i want to build this boot too. but much simplier with 3 x 100watt Solar Panel but No batterie. Only one 300 watt motor and a ruder over RC controll. :D Maby i take a little batterie in, with a switch to run with or without, for clouds
Wow, that is seriously cool! Great presentation as well! It's like watching a presentation for a capstone project for an engineering master's program. A completely insane idea would be if this little guy, or one like it, could circumnavigate the globe. Of course, there are loads of edge cases to that, not to mention some pretty serious (not as edge as you'd like) cases.
Outstanding work. Just some suggestions that i faced using similar hardware, replace the dupont jumper cables with something better, try and solder as much as you can to a PCB and secure everything because these things have caused me a lot of pain with improper connections. If you want to upgrade the microcontroller with the same IDE / footprint, you can look into getting a blue pill (STM32F103Cx) which is an ARM clone of the Arduino Nano/Mini. It has 3.3V I/O on most pins so you might need to check that before hooking everything up. You can program it with the same Arduino IDE (look up on youtube how to, some good tutorials exist) and i think most libraries are compatible. For the ESC, please check the thermals and add a heatsink to it because it looks like that had something to do with it burning up like that. And just add an action cam to your boat the next time for that sweet sweet FPV footage. Other than that, have fun. Really insipring to watch. EDIT just saw v2 plans, tl;dr; -custom PCB - this would be the best with better connectors for everything. -sonar depth sensor - JSN-SR04T & A02YYUW, ultrasonic sensors, videos are on youtube -data logging - sd card will help -upgraded microcontroller - bluepill, lets you reuse the code on an arm based mK -heatsink for the ESC / thermal profile under load
A rudder rigged to a DC windshield wiper gearbox downstream from a stationary motor might be more efficient. All I'm saying is using metered voltage per L/R to steer uses energy like a skid steer to turn, but a little worm gear will hold 15° even at the bottom of the ocean
In practice, this boat travels almost entirely in straight lines so both motors are mostly pegged at full throttle with one or the other reducing slightly if it starts to veer off course. Almost all the energy still goes to propelling the boat forwards, it’s not like one motor is braking. This setup allows the electronics to be kept pretty simple because it only has one type of motor and one type of controller. If you used a dc motor for the rudder, you’d need some type of encoder for feedback. You could use a servo too but that would complicate the control and introduce power consumption that doesn’t contribute to thrust due to holding torque requirements and constant minor adjustments. Either way, I’m sure there could be efficiency advantages to using one larger motor instead of two smaller ones, I’d love to see that in practice.
Great job on your build. I just found your vid today so by now you probably have updated the things you talked about. I'm going to check out your channel and see how this project is going. Thanks for sharing.
This morning, I was dreaming up how to take kayak propulsion to the next level (I have 2 12' fishing kayaks) and decided that a PI (or arduino or some other micro computer/controller) controlling a couple ESCs and some large RC boat thrusters on a 100AH lifePo battery could do the trick. You could go to waypoints, set different speeds, turn on a dime, and do an anchor lock. And then, here you are, doing the exact same thing, smaller scale. I'm so excited that somebody else (You) did the proof of concept for me! And you are in my old stomping grounds (I was in NW Smith county until 2020) In any event, I'm going to start specing it out and making sure the sizing is right, and start the build. It will be fun, and I expect it could work really well. I'm a software engineer, and my son and I have some good robotics experience (His team went to World's his senior year) Still it is going to be a challenge, and one screw up can destroy a $50 part. Anyway, if you want to rubber duck some ideas, I'd be delighted to assist. I really look forward to the next version you build. Richard Hasting, Canyon Lake
Sounds like an awesome project. A much more expensive version of this lol. And yeah this neck of the woods has tons of little ponds/lakes that are awesome for this type of thing. I’ve thought about building a larger scale version that could tolerate waves and saltwater to run in the ocean.
@@Clay-Builds just modify for the pelican, golden!! My current pond is 8700 acres, so I could really use some mechanical help to propel me through the water. Paddling for a mile ain't fun... That said, to be truly capable of doing this, I think I need reverse as well as forward. Does an ecs with a gpio controlling it, have the ability to go in reverse, or do I need to use relays to flip the polarity to the motor?
@@richardhasting6046 some brushless motor ESC’s can dynamically switch between forward and reverse. Pretty much all brushed motor controllers can easily switch direction, but brushed motors tend to be more water sensitive.
Separating the propellers farther will increase leverage for differential thrust, decrease power wasted needed to complete turns or heading corrections, decrease fluid flow interference because they are too close to each other, and greater separation will also increase steering accuracy in general...
Yes, it would require less of a thrust differential to rotate if the props were further apart. They’re spaced as they are because I made them as large a diameter as possible while fitting between the pontoons. To spread them apart, they’d have to either be deeper, or behind/outside the pontoons. Since this vehicle is almost always traveling in a straight line, I wasn’t very concerned with this and decided to keep it compact. The time lapses towards the end of the video show it keeping on a straight line remarkably well. Larger moment arms on the thrusters would require finer throttle adjustment to prevent overdamping the steering. Once I implement datalogging, I’ll be able to tell exactly how much/often it’s modulating the throttles to keep straight.
@@Clay-Builds Gotcha, also I was thinking you could decrease drag, and separate the thrusters by turning both of the pontoons into partial pontoons/partial ducted fan jets. If there were a way to upload drawings here in the comments, I'd show you what I mean...
Wow! That’s impressive man! There are all kinds of things running through my head after seeing this! I saw this through an article on google, and was shocked to see that you only had 1K views! I’m sure this will go viral though because it is simply brilliant! With your knowledge, & tremendous expertise, there’s no doubt in my mind that you won’t work out all the kinks, and have a working (patentable) prototype soon! I’m not necessarily into all of the “coding, & didn’t have clue about all of the “computerese,” & coordinates & such, but I subscribed just to see where you go with this in the future! Bravo! 👍🏽
Such as smuggling cocaine from holland to England without to much effort xD, let the other party reflash the device to make it come back or stack them for a pickup. I understand why YT shadow hides this but only shows it to nerds 🤣🤣
It’s probably worth reading about ‘cavitation’. You’ll probably get better speed and battery life with that setup. Maybe longer thrust tubes will be enough.
I don’t think my props are spinning fast enough to cause cavitation, but definitely something to keep in mind when I optimize them better. I’m also adding in current draw sensors and data logging so that I can quantitatively compare efficiencies of different propellor designs.
First very nice project - I would do the following improvements: - MPPT Charge controller instead of that PWM one (more efficient!) For your usage you could use an EPEVER Tracer 2606BP. The battery gets charged then at its optimum. The PWM Charger wastes a lot of energy as heat. Further you can use it for gel, flooded or LiPo/LiIon Batteries -> USB to RS485 cable. - Replace your solar panel for a more 'solid' one with an aluminium frame: Better heat management, more efficient and a good stiffness. A 30w or better 50w one should be doing its job great. Regards from germany 😉
I’ve already replaced the solar charger with an MPPT one (genasun gv-5) which is more compact/reliable/efficient than the first one I was using. And I’ll probably add more solar panels at some point, but this one keeps up well enough on a nice sunny day.
Fascinating! I do have one concern. This design might work well-enough on most lakes, but in oceans the breaking waves could flip it over, killing the solar power. You might add a weighted keel to make it self-righting. The keel would also add more stability and directionality. As on a sailboat, it would lessen the need for the motors to constantly alter the direction.
If your inlet shrouds were long cones you'd have a layer of self-cleaning capability. This will come in handy if you keep testing on lakes. Surround the panel with the sort of spikes used to keep pigeons from landing on ledges. Having a third, smaller thruster would give you better handling when you want to loiter in one spot (to lower a camera into the depths maybe). Adding ballast to the bottom of each float will increase stability. I assume they are hollow otherwise. I'd consider sizing the floats so they can be filled with foam. That way it will still float after being bitten by a gator/great white!
