Tuesday, February 24, 2015

Defeat Terrorist Drones Without Firing a Shot

Lots of news reports today of drones flying over the Eiffel Tower, French Nuclear Plants, and the US Embassy in Paris.  Of course this is irresponsible and dangerous behavior, and the idea of terrorists using drones is mentioned.  How could you stop a terrorist using a drone to carry a bomb or using a drone for some other evil purpose?

Here in the good old USA, we have the state of Oklahoma considering a law to allow you to shoot down drones over your land.  Shooting drones, particularly by homeowners, is not such a great idea as there is a lot of potential for innocent people to get hurt.  It may make sense for the military, who have over 100 years of experience with shooting down flying objects, but not for my neighbor Fred, who has a shotgun and last fired it when he went hunting with his brother in law 10 years ago.

There are certainly less dangerous but effective ways to deal with drones.  Commercially built or homebuilt drones rely on a few things to fly to a target and execute actions.  The loss of one or a few of these capabilities will cause a drone to crash.  Those of us who build and fly drones for fun know this all too well.

  1. Sensor input - barometric pressure, accelerometers, compass, airspeed indicators..
  2. GPS or GLONASS input.
  3. Radio commands.  Radio telemetry back to the operator in some cases.
  4. Onboard computer program.
  5. Correct operation of motors, speed controls, aerodynamics, and control surfaces.
Almost all of these capabilities are needed for a drone to execute a mission, so how do you deal with an evil drone?
  1. Certainly most sensors can be fooled quite easily without physically touching the drone.  I won't detail how to do this but most people with some engineering training could figure it out.
  2. GPS and GLONASS can be remotely jammed.
  3. Radios can certainly be  jammed.
  4. Onboard computers might be hard to disable remotely but there are some techniques that will work with the unhardened commercial flight control systems.
  5. Motors, control surfaces and the like are susceptible to damage with very little contact.  A piece of string or a plastic bag in the propeller will stop most drones.  There are lots of ways to stop a drone with similar low tech techniques.
So there is likely a business here for someone to make and sell drone protection systems to high profile targets like stadiums, iconic buildings and monuments, important infrastructure, and government installations.  I am sure we will hear about some startups or military contractors working in this area shortly.

Photo Credit (edited): Flickr

Saturday, February 21, 2015

Test Flight Successful

Finished the basic build of the quad and was able to get a quick successful test flight in the basement after about 5 hours of messing around with the quad, my radio, and Mission Planner.

Finally worked out all the little issues and had a successful test flight:

  1. The Deadcat needed to be balanced as it was naturally nose heavy.  Moving the battery allowed me to balance the quad.
  2. I played around with ESC calibration and radio settings for hours because the props kept spinning when I armed the quad.  My old APM equipped quad did not do this.  Well it turns out that the default for Mission Planner with Pixhawk is to make the blades spin slowly when armed.  I did manage to program the ESCs to play tunes when they power on, but that is not much use.
  3. Turning on the throttle failsafe is a waste of time as it is triggered as soon as you arm the quad, causing the arm function to fail.
  4. Use loctite on all bolts, the one place that where it was skipped (landing skids), a bolt came loose.
  5. Need to make up a cheat sheet with the meaning of all Pixhawk and ESC sounds and lights.  It is hard to understand failures otherwise.
  6. Had to reverse all the RC channels except throttle.
  7. It seems smoother than the last quad as the props are balanced but it is more twitchy on the controls, so will have to do some more tuning.

Thursday, February 19, 2015

Lots of Little Things to Complete the Quadcopter

I am making progress but am now at the stage where there are lots of little things to do to make the quad work.

