Project Logs

Collapse

• This is actually a dead project.

  R Lang • 08/02/2026 at 08:27 • 0 comments

  In the following order:

  Time - the enemy of us all, essentially. I instead focused my attention on my EE knowledge and in particular focusing on my CAD skills. I feel the basis for Tortuga is very sound and very do-able, I do plan to come back to something of a similar theme in the future but on its own dedicated PCB. I will probably stick to the ESP32 platform though.

  Workspace - I don't have a dedicated workspace and share the space with my family, this isn't the most congruent for a good work flow. It's something I'd like to re-visit at some point, co-working maker spaces in my area are few and far between, independent renting is not an option and nor is taking over a communal family area for my exclusive use. 

  Prototype - The prototype ended up being at best "sufficient", I don't like sufficient. 

  Ultimately, what I guess I'm saying in a lot of words is, I'll come back to this, but not particularly this. 

• This isn't a dead project

  R Lang • 03/14/2026 at 19:28 • 0 comments

  It's just on a back burner. Life is tumultuous at the best of times. I do plan on picking this back up at some point. 

• My initial ideas

  R Lang • 09/29/2025 at 08:54 • 0 comments

  I wanted to share this little snapshot represents my initial ideas on where I wanted to take the project, design wise. For now, I have abandoned this design and will use it when I scale up the robot - for now it's an unwarranted level of complexity for what is ultimately a small robot. In future? Who knows!

  In a future update I will share the final design along with the STLs and code examples, currently I am debating a polar vs cartesian co-ordinates system, polar would mean starting it from a nominal, known point - really not that hard to design, cartesian, being 3d would need an IMU to measure its position - again, not really that big of an integration issue, but a layer of unwarranted complexity, affecting the overall size and shape of Tortuga 1.0.

• Where from here?

  R Lang • 09/28/2025 at 13:54 • 0 comments

  So, in as much as progress, it's been a trickle rather than a flow, but a slow, and progressing trickle none the less. It is important to note, that I am entirely a hobbyist that incidentally is very interested in EE, design and gadgetry in general, hence the selection of ready to use boards, as opposed to some of the amazing but custom designs on here.

  In as much as a final shape, I have settled in some sketches as to what the final design will be like. For Tortuga 1.0, I have decided I'm going to keep it as simple as possible, the final product will be cuboid in shape, the line dragging tool for the sand will be a simple linear actuator contained within the chassis. As for final hardware decisions, the following is a non-comprehensive list of the final hardware choices:

  BOM:

  1x ESP32-C6 Microcontroller
  
  1x TB6612FNG Motor controller
  
  1x DFR1026 Battery management system
  
  1x USB C pigtail
  
  1x 7.4V LiPO battery
  
  4x N20 style motors
  
  4x 685ZZ radial ball bearings

  For now, the prototype will be USB tethered for control but ultimately, the final goal will be to either have external control from some kind of handheld controller, such as an XBOX controller or a mission planner, where the pattern is drawn in an app, with Tortuga translating these into cartesian co-ordinates. 

  What next? Well, starting work on the drive system more than likely, although I had earlier mentioned a reluctance to examine the use of differential steering, this is most likely the optimal solution for the steering issue. After that some kind of low pressure, slick tyres to spread out the weight will need to be designed. Think along the lines of a Beach wheelchair. 

  The physics behind a beach wheelchair is actually quite interesting, the force exerted by a vehicle is referred to as ground pressure, the measurable metric of ground pressure is potential mobility. Essentially, the greater the area and the lower the pressure, the easier it is for the vehicle to "float" across the surface. This is why hovercraft simply glide over most surfaces and unfortunately, is also the reason why we're unlikely to ever see a truly useful walking mecha, much to the chagrin of us Gundam enthusiasts. 

  Ground pressure is relatively simple to work as far as equations go, simply put it is: P=F/A where F is the weight in newtons (N) and A is area in m². For sand, low pressures around the 1 to 2 pascals (12-14PSI) area are considered optimal. All this is aside as the pressures exerted by Tortuga are so small, it's almost not worth going to deep into it. None the less, this is an interesting delve in to the physics behind it. If you think this is bad, you should see me proselytise about the virtues of the BS 1363 plug AKA the British 3 pin plug.