Yea this flat prop shroud design caused a lot of excess drag and it still got plants stuck to it a few times. For now I’m omitting the prop shroud entirely on the new version I’m building, but I’ll likely implement it in the future. I think a roughly conical wireframe design would perform much better. Also the props themselves could be optimized to prevent catching plants and trash.
Great project. I enjoyed watching it. Only one observation. It is always better to have the props submerged enough to not cause the props to cavitate. You will get a huge improvement in efficiency if they don't cavitate.
Thanks. In the closeup clip of the props running in the bathtub, the water level was too shallow and the shrouds were sitting on the bottom. In deep enough water, the top of the shrouds sit about 3/4” below the surface
@@Clay-Builds That's enough. It just looked like it was cavitating. Another thing to look at is any obstruction in front of the prop. Like your guard to wade off obstacles. A prop works the same as a wing and the best results are obtained from smooth flow over the foil. Any disturbance or vortices will impact the efficiency. If you find that you don't get the drive or speed for the input that you are giving you could change the grill in front of the props to be cone shaped with each member shaped with a sharp leading edge. Like little displacement boat hulls. That way the flow of the water has no or very little disturbances.
That looks like a fun project, and useful, too. Projects like your/this have a lot of different disciplines to pull together, and not easy, from the physical construction, to the software, to the data interpretation. So, congratulations. Perhaps the thruster box/deck mount could be made adjustable in height, to maintain a nice low center of gravity, like you have, but also to compensate for weight without using ballast. Alternatively, you could make a ballast tank, and fill it with water. Perhaps you could slip in a rubber grommet (not the Harbor Freight ones) where the motor wires penetrate the shrouds. A Teensy 4.1 would give you more processing power for the same code. Have you checked out Paraview? As far as the props, you could also do a study with OpenFoam.
Now, time for the Columbia to Miami deviance test. 😂 Include a 360 camera alert system and an anti buoyancy system to sink it temporarily for passersby
Clay, I was concerned about the ESC's failing as soon as you placed them in the sealed container with all other hardware. If you could go with a waterproof ESC or just mount them out of the area of risk, I think that would be a help. Also, in the V2, you may want to add a small threaded connection to each pontoon, to allow moisture to drain. I know that in theory, no moisture should be able to enter the sealed system; however, I have found that it most certainly will (if only over time). Also, there have been dozes of DIY builds very similar to yours for purposes of bathymetry. If you're not familiar with bathymetry (essentially topography in water) you will find a great deal of similar builds and great sonar options for your build. Like most autonomous rovers (land or water), most of the other DIY units use Pixhawk to save the added hassle. You could go with the less expensive APM; however, a Pixhawk flight controller would be better for your needs. Also, thank you for the excellent, comprehensive video!
Very nice. Keeping the props inside is good for forward and structural. Have you thought of putting these outside the pontons to increase turn speed and stability. The idea of sonar is excellent, you could map out the best fishing holes in the lake.
Thanks. The stability is good as is. Putting the props further apart would improve the turning radius, but it’s good enough for what I’m using it for and I wanted to keep it compact.
Call it hydronaut! Great idea,unfortunately my gut feeling says it will be toppled by big waves and then die.. try to do a cylinder that always rolls right. Also I'm pretty sure the PV is too small. And then maybe a rotateable sail plus keel would be a great idea, in place of the motors that still can be entangled.. Use LoRa / TTN to still transmit data worldwide..
Awesome work! I like the amount of math you went through to ensure it stayed bouyant. My question is: Was there a design reason that the motors only go forward with no reverse capability? This question was bugging me throughout the video. As a suggestion, have you considered making a thin keel under each pontoon to just past the depth of the propeller housing? It might help protect the shroud from impacts when running aground whilst also directing the water flow under the boat in a straight line
Adding keels is a good idea. I’ll be redesigning the parts that mount the prop shrouds anyway, it’s definitely a weak point. The ESC’s I’m using actually can’t dynamically switch directions. To make them reverse you have to shut them down, reprogram them, and restart. This is theoretically possible to do with an arduino, but it would be cumbersome and slow. If the props could reverse I’d be able to make tighter turns, but that isn’t really necessary for this application. Although it would be helpful to back off of an obstacle if it got stuck.
I would of used a Airplane Prop type propulsion. A big problem with these boats is seaweed plugs up the motor. Its also simpler all you need is a directional vane to control it.
There’s no design reason for that. It’s just a consequence of the way they’re modeled, surfaces lofted through tangent profiles with the last one being a point. This prop design is somewhat arbitrary, I plan to collect some real world data from a few different designs to optimize for power efficiency.
@@Clay-Builds Ah, that's interesting. I don't think I've ever lofted "last one being a point". I'm wondering if props being an Archimedes screw in a tube would be a cavitation free, hard to injure, hard to stop with weeds, drive system. Thanks!
The ESC’s I was using when this video was filmed were not bidirectional, so no way to reverse them without physically swapping wires around. I’ve also since ditched the 0-180 degree servo logic completely, it makes more sense to just use discrete PWM value in microseconds, 1000-2000 in this case.
Full marks for effort and programming! However for the next version, you've got to make it more hydrodynamic so that it's not just a weed collector! Lose the propguards and use or make proper swept marine propellers that don't snag, or try using impellers like jet skis. I would make proper catamaran hulls out of glass or carbon fiber and mount the motors internally in each one.
@@godwinbadu-marfo7577 Adafruit LSM303AGR is the compass I used. It also includes an accelerometer which I’m not using for anything in this project. The solar panel is just some cheap 12v panel off eBay, I think they’re all about the same. It was advertised as a 200W panel but in reality it’s closer to 50W
Esp 32 has more resources. The power consumption inside watter will be much greater than on land. Maybe that overloaded the esd. Have you checked the temperature inside the box after a long time running in the sun? this components may need cooling.
Im sure the heat will build up overtime. ~25 minutes of continuous use did not make the ESC’s noticeably warm to the touch, but this could be a problem if it runs for several hours. The next version will include current and voltage monitoring so I’ll be able to see exactly how much load is on the motors and how much solar power is being drawn. I don’t think these motors are anywhere near the 30A nominal limit, but I’d like to upgrade the ESC’s anyway.
Use sensorless brushless motors. For the most part, they’re inherently pretty waterproof and will have no electrical issues operating fully submerged. You’ll still need to worry about corrosion over time, but unlike typical brushed motors, there’s nothing inside them that could easily short out when wet. You’ll want to choose a motor with a relatively low KV value, because high torque is needed to drive a prop through water, and you don’t need very high RPM. Many brushless motors, including the ones I used in this project, are specifically intended for underwater use. I didn’t do anything special to seal them or anything.
Looks nice - if you're using the Arduino servo output to control it via PWM, does that mean it never gets up to 100% per the way that signal is formatted? Would analog output be a better option? Do you store GPS coordinates on the card before launching to set waypoints? Or does it store GPS coordinates for reading later, or both?
The servo library does use analog write to output a pwm signal to the throttle input on the esc. It’s easier to use the servo library rather than directly using analogwrite because the servo function takes a “position” value (0 corresponds to no throttle, 180 full throttle) and outputs the correct pulse width accordingly. Analogwrite takes a duty cycle value between 0-255, I’d have to add some extra calculations to convert from pulse width to duty cycle. Esc looks for a pulse width of 1000us (zero throttle) up to 2000us (full throttle). So I just use the servo library because it does that conversion for me. It can output a full throttle signal, but I have it limited to 40% because the motors would stall at higher speeds when under load in the water. Currently it only reads pre-determined coordinates from the card, no writing. But I want to implement real time data logging so the sd card would be like a “black box” with lots of different parameters for optimization and troubleshooting.