  1. Install the flight controller with special vibration reducing foam (see photo 1 below).
  2. Install the power monitor and battery velcro mounts (see photo 2)
  3. Calibrate the ESC's.  This is always a pain as you have this "one man band" performance where you are plugging the main battery in and out, watching lights on the controller, listening for musical tones and beeps, changing throttle settings, listening for more musical tones.  I usually end up doing it 5 or 6 times before it works.
  4. Calibrate the flight controller - Mission Planner complained a lot until I saw in the fine print that I had to recalibrate the compass due to a special new firmware load.  Lots of waving the quad around when you do this calibration.
  5. Installed camera mount and new camera - looks really good except the new camera wants a special power connector that I do not have.  More waiting for parts from Amazon.com.
  6. Items that remain include installing the final Spektrum receiver and PPM converter, wiring the camera, putting some mechanical pieces on the dead cat frame, installing the new professional looking cover, and then some flight testing in the basement.
Photo 1 Battery Mount and power monitor
Photo 2 Note Green Kyosho vibration absorbing foam
Photo 3 New camera that includes a 5.8 GHz video transmitter
Photo 4 Overall view of Quadcopter

Wednesday, February 18, 2015

Design Choices

Spent the last few days working through some design choices and doing some preparation work while waiting for parts.

1. I balanced all the props using the technique in this great video:

2. I soldered the battery connections on the bottom of the Deadcat PCB.  I did this by removing the paint or mask over the power and ground planes and soldering to the exposed copper.

3. Setting up the radio exposed some problems.  The Spektrum DX6i transmitter (left) that I planned to use is not ideal for quadcopters as it has only two position switches for the extra channels, while my old DX4e transmitter (right) has a handy 3 position switch which can be used to set flight modes (Stabilize, loiter, auto).  Easy to fix, use the DX4e right?  Not so fast, I got some help from a Spektrum radio developer on the RCUniverse website and I found out that the DX4e has half the power output of the DX6i!  So I may use the DX6i and fiddle with the mixing function to get more flight modes.

4. Need to reduce vibrations for the flight control system so I bought some Kyosho gel tape from Amazon.com.

I am still waiting for some critical parts - the PPM mixer for the AR610 receiver (on the boat from China), the new video camera (same source, different boat), FPV monitor (ditto).  Stay tuned for further updates.

Saturday, February 14, 2015


The quadcopter is coming together now that I have the proper parts for the Dead Cat configuration.  The arms are all attached although it was hard to get the screws to align, the new GPS is mounted on a foldable mast, and the ESC wiring is done (not shown).  I forgot to mention previously that I added carbon fiber motor mounts (about $8) as the fiberglass original parts are not very sturdy and break in a crash.

I am now waiting for a PPM adapter to arrive.  The Pixhawk only supports PPM input from the receiver (single wire interface for all channels) and the Spektrum receiver does not have this feature, so a conversion board is needed (about $8).  I am also waiting for a new camera and gimbal.

Dead Cat Config
Note the GPS in the center left of the photo

Carbon Fiber motor mount

Thursday, February 12, 2015

New Quadcopter Build: SK450 Dead Cat

Now that we are back from Florida and parts have arrived, it is time to start the new build.  I salvaged whatever I could from the Flying Flowerpot (may it rest in peace).  Two motors, the four speed controls, battery, flight control, receiver, and camera were all salvageable.  However, I decided to replace some of these parts and to change the configuration of the quadcopter to "Dead Cat" - a non-square quad setup with more room for cameras.  See the picture below for the basic setup, not fully assembled.

The other changes that I plan to make are:

  • New flight control system (Pixhawk) with more accurate GPS (LEA-6H).  I bought the genuine 3DR versions instead of the Chinese clones to ensure quality.  My old APM 2.6 system was a Chinese clone and had limitations and unknown quality.
  • New receiver (Spektrum AR610).  This receiver is highly regarded compared to the Orange R615X, which has rumors of range issues.  New Spektrum DX6i transmitter.
  • Use of a PCB as a frame to distribute battery power, bought from Hobbyking.
  • Dead Cat configuration using a conversion kit from Hobbyking.  I made a mistake and assumed that all I needed was the previously mentioned SK450 Dead Cat PCB, but it turns out that you need this plus the Dead Cat conversion kit.
  • Full monitoring of power by the control system.  This was included with the Pixhawk.
  • A proper cover instead of a flower pot, sourced from Quadrysteria.com.
  • Receiver power from the Pixhawk or separate UBEC, not from the ESC's.  This should be higher quality power with less interference.
  • Use of loctite on all screws to reduce the chance of loosening with vibration.
  • Carbon fiber props.
The new control and radio system are shown below.