• Test bed V1.

  R Lang • 08/30/2025 at 17:07 • 0 comments

  So here is an initial assembly idea of what the test bed will look like. The front bay will be for seating a battery, the finished test shell would likely have venting for batteries, maybe some kind of hexagonal pattern at the bottom. The idea of keeping it as sealed as possible is to prevent sand ingress into the motors and electronics. 

  The rear component is a hinge design to actuate the tail, Tortuga when wanting to carve into the sand will actuate its tail downwards at a set angle and drag it behind it, producing the desired line. Not shown: 

  Servo to actuate the tail

  Nose wheel steering servo

  Batteries

  ESP32 or Raspberry Pi 0w

• First print!

  R Lang • 08/28/2025 at 14:47 • 0 comments

  This is the result of the first chassis print! Having discovered an abundance of hoarded Raspberry Pi 0Ws knocking about, I am probably going to go ahead with using that platform for prototyping, I have also decided that I will go with active nose wheel steering rather than differential steering, the differential steering presented a challenge for me to implement, given that I am new to coding. I however have not given up on it and will come back to it in the future.

• A new 3d printer!

  R Lang • 08/26/2025 at 08:24 • 0 comments

  I am due to take possession of a new Flashforge Adventurer 5m, thanks to my amazing family. It represents a good cross road between feature rich, ease of use whilst remaining accessible and elementary to use. Tortuga's shell and chassis further design can now move forward! 

• It's been a while!

  R Lang • 08/09/2025 at 13:44 • 0 comments

  With university now finished, I can now spare some time to actually get this done. There's no real new changes to the concept as yet, work will restart on the CAD drawings pretty soon, I have looked at the possibility of using a company such as EasyEDA or PCBWay to produce the models, price not withstanding of course, this is not an optimal solution to my problem. To that end, it appears as if I'm in the market for a new 3d printer!

• Future ideas on mission planning

  R Lang • 04/11/2025 at 12:54 • 0 comments

  So, as it stands my research has kind of hit a bottleneck in that, for now the hardware side of thing has hit the snag of not having a readily accessible 3d printer on which to print and iterate on designs rapidly, so rather than losing momentum I've decided to take a skim at what is available for mission planning. 

  Anyone who grew up in the 80s and 90s in the UK will remember the venerable RM Nimbus running the RM version of the classic programming language Logo. (Perhaps this is where the shape of the turtle subconsciously appeared from when thinking of a friendly shaped line drawing animal?) In the programming language Logo, you were given a turtle shaped animal and a very basic set of commands which commanded the turtle to move in any direction you chose. 

  Shoot forward to today, where the likes of the Nimbus are now museum pieces and we have hardware that are magnitudes more powerful for a fraction of the cost and size, hence why we are able to do projects like this. 

  What I am after is a something that offers similar abilities to logo in that, it needs the minimal of knowledge to accomplish a lot. Examples are out there such as Python Sandbox's turtle mode where a set of very basic commands are able to make basic shapes. But beyond this, how do we go about translating pictures into cartesian coordinates that a robot knows to draw? My research has led me to this pipeline which seems to offer the highest fun to labour ratio for now, given my limited set of skills. 

  Additionally, it then begins the conversation of manual control and how this will be accomplished, but one thing at a time I guess.

• Test Bed

  R Lang • 04/09/2025 at 13:07 • 0 comments

  Although this doesn't represent the finalised design, this is generally the direction I've chosen to go. The idea of this will be to test the ability of Tortuga to drive on sand and to test out a number of things:

  A. Ingress of sand into the electronics

  B. If low pressure, slick tyres will adhere enough on sand 

  C. How accurately can Tortuga steer in such a small form factor.

  My initial thoughts are the front wheel will need more room to manoeuvre - maybe a castor based wheel, if not, depending on the level of difficulty in implementing, regenerative steering/differential steering - the same thing that allows tanks to perform "neutral turns" (tank donuts).

View all 11 project logs