Very cool. I wish I was knowledgeable about building stuff like that. I got an idea to stuff light weight foam in the pipes that way if the pipe leaks and floods the whole pipe won't fill up with water. BTW Clay I suggest watch a 13 part series called _What on earth happened_ by Ewaranon to learn that the earth is not a globe. I got it in my about tab.
Could you send it from the East coast to England? That would be too cool, huh? Charge and pump for however long it would take and keep up with it's location via GPS. Have to be made for salt water resistance.
Seems like you went thru alot of trouble waterproofing and mounting the box for the battery and electronics. Was there a particular reason for not storing the components in the water proof pontoons themselves?
I need easy access to the electronics so that I can troubleshoot, reprogram, remove SD card, charge battery, etc. The box with an openable lid is super convenient for this. The pontoon end caps are sealed shut with PVC cement. I didn’t want to put any holes in them and risk a leak since they’re critical for buoyancy.
You still using the 5V of one esc to power the 5V stuff? Get a $1 switching buck converter instead, since the esc use a linear regulator that gets hot and may die. A nice arduino compatible processor upgrade would be the esp8266 / esp32. I did use esp-now to replace the RC radio. See the quadcopter on my channel.
Im putting in a custom PCB and upgrading to a teensy LC. It’ll still use the ESC to power the teensy Vin, but the rest of the components will be running 3.3v off a regulator.
@@Clay-Builds Dont. The 5V reg of the esc is a "M" type max 500ma. Depending on bat voltage this will get very hot. At least use 5V from both esc. Brtter get a 3A 5V buck for that, 90ct on ali. No heat, no problems, solid 5V rail. The 5V->3.3V Vin reg is good, keeps 5V noise from the 3.3V rail. Keep that.
Nice stuff, looking forward to getting part2 with depth map. had thought of making similar project but never took time to do it. also I want to build too many things at same time which is very unproductive. Any link to waterproof box?
why not make that motor mounts slightly rounded. also add telemetry 4G connection then live video? nice to have features. also why not use lipo battery then you more power less weight. love it so far!
Fascinating code as well, serial.print("Fuck") during debugging when things were not working xD P.s. nice project love the hardcoded lat/long makes sure it will never stops until it gets at destination, how ever the hot glue will make sure that it won't make it from Holland to England without sinking somewhere because it's not really a material to rely on
Dude this is awesome! I am working on a custom auto pilot too but it's based on the RP2040 and micro-python. Looking forward to the new radios. Any plans for telemetry?
Nice work😃 I like the idea a lot and it seems to work really nicely. I just fear it can't do really large distances as the motors are a bit weak? I fear they could collect some sea weeds or fishing nets etc. Hard to work around such issues..
Yeah the main reason I used prop shrouds is to prevent plants/trash from getting caught up. More powerful motors would need a much bigger solar panel, but I’d love to do a larger scale version eventually.
Interesting build, Can I ask how you have made the solar panel ocean proof water resistant from waves smashing it and the epoxy type with regard to salt water?
The solar panel is waterproof, it’s intended to withstand rain. In practice it will stay totally dry unless there’s a large wave. Not sure how it would hold up to saltwater though. The epoxy I used is just gorilla glue brand 2 part 30 minute epoxy
@@Clay-Buildsthanks for reply, I will copy the code from the video or there is somewhere to download? I installed the other code with the sonar and takes 96% of 328ATmega ... there is any way to send you the progress of the progect ?
@@zuzuniagrikubanos6473 I have a 2nd video posted with some updates to this project, it has a GitHub link where you can download the code. That version has a lot more functionality than what’s shown in this video, but you could remove a lot of what you don’t need to save memory. The data-logging, sonar stuff, error detection, etc. isn’t totally necessary if you’re just wanting to build a simple boat that drives around. The code shown in this video used almost 90% of the memory the arduino pro mini (atmega 328P) IIRC. A huge portion of that was due to the libraries I used for the SD card, magnetometer, and GPS module. Maybe you could modify or find some alternative libraries to save memory, but I ended up just upgrading to a better microcontroller.
@@Clay-Builds I already removed some they are what I left on the code = #include #include #include #include #include #include #include #include #include "pindefinitions.h" #include "miscfuncs.h" #include "constvars.h" #include "manualmode.h" #include "automode.h" #include "cardlogger.h" #include "navigation.h" they are what I left on the code.I will start the connection tomorrow,,the correct pins is on GitHub link ??
@@zuzuniagrikubanos6473 the pins in that code are correct for the teensy micro controller I used in the 2nd video. You might need to make some slight changes depending on what microcontroller you’re using. There’s also an electrical schematic in the GitHub link that shows which pins connect to what. Make sure the throttle signals are connected to PWM capable pins, compass is wired to I2C pins, etc. you can mostly switch things around to whichever pins you want, but I’d make sure the timing sensitive components like the GPS serial connection, receiver communication pins that use the pulseIn() function, and the PWM outputs should each have their own timers.
The motors themselves are not sealed against water ingress. I used hot glue around the soldered joints on the back of the motor to prevent shorts between the individual phase wires. Also hot glue to fill in the drilled holes where the motors wires pass through into the electronics box. I’d use a proper potting compound if I were to redo it, but never caused a problem as it was in this video. Also notably, the pass through into the electronics box is not submerged, just keeping splashes out. Surely it would leak if below the waterline.
@@Clay-Builds Thanks for the fast and good reply! I'm doing my first underwater tests with electronics right now and I'm surprised how well hot glue seems to work all the time.
Impressionnant félicitation, si tu souhaite le tester tu peut essayer de lui faire traverser l'Atlantique, je le réceptionne en France et te le renvoi par DHL. En espérant qu'il ne finisse pas comme le Titanic... A oui j'allais oublier, très bonne référence le livre au début, j'ai lu la version française, très philosophique.
@@Clay-Builds Je vous jure que j'en suis capable, du champagne grand cru avec une bouteille de Bordeaux en cadeau. Je n'ai jamais gouté le whisky américain ce serait une grande occasion de le déguster, et avec 0 frais de douane!
I am working on a boat like this too! Is there any way you can give my the 3d model for the clamp that holds the pipes to the solar panel? Thank You in Advance and great build!
Your autopilot relies on heading too much. Look at aviation autopilots for methodology. Even 40yo aviation A/Ps will used tracking for control, not heading. Heading will always be wrong unless water current and winds are perfectly inline with path. You need to computer the vector(desired track) from start point to end point and the adjust heading to stay on the desired track. You need to correct for wind and water flow. To avoid hitting objects, the desired track must be followed, otherwise you will end up hitting the banks or running aground. As you follow the desired track, the heading will face in a direction that compensates for the wind and water flow. For example, a desired track of 180 and a wind of 270 might result in a heading between 180 and 270 depending on wind speed and water flow.
That’s true, in the presence of a strong lateral wind/current, it would make a wide arc towards the waypoint rather than a straight line. In the video I only used 4 waypoints, if I added a lot more between each of the major ones then it would greatly reduce this effect. The navigation currently assumes the direction of travel is always exactly in the direction the compass is pointing. I could also get the direction of travel from the gps module and use that for steering. That way it could steer slightly into the current/wind and keep the direction of travel exactly towards the waypoint. Only issue is I don’t think the gps heading is as accurate as the compass, especially at low speed.
Dude do you have any wiring diagram between the solar, charge controller, and ESCs? I'm building a similar project and would like to see what you did. Haven't been able to find anything helpful online. Thanks.
