Today, wireless-enabled microcontrollers are everywhere and can be had for just a few bucks. You can use them to build all kinds of connected projects more cheaply than ever before. [ROBO HUB] demonstrates this well with an incredibly simple WiFi-controlled RC car build.
The build is based around an NodeMCU ESP8266 microcontroller, paired with an L293D motor driver. This lets the microcontroller drive brushed DC motors for differential drive. Power is courtesy of three 18650 lithium-ion batteries. These parts are assembled into a 3D-printed car of sorts with four wheels. The drivetrain is rather odd, with gear motors installed on the two front wheels, and simple brushed DC motors installed on the two rear wheels. The motors on each side are paired together so the vehicle has tank-style steering.
Meanwhile, the ESP8266 is programmed so it can be controlled via a smartphone app. The touchscreen controls are not as elegant as toy RC cars of years past, but it’s pretty good for a cheap DIY build.
It’s a fairly simple project and one that any high-school student could follow along to learn something. Projects like these can be a great way to learn about everything from mechanics to electronics and even basic programming. It may not be complicated, but that makes it a great learning tool. We see a ton of projects like this on the regular, and every time they’re built, somebody is picking up some new skills.
We’ve been talking about WiFi-controlled RC cars for a long time. Way back when it was nowhere near this easy. Video after the break.
URL: https://www.instructables.com/Make-3d-Printed-Smartphone-Controlled-Wi-Fi-Car/
The glueing makes me cringe; for me the brackets should be bolted to you know they are mounted square on the platform.
I had an idea for a low cost skid steer which doesn’t use hydraulic motors. the idea is to put a pair of chain driven counter rotating drive wheels on the end of a bogey axle.
Lean the axle forward, and the vehicle moves forward, lean back and it moves back. hold the axle vertical, and it functions as a brake when a curved foot raises both wheels from the surface.
I lack the discipline to build it but if anyone wants to make a minature model go ahead.
I also am not sure why a simple chain or belt would not be sufficient for a skid steering… And in fact I am curious why skid steers are not much more popular in these type of constructions, as they have some advantages over more popular constructions with two driven wheels and a caster wheel.
It seems to me that latency would be a huge issue using WiFi for control. Unless the car moves very slowly in an obstacle free area.
FIRST Tech Challenge robots use WiFi for their control link. Unless there is radio interference the controls are very responsive. Lots of competition videos on YT if you’re curious.
FRC does too, and those robots are 150lbs and can move a few meters a second!
Each is limited to a 4 megabit connection to the driver station to ensure fairness and low latency links to all teams, and they still manage to stream a couple of cameras.
Camera latency is WAY higher than control latency. Our team’s cameras do 30fps but with 100-200ms latency. The controls meanwhile are about 16ms.
Where do you connect the front motors ?
The front motors are not connected. They are just an easy & cheap way to mount the free running front wheels.
Wich software do you use on the node MCU and on the phone ?
The screenshot shows the same controls as the software used here:
https://www.youtube.com/watch?app=desktop&v=mJEftoeSBOU
The Pi Pico W looks like a good alternative to the ESP microcontrollers for those who are trying to get away from Chinese supply chains and binary blobs.
…particularly since in most toolchains, multiple simultaneous PWM is trivial. Drive 360° modified servos with wheels mounted, or DC RC car motors via a dual reversible ESC (electronic speed controller) and you’re in business.
It’s the new belle of the ball for these type of robots. There are some pretty functional and cheap kits on amazon and the like but man oh man the kits are sooooo bad. They all use one of three ino code bases that either lack the proper libraries at this point or have unreadable (in the US) chars in code, or just have incorrect code. I helped a friend that teaches high school go through some of these for his class and it became an exercise in hilarity. There was always something catastrophically wrong with the kits to the point we quit after 8 attempts and just cooked up our own. It should be that damned hard and allllllllllllllllllllll of those youtube and tiktok videos steadily parrot bad instructions, selling kits, with bot comments on how well it worked. Just sad. One of those things I will go back and fix when I retire.
Sorry, none of this rant was directed at you or OP. I thought it was an honest walkthrough of the process. The Pi W is pretty darn sweet. Lotta bang for buck, easy to program and tons of libraries etc already and integrated wifi puts it in competition with old esp32 kinda projects. Lots of those projects are great but can be so much greater with a little oomph provided by the pi. I haven’t gotten super deep into code trickery yet well because I haven’t had to lol. It has handled those level of projects. If I need something more beefy I use a beefier platform. They are pretty cheap though so it is easy to keep a couple in the drawer for whenever inspiration rears its head :)
It is good to revisit these type of projects from time to time and see what newer hardware has to offer and also (for some of us) it gives us time to see the state of the interwebs when trying to find simple and accurate information. Just so much noise in the way of 123 ya know? Maybe I am just getting older and more grizzly lol.
Still neat, but so 2016.