Home | Product Categories | Breakout Boards | SEN-09695

Actually, what I’m seeing is not the OCVF bit set but when I read from registers 0x00 and 0x01 I acquire the same data for both. I’ll have to dig in deeper.

This should help http://www.eetimes.com/design/analog-design/4009622/The-art-of-capacitive-touch-sensing

The Wire library for the Arduino is not so great, and I would recommend directly using the AVR example.
As far as the epoxy is concerned, it’s fine. It has a dielectric constant of approximately 4, but your biggest issue will be air bubbles trapped inside the epoxy. If you can get those out you’ll have fairly uniform results. For proximity sensing, your best bet will be to have a large electrode, the largest charge current (63 uA) and the largest charge time (128 ms) of this sensor. If you have big enough electrodes, you need not use the proximity sensing mode of this sensor, which essentially ties all electrodes together to act as a proximity sensor.
For a beer pong table, if you expect to detect the ping pong ball when it lands inside the cup, forget it. The change in capacitance would be so small that you wouldn’t be able to separate it from the noise. A full cup verses an empty cup, however, is doable. The calibration is the hardest part in working with these sensors.

Yes, but you’ll need to very carefully cut one of the traces on the back of the board. On the bottom of page 3 of the datasheet they specify what’s required to change the address. Just connect the address pin to VSS, VDD, SCA or SCL to change it.

(creates public wishlist)

Quotes from datasheet:
* “ Active Low Open-drain Interrupt Output
* ”[It] is triggered any time a touch or
release of a button is detected".

If you have access to an oscilloscope you could look at the signal coming off SDA. Otherwise, I would suggest trying to write a value to a register on the sensor and then read it back. If it’s coming back as 0x00 or something other than what you wrote to that register, then you can definitively make the statement that something is amiss.

You make your own. They’re typically pads on a pcb, or a matrix of Indium Tin Oxide (or Aluminum Zinc Oxide) traces that are deposited on the inside of the glass covering a display screen, as in the case iPads, iPhones, smart phones, etc.
All you really need is a wire connecting to a conductive material. I’m currently using thin copper sheeting. Some people use aluminum foil with an alligator clip connecting the foil and the wire. It all depends upon your application.

There is now! Link.

The Arduino Wire library does a lot of things right, and one thing wrong. The right things are that it implements a correct I2C interface that grounds the lines for 0, and sets them to high-Z inputs for 1, so it will work correctly for 3.3V I2C parts. The wrong thing is that the library turns on the weak pull-up (WPU) resistors by default, which connects the I2C lines via 20K resistors to 5V. There’s a fair amount of controversy over this, and the library will likely change at some point in the future to at least make this non-default behavior. It’s also possible to modify the library yourself to turn the WPU resistors off.
But in practice, the WPU resistors are so small that they can’t transfer much current back to the device (in fact they’re so weak that they don’t really make very good I2C bus pullup resistors). And, if other I2C bus pullup resistors to 3.3V are present, as they are on this board, those will overpower the WPU resistors. For example, this board has 10K pullups to 3.3V. If you parallel those with the WPUs to 5V, no more than 3.8V will appear on the I2C lines. This is above the recommended maximum, but not as bad as putting 5V straight onto those lines. I’ve personally never had a problem running 3.3V I2C parts on a 5V Arduino when using outboard I2C pullup resistors to 3.3V, but of course your mileage may vary.