Hey everyone!
I learned loads of stuff while building the ‘HoverBot’.
http://www.rcindia.org/self-designed-diy-and-college-projects/hoverbot-xtreme/

So, after realizing that there’s lots of room for improvements, I started working on the HoverBot v2.

Here’s the new features:

1)   Semi-autonomous (using Arduino Due). Sets ESC throttle end points at each startup. Detects obstacles and stops the HoverBot.
2)   Differential steering for better control, agility and torque-balance.
3)   Full brushless setup, improved skirt and compact design.

We might add gyro stabilization (as I have a quad board now!) and onboard cam. Then there's a 2-DOF arm built waiting to be attached.
Now those who have seen the vids of HoverBot can just imagine the speed this one can achieve! Will post vids soon!

Pics:
1) Front view
2) Battery bay

More pics:
3) Lift motor
4) Thrust motors and Arduino

beautiful    Swapnil looking awsome, i know will run beautiful too.we will Wait for the Video eagerly. 

Awesome build!!!!!  The whole concept of design is futuristic. Not only design but you have also added feature of automated system. This one is completely new thing i have ever seen.

Thanks Swizera and Dharmik!


I've also written a program for auto-lift and auto-navigation(using sonar) . And it works great!
But the group of college friends (juniors) I built this for wanted it to have radio control as well. So now it's semi-autonomous.

One of my friend is working on machine-vision. When it's done I might be able to make the HoverBot autonomously chase humans. (or cats!)   

Well done Swapnil. Please post a video!!

Thanks Rohit. I have some 'testing' videos. Will post in a bit...

Here's a water test vid:

Swapnil..a million salute for your great work and dedication.. ..

Thanks Roopesh sir, I'm honoured!

its not a simple apreciation Swapnil..!! but you deserve it.. because making a full hovering bed is a tougher job.. but you made it here.. simply.. only a few can achieve this.. even you can look around the world, then can only find a very few machines as sucessful.. so, its moments of happy.. and proud.. thanks for sharing your works to us.. keep it up and go for even bigger machine.. have a nice future.. 

The 2 DOF arm turned out to be too heavy so we just added the gripper.

And the auto stop using IR sensors. The range is not much. Will soon change it to ultrasonic.

The differential thrust was achieved by using elevon mixing (with zero aileron diffetential) plus one programmable mixing. The torque is balanced by proper trimming.

And here's what differential steering can do!

This thing looks so cute. Would you mind telling us material of skirt?

The material is an old car cover dipped in a secret solution ( ) for making it totally airtight. It also has duct tape at bottom for protection on tar road.

Does anyone have experience with laser range-finders?

I was wondering about the color scheme. Would black+silver look better than black+ yellow or black+red? Or should I let it be as it is (i.e black+white)?

Here's a test bot I made to configure and check the diff mixing before applying it on the HoverBot v2. It turned out to be really fun! 

Setup:
500 RPM side-shaft brushed motors x 2
HK 30A brushed ESC                     X 2
LiPo 2200mAh 20-30C

                   

 

OMG...
u did so well sir...
Damn good work. When you say it took TWO days to make it,what exactly do you mean?
it took only TWO days to develop the design,research,assemble things together,develop the code etc etc?
please enlighten me.:-)

Nah! Two days to make the hull (that is, base + skirt) and fixing the thrust motors.
Total assembly time (including programming the arduino and setting channel mixes) was 4-5 days.

its great......but from where you got...HK 30A brushed ESC..?
   

Brushed? The HoverBot uses brushless ESCs. If you are referring to the 'FunBot' ESCs, I got them from Vivek bhaiyya (rcelectro.com)

How did I miss this awesome creation?!

Btw, the 'secret' solution maybe Silicon Spray???

Your appreciation is like an oscar award for me VC sir! 
The solution is simply permanent marker ink mixed with moderately diluted glue. I also used a layer of light and supple packing tape for use on tar roads.

can i use a brushed motor for lifting a 400 grams,
18"x8" ,thermocol base and parachute fabric skirt.?
how many rpms and voltage would it require with a prop size of 6x4(or 8x3.5 whichever is best suited).
and what will it cost?