I have a wiring diagram posted here: www.clay-p-mcpherson.com/projects/solardrone But it doesn’t show the solar charge controller because it’s honestly pretty self explanatory. A typical Solar charge controller will have 3 pairs of +/- terminals labeled “battery”, “load”, and “panel”. The “battery” connections just goes straight to the battery terminals, + to red - to black. The “panel” connections go straight to the output wires for the solar panel. My panel has a barrel jack style output connector, I stripped back its wires (there’s only 2, a + and a -), and connected them straight into the “panel” connections on the controller, polarity matters. The input side of both my ESC’s (labeled 12vin on my diagram) go to the “load” connections on the controller; they’re wired in parallel. So both red wires to to the +, both black wires go to the -. That’s all that’s plugged into the charge controller. The ESC’s each have their own DC/DC converter with 5v output that I use to power the microcontroller and everything else. Check my Chanel page, I recently posted another video about this boat which shows more footage of the wiring and upgrades. If you want to add an on/off switch, it’s best to put it in the positive side of the “load” wire, that way the battery can still charge when it’s off but the electronics won’t drain it.
@@Clay-Builds thanks, I was doing a parallel type systems, but I guess it makes way more sense and cheaper to just single up and put the ESCs in parallel.
@@Clay-Builds @Clay Builds if you have time for another question. I have some 3s batteries with 2 30 amp ESCs to some little submersible brushless motors. And have already hooked them up to a remote for throttle and differential thrust for turning. I understand what you typed above in regards to the charge controller. But I guess I have a question in regards to solar charging 3s batteries. I have a bunch of 5v 0.5 amp panels and was going to do a combination of series/parallel into the charge controller. The 3s are also 11.1v so I'm not sure how it would react with any given charge controller. Should I just redo my design using 12v batteries, and 12+ volt panel?
@@esotericcommonsense6366 it depends on your charge controller. The controller will have a specific battery voltage it is intended to operate with, some have adjustable settings with more flexibility. it’s important to use a battery of the same nominal voltage as what the charge controller is set to. The controller reads the battery voltage and that’s how it decides when to start or stop charging. If the controller is expecting a nominally 12v lead acid battery, (which is actually around 13.1v fully charged) it might try to overcharge a 3s pack which is closer to 12.5v when full. A 3s pack can also operate at much lower voltage than a 12v sla, so the controller might think it’s dead when it really isn’t. I’d either use a charge controller intended for a 3s lithium battery or switch to a 12v nominal SLA if you’re using a 12v controller like mine. As for the panels, the voltage isn’t as picky. But depending on the controller, it probably needs to be higher than 5v. Look up what the operable panel voltage range is for your controller. If you’re using 5v cells, you could always wire them in series sets of 3 to effectively get a 15v panel.
It wasn’t great. I swapped it for a GENASUN GV-5, a more reliable and efficient mppt controller. Don’t think I showed it on camera, but it was installed during the test run.
Can I pay you to help me do the same project? Just taking over the internet, being a guiding hand sorta thing. I'm looking to do the microtransat challenge
Not interested. But there’s enough information in these videos to totally recreate this project. Check the newest video too, you can download my latest code.
Enough info if you know what he’s talking about. Consulting is something you consider later in life. I’m trying to put together a solar raft but way bigger than this. I’m going to buy a sailboat soon and make it autonomous and able to cross the Atlantic way faster than average.
This is still an ongoing project. I have some improvements to make, and I’ll clean up the code a little. I don’t have a repo yet but I do plan to post the code/cad/ and other data somewhere soon. Any specific details you’re interested in?
Haha I’m not interested in selling anything. This project is totally open source however. See the newer videos for improved code and electronics information.
@@ehsankf vegetation is pretty unavoidable which is why I always use some type of prop shroud for small scale boats. Your best bet for above the surface obstacles is probably LiDAR but that gets expensive quick.
Well that would’ve been too easy and would’ve required a more expensive flight controller. Plus I plan for it to have a lot more functionality than just autonomous driving.
Fascinating custom autopilot! Thanks for sharing
Thanks a lot. I’m a fan of your channel, you’ve got some awesome projects.
@@Clay-Builds Nice for drogs smuggling :D
i want to build this boot too. but much simplier with 3 x 100watt Solar Panel but No batterie. Only one 300 watt motor and a ruder over RC controll. :D
Maby i take a little batterie in, with a switch to run with or without, for clouds
Put you a camera on it and tell it to circle the world
I bet that pretty kitty had a big hand (well, paw) in helping you get your gumption back 🙂
An essential design consultant
beware of ducks. They try to get onto stuff like that.
Wow, that is seriously cool! Great presentation as well! It's like watching a presentation for a capstone project for an engineering master's program. A completely insane idea would be if this little guy, or one like it, could circumnavigate the globe. Of course, there are loads of edge cases to that, not to mention some pretty serious (not as edge as you'd like) cases.
Thanks. It would be awesome to build an ocean version at some point.
Outstanding work. Just some suggestions that i faced using similar hardware, replace the dupont jumper cables with something better, try and solder as much as you can to a PCB and secure everything because these things have caused me a lot of pain with improper connections.
If you want to upgrade the microcontroller with the same IDE / footprint, you can look into getting a blue pill (STM32F103Cx) which is an ARM clone of the Arduino Nano/Mini. It has 3.3V I/O on most pins so you might need to check that before hooking everything up. You can program it with the same Arduino IDE (look up on youtube how to, some good tutorials exist) and i think most libraries are compatible.
For the ESC, please check the thermals and add a heatsink to it because it looks like that had something to do with it burning up like that.
And just add an action cam to your boat the next time for that sweet sweet FPV footage. Other than that, have fun. Really insipring to watch.
EDIT just saw v2 plans, tl;dr;
-custom PCB - this would be the best with better connectors for everything.
-sonar depth sensor - JSN-SR04T & A02YYUW, ultrasonic sensors, videos are on youtube
-data logging - sd card will help
-upgraded microcontroller - bluepill, lets you reuse the code on an arm based mK
-heatsink for the ESC / thermal profile under load
Cool build. I partucularly like that you reviewed what worked as expected and what might be improved.
Thank you. Definitely plenty of improvements to be made.
A rudder rigged to a DC windshield wiper gearbox downstream from a stationary motor might be more efficient.
All I'm saying is using metered voltage per L/R to steer uses energy like a skid steer to turn, but a little worm gear will hold 15° even at the bottom of the ocean
In practice, this boat travels almost entirely in straight lines so both motors are mostly pegged at full throttle with one or the other reducing slightly if it starts to veer off course. Almost all the energy still goes to propelling the boat forwards, it’s not like one motor is braking. This setup allows the electronics to be kept pretty simple because it only has one type of motor and one type of controller. If you used a dc motor for the rudder, you’d need some type of encoder for feedback. You could use a servo too but that would complicate the control and introduce power consumption that doesn’t contribute to thrust due to holding torque requirements and constant minor adjustments.
Either way, I’m sure there could be efficiency advantages to using one larger motor instead of two smaller ones, I’d love to see that in practice.
Great job on your build. I just found your vid today so by now you probably have updated the things you talked about. I'm going to check out your channel and see how this project is going. Thanks for sharing.
great design, great idea, i think you should just protect the motors against the ocean Algae, it will disturb/annoy the motors, or may stop it,
This morning, I was dreaming up how to take kayak propulsion to the next level (I have 2 12' fishing kayaks) and decided that a PI (or arduino or some other micro computer/controller) controlling a couple ESCs and some large RC boat thrusters on a 100AH lifePo battery could do the trick. You could go to waypoints, set different speeds, turn on a dime, and do an anchor lock. And then, here you are, doing the exact same thing, smaller scale. I'm so excited that somebody else (You) did the proof of concept for me!
