Thanks everyone.  I appreciate it.

Thanks.  I'm really just trying to figure out if it is possible to damage the IC by overclocking.  Seems unlikely.  The reason I might want to "take the chance" is that I can't accomplish my goals by limiting myself to 1MHz clock.  I need more speed, and that's why I'd like to go faster, at the possible expense of some accuracy.

I'm using a Microchip MCP3201-C ADC (analog to digital converter) in an audio sampling circuit.  The datasheet says that the maximum SPI clock frequency for communicating with the IC is 1.6MHz at 5V operating voltage.  For my application, I'd like to communicate faster at 2MHz.  I am wondering what the ramifications of exceeding the datasheet's clock frequency might be.

Can using a 2MHz SPI clock speed possibly damage the IC?

Or would exceeding the 1.6MHz speed simply cause the analog sampling to be less accurate?  I can certainly understand that the datasheet's stated INL and DNL performance characteristics would not be guaranteed if I overclock the communication.  But accuracy isn't necessarily critical for my application, and I can't hear any audio sampling quality difference between SPI speeds of 1MHz, 2MHz, or even 8MHz.

In short, if I'm satisfied with the performance, is there any danger that I can damage the IC by using a 2MHz SPI clock speed?  Thanks for any insight anyone might have!

I agree that if the requirement requires overlay onto an HD signal, then an Arduino isn't going to cut it.  I designed the Video Experimenter shield for Arduino, but this only works for a composite video signal.  It uses an LM1881 sync separator chip to detect the timing of the video signal, then in Arduino code you can overlay low-res monochrome pixels onto the video.  It works well for many applications, but is decidedly low-tech!
http://nootropicdesign.com/ve

I built this as a project a while back, but with Adafruit's encouragement, I brought it to market as a product.  Adafruit will be selling them, too. 

http://nootropicdesign.com/matrixbackpack/

If you have a large RGB LED panel, you know how the wiring can become a real rat's nest.  This tiny board snaps onto the back of the Adafruit 16x32 RGB panel to eliminate the mess.  The ATmega328 comes preloaded with a nice animation, and it's easy to control the panel with Adafruit's library.

 

If you have the Adafruit panel, this is a must-have!  Easy-to assemble kit, and also available pre-assembled.  All the product info is here: http://nootropicdesign.com/matrixbackpack/

You only need one serial port for the project.  I used Serial for debugging and Serial1 for the XBee.  Just eliminate all the calls to Serial, then change Serial1 to Serial. 

azad200517, I'm the designer of the Video Experimenter shield.  There is only enough SRAM in the Arduino to capture one frame of low-res monochrome, so you can't capture video, but you can capture individual video frames as demonstrated in this project: http://nootropicdesign.com/projectlab/2011/03/20/video-frame-capture/.

You can write the frame buffer to a computer over serial, but that is going to be very slow.

I built this XBee controlled robotics platform using a 4WD chassis.  There is no microcontroller on the remote, just an XBee connected directly to a joystick and two buttons for controlling the lights.  All project details and video at:
http://nootropicdesign.com/projectlab/2012/02/18/xbee-robotics-platform/

Using the OpenCV library for Android, an IOIO board, and an Arduino I was able to display camera frames from the phone on the Adafruit 32x16 RGB matrix panel display.  The frame rate is slow (the IOIO is the bottleneck), but it looks pretty good.  All project details here:
http://nootropicdesign.com/projectlab/2012/01/22/displaying-android-video-on-led-matrix/

Yeah, guys, the Hackvision is completely open source, just like all my products.  Not sure why anyone would think otherwise.  Just look at the product page:  http://nootropicdesign.com/hackvision/

Actually, you should not source or sink more than 6mA per pin on a 74HC595 and not more than 70mA for the whole chip.  It's not intended to drive LEDs.  If you have all 7 segments on at 20mA, that's 140mA. 

I designed the Digit Shield to use a single 595 and a 74LS247 BCD to 7-segment driver to multiplex four 7-segment displays.  Four of the 595 outputs control PNP transistors sourcing current to the common anode displays.  The other 4 outputs drive the 4 inputs of the 74LS247 to set the value of the current digit.  This approach only requires 3 Arduino pins, plus one more to control the decimal points.  The shield uses pins 2,3,4,5.

More info: http://nootropicdesign.com/digitshield/

Yeah, I thought long and hard about all of that, consulted a lot of people, my lawyer, etc.

Read this blog post about my anxious adventures: http://nootropicdesign.com/projectlab/2011/09/16/watching-the-feds-watch-me/

This is why I'm only shipping a bag of parts and nothing else....

You may have seen this when I made this as a project, but now it's a product.  The Defusable Clock kit lets you build your own dangerous looking device.  Only the electronics are in the kit, not the scary-scary looking parts. 



This is based on Arduino technology (powered by an ATmega328 with Arduino bootloader so you can hack it!).  Kit is $32.95 in the nootropic design store.

Also see the full product description here.

The vertical sync is not quite standard.  So some devices will not work.  I have not found a device that doesn't work, but your USB device may be too picky about the standard.

This looks really nice.  I'm really interested in the beat detection code, but I can't find it anywhere on your site.  Are you going to let us see the code?