Project Logs

Collapse

• Designing the wireless charging and battery system

  Ben • 03/21/2025 at 12:14 • 0 comments

  In this post It’s time to focus on the wireless charging and battery system of the sea scooter.

  If you are interested the below video documents this stage of the build. So please do check it out!

  What is the power system

  I've broken the power system down into 2 main subsystems. 

  1 – The power storage subsystem. Comprising of a battery pack and battery management system.

  2 – The wireless charging subsystem.  All about efficiently getting enough power at the right voltage though the hull wirelessly to charge the batteries in a reasonable time scale.

  Designing the power management subsystem

  I first takeled the design of the power management system as this will have a knock on impact to the wireless charging subsystem design. The first decision was what voltage and maximum current I want the system to run at. If you’ve followed the previous instalments of the build you might know the answer.

  Quite early on in the build I decided to design the system to run at around 24 volts with a maximum 15 Amps current draw on the basis of the characteristics of the motor and it being still a safe voltage to work with.

  To calculate the Watt-hours capacity. I’m going to approximate the power consumption as the maximum expected power consumption of the Motor. Following the results of the magnetic gear propeller testing I arrived at a max current draw before the magnetic gear stalled at around 4 amps. So I’m going to work on the basis that the max average power consumption of the sea scooter will be 4 amps. So the total power consumption would be voltage times current so 24 times 4. Lets call it 100w.

  The next question is what is the minimum time I want the scooter to run for on a full charge?

  I think anything under 30 mins would be rather frustrating so i've picked that as the absolute minimum run time. Which if it’s running flat out at 100w would consume. 100 times 0.5 which is 50 Watt hours of capacity. 

  I've decided to design my own battery pack that will snugly fit into the hull. It will be a fun mini project on it’s own.

  Right so given that I’m making my own battery pack the first thing to decide was what battery chemistry do I want to use. As this would affect the way that I’d have to wire the battery up, volume and storage capacity, also the charging circuit.

  To cut a long story short I ended up deciding on LiFePO4. Principle on  the basis of high energy density. High discharge rate. one, if not the most stable of the mainstream Li-ion battery chemistries. Modular form factor. And..well.. I found a job lot available at a really good price on eBay. :-)

  This is what a single LiFePO4 cell looks like. 

  There are a few common form factors you can by the most common is called a 18650. Which means that it is has an 18mm diameter and 65 mm long. This one is actually a 26650 which, as you've probably guessed, it means it’s got  a larger diameter of 26mm and a length of 65mm. These batteries have a nominal voltage of 3.2 v. Which is importantly is different to the 3.7 volts per cell  your probably used to if own a LiPo battery for an RC car or plane. These one’s have a capacity of 2.5 Ah and a somewhat unnecessarily high discharge rate rating of 42 A. Apparently they were originally spec to go in an EV car battery. I think I paid around $3 per battery.

  Ok so I’ll need a battery that runs at around 24 volts.

  We can increase the voltage though the magic of simply connecting the batteries in series

  Ok so 8 cells is the magic number to get us over 24 volts. So this battery configuration would give us 26 volts times 2.5 amp hours equals 65 Wh of energy storage capacity. Which should give us a minimum of 40 minutes run time under normal conditions, so more than the 30 min minimum run time I was aiming for.

  So it looks like this battery pack that I will need to build will need to be wired...

  Read more »

• Designing an underwater wireless control system

  Ben • 03/06/2025 at 14:43 • 0 comments

  What is the control system

  So I've been continuing to focus on meeting the second design constraint . Working out how to wirelessly control the sea scooter…. underwater.

  if your interested in the detail of this phase of the project then you might want to check out the video below that goes though it all 

  I've broken this down into 3 subsystems.

  1 - The master on off switch or ‘key’. Which isolates the power source from the rest of the system. Reducing the passive current draw when the sea scooter is of as well as providing an emergency off button that is triggered in the event the sea scooter gets out of control, say you let go of it while underwater

  2 – the throttle and axillary function to make the sea scooter change spend and trigger any other auxiliary functions that we might want to do periodically such as synchronising the sea scooter with a base station or running a special test program.

  3 - the sea scooters ‘brain’ or central control system that takes in inputs and concerts these to output  signals to control other systems like the motor.

  Why can we just use Bluetooth or wifi..

  When someone say’s that they want to make something wireless the first thing that comes to mind is why not use wifi or Bluetooth to achieve this. I mean they are pretty main stream technologies and seem to be integrated into just about anything.

  the first thing to note about WiFi and Bluetooth is that these both work at 2.4GHz and higher frequency, which is the same as a microwave oven. This is because it’s a frequency that water is very good at absorbing and turning this into heat.

  So when you try to send a Wi-Fi or Bluetooth signal through water, which is often about ten thousand times weaker than your microwave oven the water molecules almost immediately absorb this radiation and converts it into heat. Not very useful.

  So if you work it though Wi-Fi and Bluetooth have an theoretical range of a few of millimetres though water. And as everyone I’m sure knows trying to use wifi or Bluetooth at the extreme of it’s range is often rather flaky and not something you can particularly count on.

  So I've gone for something a little different and once again turn to the power of magnets to save the day. As interestingly the magnetic permeability of water  is almost the exactly the same as air. So I've designed a control system that can sense magnetic fields above a certain threshold at a particular point in space so that I can transmit the intention to switch a button or increase a throttle.  

  Sensing magnetic fields is of course something that is not a new idea and there are quite a few ways to achieve it. Two popular, simple and well proven methods are firstly using a reed switch.  Which acts like a magnetic switch that closes when a magnetic field is near . The second method uses a hall sensor to sense the strength of the local magnetic field.

  Both sensors have their advantages.