And you are in my old stomping grounds (I was in NW Smith county until 2020) In any event, I'm going to start specing it out and making sure the sizing is right, and start the build. It will be fun, and I expect it could work really well.
I'm a software engineer, and my son and I have some good robotics experience (His team went to World's his senior year) Still it is going to be a challenge, and one screw up can destroy a $50 part.
Anyway, if you want to rubber duck some ideas, I'd be delighted to assist.
I really look forward to the next version you build.
Richard Hasting, Canyon Lake
Sounds like an awesome project. A much more expensive version of this lol. And yeah this neck of the woods has tons of little ponds/lakes that are awesome for this type of thing. I’ve thought about building a larger scale version that could tolerate waves and saltwater to run in the ocean.
@@Clay-Builds just modify for the pelican, golden!! My current pond is 8700 acres, so I could really use some mechanical help to propel me through the water. Paddling for a mile ain't fun...
That said, to be truly capable of doing this, I think I need reverse as well as forward. Does an ecs with a gpio controlling it, have the ability to go in reverse, or do I need to use relays to flip the polarity to the motor?
@@richardhasting6046 some brushless motor ESC’s can dynamically switch between forward and reverse. Pretty much all brushed motor controllers can easily switch direction, but brushed motors tend to be more water sensitive.
i like the simplicity of your boat design. im gonna try the same, but with ardupilot and a flightcontroller and a companion computer
Thanks, ardupilot would definitely be a much easier way to automate it.
I like the commented debug log on line 156 a lot!
Nice build! Looking forward to watching this project grow!
Awesome. Thanks for sharing. I recommend redundancy for steering with a rudder incase you lose a motor so you don’t get stuck going in circles.
Samson I suggest watch a 13 part series called _What on earth happened_ by Ewaranon to learn that the earth is not a globe.
I got it in my about tab.
Separating the propellers farther will increase leverage for differential thrust, decrease power wasted needed to complete turns or heading corrections, decrease fluid flow interference because they are too close to each other, and greater separation will also increase steering accuracy in general...
Yes, it would require less of a thrust differential to rotate if the props were further apart. They’re spaced as they are because I made them as large a diameter as possible while fitting between the pontoons. To spread them apart, they’d have to either be deeper, or behind/outside the pontoons. Since this vehicle is almost always traveling in a straight line, I wasn’t very concerned with this and decided to keep it compact. The time lapses towards the end of the video show it keeping on a straight line remarkably well. Larger moment arms on the thrusters would require finer throttle adjustment to prevent overdamping the steering. Once I implement datalogging, I’ll be able to tell exactly how much/often it’s modulating the throttles to keep straight.
@@Clay-Builds Gotcha, also I was thinking you could decrease drag, and separate the thrusters by turning both of the pontoons into partial pontoons/partial ducted fan jets. If there were a way to upload drawings here in the comments, I'd show you what I mean...
Really methodical work and nicely documented - well done!
Thanks
Wow! That’s impressive man! There are all kinds of things running through my head after seeing this!
I saw this through an article on google, and was shocked to see that you only had 1K views! I’m sure this will go viral though because it is simply brilliant!
With your knowledge, & tremendous expertise, there’s no doubt in my mind that you won’t work out all the kinks, and have a working (patentable) prototype soon!
I’m not necessarily into all of the “coding, & didn’t have clue about all of the “computerese,” & coordinates & such, but I subscribed just to see where you go with this in the future!
Bravo! 👍🏽
Thanks haha, I’ll post another video once I’ve made some improvements. Not sure what article you’re referring to.
Such as smuggling cocaine from holland to England without to much effort xD, let the other party reflash the device to make it come back or stack them for a pickup. I understand why YT shadow hides this but only shows it to nerds 🤣🤣
soo much good information - thank you.
It’s probably worth reading about ‘cavitation’.
You’ll probably get better speed and battery life with that setup.
Maybe longer thrust tubes will be enough.
I don’t think my props are spinning fast enough to cause cavitation, but definitely something to keep in mind when I optimize them better. I’m also adding in current draw sensors and data logging so that I can quantitatively compare efficiencies of different propellor designs.
Nice project! I like the idea of using pipes for buoyancy :)
Thanks. They’re cheap and easy, but a little on the heavy side compared to foam or fiberglass.
I do love Arduino. This is a beautiful build. Thanks for pushing through and sharing this with us. :)
Absolutely
First very nice project - I would do the following improvements:
- MPPT Charge controller instead of that PWM one (more efficient!)
For your usage you could use an EPEVER Tracer 2606BP.
The battery gets charged then at its optimum. The PWM Charger wastes a lot of energy as heat.
Further you can use it for gel, flooded or LiPo/LiIon Batteries -> USB to RS485 cable.
- Replace your solar panel for a more 'solid' one with an aluminium frame: Better heat management, more efficient and a good stiffness. A 30w or better 50w one should be doing its job great.
Regards from germany 😉
I’ve already replaced the solar charger with an MPPT one (genasun gv-5) which is more compact/reliable/efficient than the first one I was using. And I’ll probably add more solar panels at some point, but this one keeps up well enough on a nice sunny day.
Awesome build.. looking forward for version 2.
Thank you
Fascinating! I do have one concern. This design might work well-enough on most lakes, but in oceans the breaking waves could flip it over, killing the solar power. You might add a weighted keel to make it self-righting. The keel would also add more stability and directionality. As on a sailboat, it would lessen the need for the motors to constantly alter the direction.
This is only intended for lakes/ponds, although I’d love to build an ocean version at some point.
needs a "turtle mode" flip-over mechanism
It's brilliant! You're brilliant! Keep up the good work.
Thank you
If your inlet shrouds were long cones you'd have a layer of self-cleaning capability. This will come in handy if you keep testing on lakes. Surround the panel with the sort of spikes used to keep pigeons from landing on ledges. Having a third, smaller thruster would give you better handling when you want to loiter in one spot (to lower a camera into the depths maybe). Adding ballast to the bottom of each float will increase stability. I assume they are hollow otherwise. I'd consider sizing the floats so they can be filled with foam. That way it will still float after being bitten by a gator/great white!
Yea this flat prop shroud design caused a lot of excess drag and it still got plants stuck to it a few times. For now I’m omitting the prop shroud entirely on the new version I’m building, but I’ll likely implement it in the future. I think a roughly conical wireframe design would perform much better. Also the props themselves could be optimized to prevent catching plants and trash.
Great project. I enjoyed watching it. Only one observation. It is always better to have the props submerged enough to not cause the props to cavitate. You will get a huge improvement in efficiency if they don't cavitate.
Thanks. In the closeup clip of the props running in the bathtub, the water level was too shallow and the shrouds were sitting on the bottom. In deep enough water, the top of the shrouds sit about 3/4” below the surface
@@Clay-Builds That's enough. It just looked like it was cavitating. Another thing to look at is any obstruction in front of the prop. Like your guard to wade off obstacles. A prop works the same as a wing and the best results are obtained from smooth flow over the foil. Any disturbance or vortices will impact the efficiency. If you find that you don't get the drive or speed for the input that you are giving you could change the grill in front of the props to be cone shaped with each member shaped with a sharp leading edge. Like little displacement boat hulls. That way the flow of the water has no or very little disturbances.
That's a very interesting and good design in my opinion!
Thanks
That looks like a fun project, and useful, too.
Projects like your/this have a lot of different disciplines to pull together, and not easy, from the physical construction, to the software, to the data interpretation. So, congratulations.
Perhaps the thruster box/deck mount could be made adjustable in height, to maintain a nice low center of gravity, like you have, but also to compensate for weight without using ballast. Alternatively, you could make a ballast tank, and fill it with water.
