Friday, December 9, 2011

PCB PROTOTYPING : THE ART AND CRAFT (part - I)

PCB prototyping is an important skill that every electronic enthusiast should learn in order to make some real sense out of his designs.Most of our designs are based on single or double sided PCB boards because more then two layers are almost impossible to be fabricated at home .Besides this there are some other things that an electronic enthusiast should know like high speed electronics and RF circuits have many issues and even the slightest of mistake can lead to the whole design failure as the designs would work on simulators but they fail in the real world due to bad prototyping.
    Most of the people have used toner transfer method to fabricate single as well as double sided boards ,SMD boards could also be fabricated using this method but as we move to more complex designs this method fails badly.Here i would talk about a more suitable and industrial method for PCB fabrication .This method is based on the principal of photo-lithography and is widely used in electronics industries and this method is so cheap to work with and so efficient to give you professional class designs and you would never want to go any pcb prototyping services again ... i would be coming up with detail description of the work i have done using photo-lithography method in the next post and would be giving you a HOW-TO on the famous BUS PIRATE pcb layout using the photo method....
see you next time....till then keep learning...

Here are some reference books you people would love to learn from:
  • printed circuit board designer reference by Christopher t. robertson.
  • Printed circuit handbook by Clyde f.coombs. 
        For high speed designs you may refer these :
  • Signal integrity issues & printed circuit board design by Douglas brooks
  • Signal integrity-simplified by Eric bogatin is the book in depth about signal integrity..
  • High speed signal propagation by howard w.johnson.

Sunday, December 4, 2011

Kinect : The Myth And Its Truth

You would be surprised to know that microsoft's KINECT being the most favorite device for gamers all around the world this year has a very simple yet sophisticated underlying technology which is unfortunately being misunderstood by many people who tried hacking the device.....there is a myth about KINECT that it uses TIME OF FLIGHT Sensor (IR Camera) to calculate the distance of each and every point in the point cloud . Well i must tell you that its a total bullshit....


First of all the price of kinect is just 169$ which makes it around 8500 (INR). now a time of flight sensor would cost much more to cover the whole price of making the device plus the profit margin. besides the IR CAMERA it also has a RGB camera + an IR projector.It would have been impossible for microsoft to sell the device at this price if it would be having a time of flight sensor.
   Actually KINECT uses 3D MODELLING THROUGH STRUCTURED LIGHTENING. As we know that there is an IR projector in the KINECT which projects a known Pattern of IR beams , now we will consider a most simple pattern of POINT MATRIX in which each point is equidistant from its neighbouring points.when this matrix pattern falls on an object the points which were equidistant and well alligned before would now look garbled and misalligned .This misallignment of points is the key to the whole 3D Recognition thing. When the IR camera captures the whole pattern which is ofcourse different from the projected pattern ,the image processor in kinect calculates the changes in the captured pattern like the amount by which the points are misalligned and then on the basis of these calculations the 3D model of the world around kinect is generated. you must have studied reading topological maps in 8th standard where were asked to calculate the steepness or height of a hill or a mountain by just observing the CONTOUR LINES of that physical structure .The more the contour lines are far from each other the more STEEP is the structure... The same happens in this Structured Lightening Technology. The point cloud captured through the IR Cam is processed by the Image processor and using the algorithms contour points can be differentiated .In fact the IR cam could be replaced with a simple WEB CAMERA with above 1.3 MegaPixel res. and 30 FPS frame rate and a x-ray(IR FILTER) film patch in front of its lens. The RGB Camera is optional but it helps in Gesture recognition.For more information read this wiki (http://en.wikipedia.org/wiki/Structured-light_3D_scanner) .

I would be coming up with a DIY structured light 3D scanner project Soon.....just wait and keep learning ......