You can certainly do that. My 1st hovercraft used a 12V 12000 RPM brushed motor and could easily lift around 1 Kg with a 3-blade 4x3 prop.

The motor+prop combo cost me Rs. 390. I moved on to brushless because I didn't have a brushed ESC then and wanted a linear lift.

I hav a small rc heli (cheap chinese toy.),both of its motors are working(when connected directly to a regulated voltage power supply).
It has counter rotating blades. can i use its motor(s) with a new prop for lifting the hovercraft?
Its worthy enough to mention  here that the motors are very small (1cm dia,2cm height).
Also i would like to know whether it can be used for propulsion system(if not for lifting)?

O.K, let me tell you that I tried a range of DC motors for the thrust system to keep weight low.
(3V, 6V 8000 RPM, 9V 15000 RPM, 12V 12000 RPM, 12V 16000 RPM etc). Only the 12V 16000 RPM motor gave good enough thrust without heating up.

Because, even though the hovercraft is hovering (frictionless), the thrust motors have to overcome inertia.

I have micro helis too. Their motors certainly can't be used for a hovercraft of size 18" x 8".

can u please tell me the dimensions and skirt specs of this build?it would be great if u can give some pics/videos of it

The skirt design is really simple. The dimensions are (side length in inches) x (6 to 8 inches). Mine is 12x8
The skirt corner design is shown in the pic below.

Hi Guys
I have made a hovercraft. I want to know how to connect tow ESC to the receiver? I will be connecting the thrust ESC to pit channel. Pls tell me how can I connect it.

very nice build swapnil.
i am too making a hovercraft i have question.
how did u achieve torque balancing.did u use counter rotating props..
waiting for your reply..

No, I used a small 3-blade prop for lift with a differential thrust setup.

thanks

can u pls explain what is this elevon mixing ?  and the other mixing that u tried on your funbot ? 

I know this is a pretty old thread but I wanted to ask something as I am working on something similar.
How did you interface the RC system with the arduino due? The way I see it is that the receiver outputs PPM and the arduino outputs a modified PPM to the escs right? How does one accomplish that reliably?

Hmmm...I don't really understand your query. I analyzed the signals from both the systems on my CRO a long time ago and they were both definitely PWM signals (which is what a servo/ ESC expects).

You don't need to go after the PPM signal unless you need to use more than 2 channels with the arduino.

Also, I did not use an arduino Due (which uses an ARM Cortex M3). I used the Duemilanove and might have shortened it to 'due' while typing.

Arduino generated 'servo-signals' are very reliable and work flawlessly with RC servos and ESCs. I've worked with the servo library a lot since this project. It's very convenient to use and doesn't mess with the rest of your code or hardware. 

The receiver outputs PPM definitely. PPM may look like PWM on the scope as the signals are similar. Correct me if I am wrong
Did you use interrupts?

I'm afraid you have a misconception.

The output on each channel is most certainly PWM.
PPM utilizes time division multiplexing to pack multiple (about 9 - 10) PWM signals in a single time-frame. There is indeed a PPM signal at the receiver output before it is de-multiplexed for separating channels. And one needs to tap that in order to decode more than 2 channels with the arduino.

If you still have doubts in this regard you should seriously consider brushing-up your theoretical basics or better yet try working with and analyzing the actual hardware and related signals.

I did use interrupts. But it was completely handled by the servo library.   

Now I get it. I actually do not own a scope and got my theories a bit messed up. My Tx and Rx communicate through PPM. And the Rx outputs PWM.

Another query though. Did you make use of the pulse-in function?

Yes.  It's good enough for upto 2 channels.

What's differential thrust mechanism .?

Differential Thrust, if I am not mistaken, is all about controlling a craft with varying propeller thrusts on the port / starboard (left/right) sides. It negates the use of a rudder.