Perhaps you could slip in a rubber grommet (not the Harbor Freight ones) where the motor wires penetrate the shrouds.
A Teensy 4.1 would give you more processing power for the same code.
Have you checked out Paraview? As far as the props, you could also do a study with OpenFoam.
I came here, bc thumbnail looked like another rctestflight video :)
Cool build anyways.
Now, time for the Columbia to Miami deviance test.
😂
Include a 360 camera alert system and an anti buoyancy system to sink it temporarily for passersby
So rad!!!! You should see if it can make it down the Mississippi
Clay, I was concerned about the ESC's failing as soon as you placed them in the sealed container with all other hardware. If you could go with a waterproof ESC or just mount them out of the area of risk, I think that would be a help.
Also, in the V2, you may want to add a small threaded connection to each pontoon, to allow moisture to drain. I know that in theory, no moisture should be able to enter the sealed system; however, I have found that it most certainly will (if only over time).
Also, there have been dozes of DIY builds very similar to yours for purposes of bathymetry. If you're not familiar with bathymetry (essentially topography in water) you will find a great deal of similar builds and great sonar options for your build.
Like most autonomous rovers (land or water), most of the other DIY units use Pixhawk to save the added hassle. You could go with the less expensive APM; however, a Pixhawk flight controller would be better for your needs.
Also, thank you for the excellent, comprehensive video!
Very nice. Keeping the props inside is good for forward and structural. Have you thought of putting these outside the pontons to increase turn speed and stability. The idea of sonar is excellent, you could map out the best fishing holes in the lake.
Thanks. The stability is good as is. Putting the props further apart would improve the turning radius, but it’s good enough for what I’m using it for and I wanted to keep it compact.
Call it hydronaut!
Great idea,unfortunately my gut feeling says it will be toppled by big waves and then die.. try to do a cylinder that always rolls right. Also I'm pretty sure the PV is too small.
And then maybe a rotateable sail plus keel would be a great idea, in place of the motors that still can be entangled..
Use LoRa / TTN to still transmit data worldwide..
Awesome work! I like the amount of math you went through to ensure it stayed bouyant.
My question is: Was there a design reason that the motors only go forward with no reverse capability? This question was bugging me throughout the video.
As a suggestion, have you considered making a thin keel under each pontoon to just past the depth of the propeller housing? It might help protect the shroud from impacts when running aground whilst also directing the water flow under the boat in a straight line
Adding keels is a good idea. I’ll be redesigning the parts that mount the prop shrouds anyway, it’s definitely a weak point.
The ESC’s I’m using actually can’t dynamically switch directions. To make them reverse you have to shut them down, reprogram them, and restart. This is theoretically possible to do with an arduino, but it would be cumbersome and slow.
If the props could reverse I’d be able to make tighter turns, but that isn’t really necessary for this application. Although it would be helpful to back off of an obstacle if it got stuck.
I would of used a Airplane Prop type propulsion. A big problem with these boats is seaweed plugs up the motor. Its also simpler all you need is a directional vane to control it.
Yea that would definitely help with tangles but might sacrifice some power efficiency
Lotta nice thought-out comments here! I was curious why the prop tips have kind of a spoonbill effect on the ends?
There’s no design reason for that. It’s just a consequence of the way they’re modeled, surfaces lofted through tangent profiles with the last one being a point.
This prop design is somewhat arbitrary, I plan to collect some real world data from a few different designs to optimize for power efficiency.
@@Clay-Builds Ah, that's interesting. I don't think I've ever lofted "last one being a point".
I'm wondering if props being an Archimedes screw in a tube would be a cavitation free, hard to injure, hard to stop with weeds, drive system.
Thanks!
For differential thrust throttle values (0-180), have you tried using reverse thrust as the range for each motor.
So, -180 to 0 to 180...
The ESC’s I was using when this video was filmed were not bidirectional, so no way to reverse them without physically swapping wires around. I’ve also since ditched the 0-180 degree servo logic completely, it makes more sense to just use discrete PWM value in microseconds, 1000-2000 in this case.
Full marks for effort and programming!
However for the next version, you've got to make it more hydrodynamic so that it's not just a weed collector!
Lose the propguards and use or make proper swept marine propellers that don't snag, or try using impellers like jet skis.
I would make proper catamaran hulls out of glass or carbon fiber and mount the motors internally in each one.
Very good job! Please further details about motors. Thanks.
Got them here:
VGEBY1 RC Brushless Motor,... www.amazon.com/dp/B07W8TXD7L?ref=ppx_pop_mob_ap_share
@@Clay-Builds thank you!
We're very fascinated by your great effort and endurance. Thanks indeed. Please would you share the software and some guide?
I’ll put the software on GitHub once I’ve cleaned the code up a little. I show all of it in the video, you could just pause and read it.
Superb. Thank you.
Great work you got here!
Thank you
Clay .... can you share purchase link on the solar panels and also i did not see the details of the compass used
@@godwinbadu-marfo7577 Adafruit LSM303AGR is the compass I used. It also includes an accelerometer which I’m not using for anything in this project.
The solar panel is just some cheap 12v panel off eBay, I think they’re all about the same. It was advertised as a 200W panel but in reality it’s closer to 50W
@@Clay-Builds thanks a lot for this detail
Esp 32 has more resources.
The power consumption inside watter will be much greater than on land. Maybe that overloaded the esd.
Have you checked the temperature inside the box after a long time running in the sun? this components may need cooling.
Im sure the heat will build up overtime. ~25 minutes of continuous use did not make the ESC’s noticeably warm to the touch, but this could be a problem if it runs for several hours. The next version will include current and voltage monitoring so I’ll be able to see exactly how much load is on the motors and how much solar power is being drawn. I don’t think these motors are anywhere near the 30A nominal limit, but I’d like to upgrade the ESC’s anyway.
That is awesome! Well done dude 👍
Thank you
Excellence unveiled!
Awesome design. I am working on an ASV for school. How did you waterproof the motors in your thrusters? Any tips would be appreciated.
Use sensorless brushless motors. For the most part, they’re inherently pretty waterproof and will have no electrical issues operating fully submerged. You’ll still need to worry about corrosion over time, but unlike typical brushed motors, there’s nothing inside them that could easily short out when wet. You’ll want to choose a motor with a relatively low KV value, because high torque is needed to drive a prop through water, and you don’t need very high RPM.
Many brushless motors, including the ones I used in this project, are specifically intended for underwater use. I didn’t do anything special to seal them or anything.
This is really cool
Thanks, version 2 coming soon with major upgrades
Looks nice - if you're using the Arduino servo output to control it via PWM, does that mean it never gets up to 100% per the way that signal is formatted? Would analog output be a better option?
Do you store GPS coordinates on the card before launching to set waypoints? Or does it store GPS coordinates for reading later, or both?
The servo library does use analog write to output a pwm signal to the throttle input on the esc. It’s easier to use the servo library rather than directly using analogwrite because the servo function takes a “position” value (0 corresponds to no throttle, 180 full throttle) and outputs the correct pulse width accordingly. Analogwrite takes a duty cycle value between 0-255, I’d have to add some extra calculations to convert from pulse width to duty cycle. Esc looks for a pulse width of 1000us (zero throttle) up to 2000us (full throttle). So I just use the servo library because it does that conversion for me.
It can output a full throttle signal, but I have it limited to 40% because the motors would stall at higher speeds when under load in the water.
Currently it only reads pre-determined coordinates from the card, no writing. But I want to implement real time data logging so the sd card would be like a “black box” with lots of different parameters for optimization and troubleshooting.
@@Clay-Builds Makes sense, and very cool, thanks!
Hi. This is a very interesting project. Good luck.