  The reed switch is super simple but only giving binary feedback of whether a magnetic field above a certain threshold is nearby, but consequently fairly robust to changes in the local magnetic field due to other sources of magnetism such as the motor, magnetic gearbox even the earths magnetic field.

  While the hall sensor can sense the size of the magnetic field as well but this requires more complex electronics and often benefits from being connected to a microcontroller to read the output from the sensor and decide what to do with the signal.  Hall sensor are much more sensitive than reed switches, which is great, but this also makes them more susceptible to magnetic noise from other sources of magnetism that are in the local area.

  On this basis I’m going to for now at least make the master on off switch a reed switch. As I want it to activate independently of any other electronics and have a high threshold...

  Read more »

• Designing a Non-contact drive system

  Ben • 02/28/2025 at 15:43 • 0 comments

  In this update I've been looking at the non-contact drive system for the propeller. where I have combined a non-contact drive system that incorporates reduction gear box, known as a magnetic gear, to better match the motor with the propeller. 

  If you are interested complete details of this part of the build process is documented in the below video

  A magnetic gear is made up of three components a inner rotor, field concentrator and outer rotor. As the inner rotor rotates the outer magnetic rotor rotates in the opposite direction by an amount that is equal to the ratio of the outer to inner number of magnetic pole pairs. In this design the ratio is 4:1. 

  If you’re interested in more details of the construction of the magnetic gear, how it works and performs check out video below.

  Incorporating the magnetic gear into the hull.

  For this design to work I needed to be able to incorporate the field concentrator layer that contains 5 soft iron field concentrators into the wall of the hull that I need to be able to printed together with the hull in one go.

  The simplest setup I could come up with to test this integrated magnetic gear design was to create an access port in the hull and add a motor and motor mount. Powering the motor remotely using an external power supply.

  I also need to cut some 5mm thick strips of low carbon iron to use as the magnetic concentrators. To make this easy for myself I decided to build cross cutting jig for my DeWalt mini grinder. 

  This design also incorporates a fence to make cutting the same dimension easy. As you can see it does the job fairly well. The only slight addition worth incorporating in to add a bit of metal angle to the fence and square edge that are in contact with the part being cut to prevent the plastic from melting.

  If you want to see how the cross cutting jig is designed and put together check out the video I made on it below.

  I then glued the shroud on to the hull and then coated the surface in epoxy. This time I tried some readily available 5 minute epoxy that you can get from the local hardware store rather than flexible epoxy as I’m not expecting there to be much elastic deformation of the plastic this time around compared with the pressure testing video and it’s much, much quicker to cure.

  only thing left to do was to cut out the plastic window and  make some holes to pass though the power cable and attach it to the hull and we are good to start testing.

  Testing rig design

  I need some sort of test rig to measure how much thrust the test hull propeller combo actually produces. So I decided to take the biggest plastic box I could find add a gantry and sled to it and attach the test hull underneath it. That way I can constrain the motion to one dimension to measure static thrust. The only other piece I needed was to actually measure the thrust. I could buy a load cell, calibrate it, attach it to a micro controller and some sort of display. Or I could just repurpose a luggage scale that I have which has a resolution of 10 grams or 0.1 newtons which should do the job. Just needed to add a pully to ensure to transfer the force from horizontal to vertical so that I could read the display and keep it nice and dry.

  Propellers to test

  I decided to test 4 contrasting designs for propellers to see which is the best match. These designs where some of the top ranking prop designs for the rctestflight propeller competition that was run in the summer. Thanks to rctestflight and the participants for making the designs available.  If you’re interested in the detailed test of how these and a whole bunch of other propellers performed then you should definitely take a look at rctestflight’s channel and the final competition video.

  Lets print out the props. 

  Read more »

• Waterproofing the Hull

  Ben • 02/28/2025 at 11:30 • 0 comments

  Ok so the first challenge was to see if I can actually make a 3D print waterproof.  if not this project is sunk in the water. 

  For this to work I'm not just talking about whether the print can hold water. But need something that can stand up to some serious water pressure without springing a leak....

  if your interested in the phase of the build then be sure to check out the video below as it has all the details of the build so far.

  I needed a pressure test rig for that.

  Ok so to test whether the sea scooter can to handle this kind of pressure. i first had to design a pressure testing rig!

  I went for a design that consists of a polycarbonate cylinder with 3D printed endcaps  also out of polycarbonate. well actually polycarbonate carbon fiber in the end... These form a tight seal with the cylinder using a large silicone O-ring. With either end secured in place with 6 pieces of M10 threaded rod. It was then filled with water and a ¼ inch  air hose fitting taped into the top so that I could connect it to an air compressor to pressurize the test rig. 

  The samples where attached to this load spring and submerged in the tank to measure how buoyant they where,  and how this changes over time. The idea being that either water iwould ingress and compress the air in the cavity – reducing the buoyancy, or water would ingress and air would leave as a stream of bubbles – reducing the buoyancy. As  a back up I also weighted the samples before and after to measure if they took on water.

  I tested a scaled down, simplified version of a sea scooter hull to make the testing as representative as possible. (white object above) 

  If you do a quick search you’d find quite a bit of material on ways to water proof a 3d print. It’s not all entirely consistent. but after a bit of thought I think I found a good set of parameters to test. (details in the above video).

  I tried ASA as the main contender and PLA as a reference. Trying these at a few different pressures up to a few Atmospheres (Atm) water pressure and also few post processing techniques like acetone smoothing and epoxy coating. So I needed quite a few test hulls......!

  There where some surprising results. 

  But post processing was definitely the way to go.  The Epoxy coating in particular was able to stand up to more pressure than you might expect. 

View all 4 project logs