Thursday, November 20, 2014

Toaster Oven Modification for PCB Reflow

This is a video on a toaster oven modification for PCB reflow purposes. I got a toaster oven at Walmart for $20 and added a reflow controller called the Controleo from For folks who don't know what reflowing is, you can read about it here. I wanted to  make some PCBs at home using SMD components, and decided to give this project a try.

The Controleo setup was pretty painless and inexpensive. If you do decide to do this modification, please do so at your own risk. AC voltages can bite you if handled improperly. Happy reflowing!!

Wednesday, September 4, 2013

Dynam Hawk Sky Review

I bought a Dynam Hawk Sky a couple of weeks ago from and here
is a little review.
The package was delivered without any issue. The foam parts were packaged well. Plugged in the LiPo to charge on the provided balance charger. The charger gave a wierd high-pitched sound when charging, I guess that was normal as the 2 LEDs were green. It took about 45 minutes to charge. The instruction manual is poorly written, so I would suggest looking for some help online. If you have previous RC experience, most of the construction is common sense. The horizontal stabilizer and the rudder were attached using 20 minute epoxy ensuring that the parts were square. The rudder and elevator clevises are a couple of brass attachements with a screw to hold the push-rod in place. It was quite flimsy in my opinion, will get to that in a minute. I let the epoxy harden overnight. The aeileron servos/push-rods came installed. The wing installs by sliding each side into the fuselage after connecting the servo leads. There is a plastic tube that goes between the wings to add some strength. The wings are held in place by the tightness between the wing root and fuselage. I found that a little unsettling, hence added a couple of pieces of electrical tape for good measure. I would have hated for the wings to come loose in flight :)

I wasn't too sure about the 2.4GHz radio system that came with the plane, so I replaced the transmitter/receiver with a Tactic system from another plane that I had. The stock servos worked with the Tactic system, so I left them in place.

The motor connections required a bit of maneuvering as the motor wires had to be pulled through the foam area on the mount into a vent hole just under the motor mount (see pic above). This required a bit of patience as the hole was very small.The first flight was uneventful. The brushless motor has plenty of power for a good climb. The plane is very dosile and it is difficult to do rolls with the huge wingtip dihedral.

My first take-off was on asphalt using the landing gear. Later did hand-launches. It glides really well as I was flying at 1/2 power with a flight time of about 20 minutes. The stalls are predictable, with a bit of spiral to the left. Since the plane glides very well without a significant drop in airspeed, the landing descent was started early on idle. I had to use a bit of down elevator to keep the decent as it would have otherwise over-shot my landing mark. The rudder works well at low airspeeds as well. The first landing was a bit hard and the elevator clevise came loose. I ended up twisting the pushrod in the hole of the control horn to keep it in place for the second flight. I would recommend that this be done for all surfaces. I tested a payload capability with a 160g battery pack, and the plane handled it without issue. I will eventually be using the plane for my Ardupilot Mega autopilot. There is a generous amount of room in the fuselage as the elevator and rudder servos are mounted outside.

Ultimately, I think that the Hawk Sky is a great beginner plane and defintitely a good platform for an autopilot system. It looks good too!

Tuesday, June 4, 2013

Parrot AR Drone


This was a test of the Parrot AR drone using the ROS teleop package controlled from a PC running Ubuntu. The drone has a forward and downward facing camera. Currently working developing ROS packages to conduct image recognition for applications in SLAM (Simultaneous Localization And Mapping).

Wednesday, December 12, 2012

Ardupilot Mega first test

This is the first test of the assembled Ardupilot Mega autopilot board. All the sensors and GPS worked great. It was really cool to see the system tracking on the map in real-time while I drove. I was relieved to see that my soldering was ok, and that I had not fried something while assembling the board :)

Sunday, November 18, 2012

Tuesday, October 23, 2012

Ardupilot Mega kit

I assembled the Ardupilot Mega kit from in a couple of evenings. The soldering was not difficult, just had to take care in not applying too much heat onto the board. All the sensors seem to be functioning and the GPS got a fix. DIYdrones is a great place to start on learning all about drones, autopilot systems, etc. It does help to have some knowledge about RC planes in order to understand how to integrate an autopilot system into a plane. However, there are tons of resources on how to do that for a newbie. I will be putting this system on a Dynam Hawk Sky from I have read some rave reviews about this plane being used for an autopilot system and is a suitable substitute for the Multiplex Easy Star.

Wednesday, March 28, 2012

Xerox Project - Bag Security Device


A few months ago, some college colleagues and I started work on a prototype for patent number 7,535,358 B2, Method and Apparatus for Electronically Tracking Luggage. This was the Xerox fellowship project that our group had to complete for the semester. Having led the effort and tackling the hardware aspect, I decided to publish a small report on the work that was accomplished along with the design details. All the hardware that was used is open source and cost around $250. Here is a list of the hardware used:

Arduino UNO
Sparkfun GSM cellular module with patch antenna
Adafruit SD logger Shield
EM-406A GPS module
6 AA Rechargeable batteries
Photocell (light sensor)
Reed switch
Medium OtterBox for the enclosure

The device would be placed in a piece of luggage with the photocell and reed switch as the primary triggers. The SD card shield would record an entry onto the SD card every minute, with the Unix timestamp, light sensor and photocell value, latitude and longitude. The time is kept by the RTC on the SD logger shield. If the light sensor or the reed switch thresholds were surpassed, the system would immediately send an SMS through the GSM module to a Twitter account to indicate that the bag had been opened.

Here is the Arduino code for the final device…excuse me for the crappy coding :)

The logged data can be analyzed in the following ways:
The owner of the bag could upload the logged data using a simple uploader script to a website that we built. Once uploaded the data is parsed and plotted indicating spikes in the light sensor or reed swtich readings along with a Google map showing the locations in the spikes. It is always possible that the GPS module would not have a fix as the bag enters structures that inhibit GPS fixes.

A webpage had also been constructed to scrub the last 10 tweets posted via SMS to the Twitter account. Once scrubbed, the Tweet was parsed for the timestamp and latitude/longitude data. The information was then pushed onto the Google maps API where the location information was pinned. One of the hardest things to do was to fit everything in the Otterbox without shorting/crushing my electronics. It worked out well. The Otterbox is an awesome enclosure btw.

Please contact me with any questions regarding this project. I would be more than happy to assist. Thanks to all the folks at Adafruit for your assistance and awesome documentation. Keep doing what you are doing, you are inspiring budding engineers such as me to try stuff they have never before. Your prompt responses to my forum questions really helped in getting this thing to work. This device is a bit ahead of its time in terms of regulations to be used on an airplane. Once the Feds allow cellphone use on airplanes, this device would be able to fly!