Thank you
Very cool. I wish I was knowledgeable about building stuff like that.
I got an idea to stuff light weight foam in the pipes that way if the pipe leaks and floods the whole pipe won't fill up with water.
BTW Clay I suggest watch a 13 part series called _What on earth happened_ by Ewaranon to learn that the earth is not a globe.
I got it in my about tab.
This is very cool!
Thank you
really kewl project man
nice job!
hm.. nice done! I thinnk it can be also useful as a bait boat in fishing.. also please think about swath variant for the seas..
Fantastic video and build!
Thanks a lot
I can`t wait to see the second part - when?
Could you send it from the East coast to England? That would be too cool, huh? Charge and pump for however long it would take and keep up with it's location via GPS. Have to be made for salt water resistance.
I'll let you know when it's on the way.
Very cool
Seems like you went thru alot of trouble waterproofing and mounting the box for the battery and electronics. Was there a particular reason for not storing the components in the water proof pontoons themselves?
I need easy access to the electronics so that I can troubleshoot, reprogram, remove SD card, charge battery, etc. The box with an openable lid is super convenient for this. The pontoon end caps are sealed shut with PVC cement. I didn’t want to put any holes in them and risk a leak since they’re critical for buoyancy.
You still using the 5V of one esc to power the 5V stuff? Get a $1 switching buck converter instead, since the esc use a linear regulator that gets hot and may die.
A nice arduino compatible processor upgrade would be the esp8266 / esp32. I did use esp-now to replace the RC radio. See the quadcopter on my channel.
Im putting in a custom PCB and upgrading to a teensy LC. It’ll still use the ESC to power the teensy Vin, but the rest of the components will be running 3.3v off a regulator.
@@Clay-Builds Dont. The 5V reg of the esc is a "M" type max 500ma. Depending on bat voltage this will get very hot. At least use 5V from both esc.
Brtter get a 3A 5V buck for that, 90ct on ali. No heat, no problems, solid 5V rail.
The 5V->3.3V Vin reg is good, keeps 5V noise from the 3.3V rail. Keep that.
cool stuff! I subbed :) hoping to see more of this in the future!
Thanks, I’ve got big plans for this boat and I always have several other projects in the works.
@@Clay-Builds show it all to us 😁 I'm very excited to see what else you're making
sweet build man, i like it. Could you keep the GPS antenna long by gluing it to the top of a pontoon?
Yeah I could bundle that wire up higher. I’m concerned if I cut out some length and solder it back together it could mess with the signal.
Nice stuff, looking forward to getting part2 with depth map. had thought of making similar project but never took time to do it. also I want to build too many things at same time which is very unproductive.
Any link to waterproof box?
why not make that motor mounts slightly rounded. also add telemetry 4G connection then live video? nice to have features. also why not use lipo battery then you more power less weight. love it so far!
They’ll be streamlined, I’m redoing them.
@@Clay-Builds looking forward to next video! and a night mission
Fascinating code as well, serial.print("Fuck") during debugging when things were not working xD P.s. nice project love the hardcoded lat/long makes sure it will never stops until it gets at destination, how ever the hot glue will make sure that it won't make it from Holland to England without sinking somewhere because it's not really a material to rely on
Haha good catch. left that in there from troubleshooting. The hot glue held up ok so far, but definitely not a seaworthy solution.
Dude this is awesome! I am working on a custom auto pilot too but it's based on the RP2040 and micro-python. Looking forward to the new radios. Any plans for telemetry?
I’d like to add telemetry but it isn’t doable with the current radio I’m using. Considering some upgrade options
Nice work😃 I like the idea a lot and it seems to work really nicely. I just fear it can't do really large distances as the motors are a bit weak? I fear they could collect some sea weeds or fishing nets etc. Hard to work around such issues..
Yeah the main reason I used prop shrouds is to prevent plants/trash from getting caught up. More powerful motors would need a much bigger solar panel, but I’d love to do a larger scale version eventually.
@@Clay-Builds I just installed 550W panels on my house walls.. they are 2,30m * 1,10m 😆 good scale I would say
i want to do the same thing except having it out on the ocean instead
That's even more impressive.
good job love your work
Hey man what if it flips upside down? Or we just sticking to pleasant lake trips?
It’s stable enough for any wave it would encounter on a lake like this. Not built for the ocean.
Well done 👍
Interesting build, Can I ask how you have made the solar panel ocean proof water resistant from waves smashing it and the epoxy type with regard to salt water?
The solar panel is waterproof, it’s intended to withstand rain. In practice it will stay totally dry unless there’s a large wave. Not sure how it would hold up to saltwater though. The epoxy I used is just gorilla glue brand 2 part 30 minute epoxy
Chad 🗿
Awesome work!!!
There is a potential in the code so that the activation point is saved and then always return to this point ??
It doesn’t currently do that, but it easily could with a just few more lines of code.
@@Clay-Buildsthanks for reply, I will copy the code from the video or there is somewhere to download?
I installed the other code with the sonar and takes 96% of 328ATmega ... there is any way to send you the progress of the progect ?
@@zuzuniagrikubanos6473 I have a 2nd video posted with some updates to this project, it has a GitHub link where you can download the code. That version has a lot more functionality than what’s shown in this video, but you could remove a lot of what you don’t need to save memory. The data-logging, sonar stuff, error detection, etc. isn’t totally necessary if you’re just wanting to build a simple boat that drives around.
The code shown in this video used almost 90% of the memory the arduino pro mini (atmega 328P) IIRC. A huge portion of that was due to the libraries I used for the SD card, magnetometer, and GPS module. Maybe you could modify or find some alternative libraries to save memory, but I ended up just upgrading to a better microcontroller.
@@Clay-Builds I already removed some
they are what I left on the code =
#include
#include
#include
#include
#include
#include
#include
#include
#include "pindefinitions.h"
#include "miscfuncs.h"
#include "constvars.h"
#include "manualmode.h"
#include "automode.h"
#include "cardlogger.h"
#include "navigation.h"
they are what I left on the code.I will start the connection tomorrow,,the correct pins is on GitHub link ??
@@zuzuniagrikubanos6473 the pins in that code are correct for the teensy micro controller I used in the 2nd video. You might need to make some slight changes depending on what microcontroller you’re using. There’s also an electrical schematic in the GitHub link that shows which pins connect to what. Make sure the throttle signals are connected to PWM capable pins, compass is wired to I2C pins, etc. you can mostly switch things around to whichever pins you want, but I’d make sure the timing sensitive components like the GPS serial connection, receiver communication pins that use the pulseIn() function, and the PWM outputs should each have their own timers.
How did you seal the cable exits of the motors? Which specific glue did you use?
The motors themselves are not sealed against water ingress. I used hot glue around the soldered joints on the back of the motor to prevent shorts between the individual phase wires. Also hot glue to fill in the drilled holes where the motors wires pass through into the electronics box. I’d use a proper potting compound if I were to redo it, but never caused a problem as it was in this video. Also notably, the pass through into the electronics box is not submerged, just keeping splashes out. Surely it would leak if below the waterline.
@@Clay-Builds Thanks for the fast and good reply! I'm doing my first underwater tests with electronics right now and I'm surprised how well hot glue seems to work all the time.
Impressionnant félicitation, si tu souhaite le tester tu peut essayer de lui faire traverser l'Atlantique, je le réceptionne en France et te le renvoi par DHL. En espérant qu'il ne finisse pas comme le Titanic...
A oui j'allais oublier, très bonne référence le livre au début, j'ai lu la version française, très philosophique.
