and requires less memory than int that is 16 bit, but that will most likely slow the code down on a ESP32 or ARM processor without saving any memory.
Right, I now understand your take on this; you're relying on the compiler to give you the most efficient implementation of 'int' without having to really dig into the detail. Understood. And yes, an 8-bit int is going to be more efficient if for not other reason that it is a single byte! But drat that 255 limit!
I purchased some of these, but I can't figure out how to set it up with a monitor. And I don't think I'm up for the Arduino task since I've never done anything similar. But the 4 cables are confusing since the boards I have seen only have 1 input 🤷🏽. This is going over my head.
These 4-wire sensors normally give out a small signal when the switch is closed - as far as I know, only really suitable for detection with a microcontroller.
However, I also have one with just 2 wires and when the float moves up (because the water level has risen) it closes the reed switch inside the sensor with its built-in magnet. That's what you need. Note that the switch inside the sensor is NOT capable of carrying anything other than a signal-level (ie very low) current/voltage.
Ralph,you should make a video on how to create a file of your circuit to send to a pcb maker. i find this rather difficult ...
I have done a couple of videos that demonstrate this (even the promo videos show this in some cases). EasyEDA do have some good help pages that walk you trhough it.
I wonder if the infra red LED would promote algae growth?
I would have thought that infra red would be invisible to algae; after all, they need sunlight to grow and water would filter the IR light anyway, wouldn't it?
@@RalphBacon I wish I could make that assumption. From what I have Googled it sounds like infrared could promote growth of algae. It's too bad as otherwise these sensors look like the perfect solution.
I need some little help sir, is it possible to use this to give logic high or low for the digital I/O of pic184550.
PIC chips work in the same way for GPIO as Arduino does. You can detect HIGH and LOW states just the same.
Nice sensor Ralph. First I thought this would be exactly what I was searching for but it’s not. I’m still searching for a sensor that can tell me how many water is left in the rain catcher. Thought about ultrasound sensors to get it with the distance between the top and the water level but they will rust within a few weeks. So if you come along a solution for that please let me know.
How about a waterproof ultrasonic sensor, Michael? Like this one, but I've heard some rumours that the v2.0 is not as good as the original: www.banggood.com/SN-SR04T-DC-5V-Ultrasonic-Module-Distance-Measuring-Transducer-Sensor-IO-Port-Waterproof-High-Accura-p-1376233.html?p=FQ040729393382015118&
I'm thinking of getting one (both a v2.0 and an original if I can still find one).
Same request here ! currently i'm working on a waterproof ultrasonic sensor (like the one Ralph presents), (and keep looking for an ultrasonic sensor with most narrow transmission angle) . Other ideas i have to check (if my current project fails), include a laser distance sensor, and as an ultimate solution a differential air pressure sensor (will need plenty of math on this for compensation with current atmospheric pressure) capacitance sensor is another idea (not sure if the quality of water will affect the sensor)
@@RalphBacon I' m already working on this sensor. It's transmission angle is ~45 degrees, and i'm not sure (have not test it yet) if the walls of the tank (in my case 1,5 meter diameter) will affect the measure. Conversion of the distance to liters is made with a ''map'' command as soon as you determine how many liters take to raise thee level into 1 cm.
sfour3003 nice Sensor. But for my little 500 Liter water tank it needs to have a very narrow angle. Otherwise it won’t work. I also thought about a laser distance measurement with maybe some styrofoam floating on top of the water because I doubt the water will reflect the laser in a useful and reliable manner. Any other ideas are highly appreciated.
@@Hasitier depending on the type of your tank, there might be only few solutions. 1. a weight sensor (if possible to be placed under the tank. 2. an external stripe of cooper to act as a capacitor (if your tank is above ground) and finally 3. an air pressure meter, in combination with a tapped pipe in one end and submerge it to water in order to measure the pressure of the air trapped in the pipe. I hope the shape of your tank will let you calculate its area.
is it prone to being covered with the same minerals and becoming blind?
Not as far as I can tell. I've had this on my workshop roof since I did this video (several months if not 1 year by now) and it is still shiny and new. I think the rain actually keeps it clean in this case. It very much depends on the quality of the water it's detecting I suppose. And, in the same way that I have to occasionally clean the front of my CCTV cameras, I guess this would need the occasional wipe too.
Great video. Thanks for sharing.
Thanks for watching!
Very informative video
Glad you found it useful, Muhammad, thanks for posting. 👍
dual sensor for hugh water tank? full/refill, how to do
Indeed. How far have you got so far in your research?
Is it posible to use for fuel ?
I suspect any liquid would be detectable - but like all things, you will have to test that out on the workbench first.
I like your explanation of uint16_t.
Learned something new.
Thank you!
Excellent, now you know about it you can use it. Don't forget to use other types too, 8 bits, 32 bits, even 64 bits (known as a long long, yes really).
can I immerse it in boiling water?
Well, there's one sure way of finding out! But seriously, if there is a datasheet you need to read it.
It is the PITA guy again.....I thought a water level detector had more an indication of the water level. This is more a water present sensor? As such you would need two to control/indicate water level (say low and high).
Well Kenn, the title does say *Water Detector* in the video title and you could interpret that to read "a predefined water level detector") - not a measure! You are correct that you would need two to detect low and high (eg if you wanted to fill a pond once it has dropped to level A but then want to stop filling once it has reached level B. Hysteresis, so to speak!
bangoo has them on sale now, -35% at 11$ hahaha
Good to know!
Ralph, I'm trying to get in touch with you. Sent an email to your webmaster account. Love your stuff.
Yup, sent it straight to the trash bin, John. Just kidding. Reply winging its way to you as we speak.
@@RalphBacon lol. I absolutely understand that trash bin. Plus I own a domain and often don't get stuff sent to that email. When I sent to webmaster @ it didn't bounce but that doesn't mean it is seen. 😁
Your'e only one man time is your enemy.
You're not wrong there. Researching, designing and building take a long time. I need more staff. I've asked Benny if he would do some overtime but he just rolled over and wanted his tummy tickled. No work ethic, that cat.