Je remplirai les pontons de whisky américain et je l'enverrai en France si vous me le renvoyez plein de champagne français
@@Clay-Builds Je vous jure que j'en suis capable, du champagne grand cru avec une bouteille de Bordeaux en cadeau. Je n'ai jamais gouté le whisky américain ce serait une grande occasion de le déguster, et avec 0 frais de douane!
Hello, I love it! With my students we are making a similar model and I wanted to know if you could send us the file of the design of the propellers
I am working on a boat like this too! Is there any way you can give my the 3d model for the clamp that holds the pipes to the solar panel? Thank You in Advance and great build!
Great project. May I ask how many watt pick is the solar panel?Thanks
The panel was advertised as 200w but in reality produced about 40w at the most
Your autopilot relies on heading too much. Look at aviation autopilots for methodology. Even 40yo aviation A/Ps will used tracking for control, not heading.
Heading will always be wrong unless water current and winds are perfectly inline with path.
You need to computer the vector(desired track) from start point to end point and the adjust heading to stay on the desired track. You need to correct for wind and water flow. To avoid hitting objects, the desired track must be followed, otherwise you will end up hitting the banks or running aground.
As you follow the desired track, the heading will face in a direction that compensates for the wind and water flow. For example, a desired track of 180 and a wind of 270 might result in a heading between 180 and 270 depending on wind speed and water flow.
That’s true, in the presence of a strong lateral wind/current, it would make a wide arc towards the waypoint rather than a straight line. In the video I only used 4 waypoints, if I added a lot more between each of the major ones then it would greatly reduce this effect.
The navigation currently assumes the direction of travel is always exactly in the direction the compass is pointing. I could also get the direction of travel from the gps module and use that for steering. That way it could steer slightly into the current/wind and keep the direction of travel exactly towards the waypoint. Only issue is I don’t think the gps heading is as accurate as the compass, especially at low speed.
Clay, I’m trying to build something similar to use on our lake in northern WI. Would you be willing to share your code? Thanks, Mike L
Check my newest video about this robot, there’s a GitHub link in the description with code.
Where can ı get the 3d print and software files is this projec open sourced ?
I’ll be making a part 2 soon with lots of improvements and upgrades, I’ll have downloads for code cad and schematics along with it.
Thank you so much I'm waiting
hello cat, you are welcome
Dude do you have any wiring diagram between the solar, charge controller, and ESCs? I'm building a similar project and would like to see what you did. Haven't been able to find anything helpful online. Thanks.
I have a wiring diagram posted here: www.clay-p-mcpherson.com/projects/solardrone
But it doesn’t show the solar charge controller because it’s honestly pretty self explanatory. A typical Solar charge controller will have 3 pairs of +/- terminals labeled “battery”, “load”, and “panel”. The “battery” connections just goes straight to the battery terminals, + to red - to black. The “panel” connections go straight to the output wires for the solar panel. My panel has a barrel jack style output connector, I stripped back its wires (there’s only 2, a + and a -), and connected them straight into the “panel” connections on the controller, polarity matters. The input side of both my ESC’s (labeled 12vin on my diagram) go to the “load” connections on the controller; they’re wired in parallel. So both red wires to to the +, both black wires go to the -. That’s all that’s plugged into the charge controller. The ESC’s each have their own DC/DC converter with 5v output that I use to power the microcontroller and everything else.
Check my Chanel page, I recently posted another video about this boat which shows more footage of the wiring and upgrades. If you want to add an on/off switch, it’s best to put it in the positive side of the “load” wire, that way the battery can still charge when it’s off but the electronics won’t drain it.
@@Clay-Builds thanks, I was doing a parallel type systems, but I guess it makes way more sense and cheaper to just single up and put the ESCs in parallel.
@@Clay-Builds @Clay Builds if you have time for another question.
I have some 3s batteries with 2 30 amp ESCs to some little submersible brushless motors. And have already hooked them up to a remote for throttle and differential thrust for turning. I understand what you typed above in regards to the charge controller. But I guess I have a question in regards to solar charging 3s batteries. I have a bunch of 5v 0.5 amp panels and was going to do a combination of series/parallel into the charge controller. The 3s are also 11.1v so I'm not sure how it would react with any given charge controller. Should I just redo my design using 12v batteries, and 12+ volt panel?
@@esotericcommonsense6366 it depends on your charge controller. The controller will have a specific battery voltage it is intended to operate with, some have adjustable settings with more flexibility. it’s important to use a battery of the same nominal voltage as what the charge controller is set to. The controller reads the battery voltage and that’s how it decides when to start or stop charging. If the controller is expecting a nominally 12v lead acid battery, (which is actually around 13.1v fully charged) it might try to overcharge a 3s pack which is closer to 12.5v when full. A 3s pack can also operate at much lower voltage than a 12v sla, so the controller might think it’s dead when it really isn’t. I’d either use a charge controller intended for a 3s lithium battery or switch to a 12v nominal SLA if you’re using a 12v controller like mine. As for the panels, the voltage isn’t as picky. But depending on the controller, it probably needs to be higher than 5v. Look up what the operable panel voltage range is for your controller. If you’re using 5v cells, you could always wire them in series sets of 3 to effectively get a 15v panel.
Is that solar controller reliable? that cheap controller seems easily damage battery.
It wasn’t great. I swapped it for a GENASUN GV-5, a more reliable and efficient mppt controller. Don’t think I showed it on camera, but it was installed during the test run.
Can I pay you to help me do the same project? Just taking over the internet, being a guiding hand sorta thing. I'm looking to do the microtransat challenge
Not interested. But there’s enough information in these videos to totally recreate this project. Check the newest video too, you can download my latest code.
Enough info if you know what he’s talking about. Consulting is something you consider later in life. I’m trying to put together a solar raft but way bigger than this. I’m going to buy a sailboat soon and make it autonomous and able to cross the Atlantic way faster than average.
Cool project. Die you epoxy your pontons ? 😊
The end caps are held on with PVC cement. Epoxy for the 3D printed mounts.
@@Clay-Builds nice but i would prefer stainless Steel and a Rudder with Air props
Good for smuggling cocaine from Columbia to Australia
Need a camera on it
I plan to for next time
Do you have all this info in a repo to see details?
This is still an ongoing project. I have some improvements to make, and I’ll clean up the code a little. I don’t have a repo yet but I do plan to post the code/cad/ and other data somewhere soon. Any specific details you’re interested in?
@@Clay-Builds Yes please. I have used Mission planner. But this custom code is really . I am new to arduino and this help me a lot.
are you willing to share links to the 3D printed parts?
Hi
Its Awesome...man
How it cost..?
Could I buy a setup like that from you? I'm making a treasure hunting marine drone, albeit it's bigger with 600 watts of solar on it
Haha I’m not interested in selling anything. This project is totally open source however. See the newer videos for improved code and electronics information.
very nice and outstanding arduino...wow....but why your not to sharing the code.ino for us.......very sadness
I'd just use the PWM outputs on the Arduino to control the motors, no need for timing?
This does use the PWM outputs to control the motors.
Is there any obstacle avoidance technology on this pilot?
Nope, it’s flying blind.
@@Clay-BuildsThanks for reply. What do you recommend for avoidance from debris, vegetation, and other dynamic obstacles?
@@ehsankf vegetation is pretty unavoidable which is why I always use some type of prop shroud for small scale boats. Your best bet for above the surface obstacles is probably LiDAR but that gets expensive quick.
@@Clay-Builds Does Lidar capture the obstacles floating on the water surface or hidden under the water?
@@ehsankf surface obstacles. There’s some small scale hobby level LiDAR units out there. Even some robot vacuum cleaners use them for SLAM.
would have been much easier using ardurover
Well that would’ve been too easy and would’ve required a more expensive flight controller. Plus I plan for it to have a lot more functionality than just autonomous driving.