Notes - Alex Hiam - Fall 2011
Designing and Building a Custom Handheld Embedded Device
--------------------------------------------------------

9/9:
-Completed a snazzy block diagram of the device.

-Decided (for now at least) on the Atmel ATmega328. Still unsure about
 other components.

-Another big consideration is the power supply level. Here's one simple 
 approach that I hope to be able to get away with:
  
   USB ----> LiPoly Charger ----> LiPoly ----> 5V DC boost ----> VCC
 (5v in)    (MCP73831 or eq.)     (3.7v)     (MCP1640 or eq.)    (5v)

 The problem with this scheme is that I would only have a 5v regulated 
 supply, which means extra components to communicate with any 3.3v 
 components. The other simple options I can come up with right now are
 to use a buck/boost:

   USB ----> LiPoly Charger ----> LiPoly ----> 3.3V DC buck/boost ----> VCC
 (5v in)    (MCP73831 or eq.)     (3.7v)   (v<3.3v, v>=3.3v -> 3.3v)   (3.3v)

 and have a single 3.3v regulated supply, but I could then not use any parts
 with a Vmin > 3.3v, or simply put a regulator after the boost:
  
   USB ----> LiPoly Charger ----> LiPoly ----> 5V DC boost ----> 3.3V regulator ----> 5v/3.3v supply
 (5v in)    (MCP73831 or eq.)     (3.7v)     (MCP1640 or eq.)    (LM317 or eq.)       


9/10:
-I'm leaning towards using a digital temp/RH combination sensor, mainly because 
 I can get a more accurate reading without having to do my own calibration. This
 may mean more software if I choose a part that doesn't use a standard communication
 protocol; I could use the DHT11 (same as the RHT02), as I already have a mostly 
 written library for it from another project. It's big though... perhaps its 
 smaller cousin, the DHT11.

-Thinking about using the ST7565 LCD (from adafruit). It uses a nice serialised 
 protocol; the only problem is that it has no on-board memory, so the MCU has
 to buffer every pixel, which may be a space concern. Maybe I can keep that in
 the EEPROM?


9/12:
-Thinking about range sensing - would be cool to have both long-range ultrasonic 
 and short-range infrared. 
-Ultrasonic rangefinder modules are pricey... I think I'll take a stab at writing 
 a library for ultrasonic rangefinding with a transducer and figuring out any 
 supporting hardware.


9/13:
-Looking for parts - thinking right now about the analog input circuit and the power
 supply circuit.
-What I'm thinking for analog in:
 -Multiplexed input using TS3A5017DR dual 4:1 analog mux.
  -One mux for input to select mic, VLF ant., or input (3.5mm phone jack?)
  -Other mux for input from mux 1 through choice of high-gain preamp, direct in, diode protected divider ...
 Probably overcomplicated...
    

9/14:
-A cool idea:
 -Hardware scheme:
  -An IR receiver/decoder -> software-serial in
  -Software serial out -> NPN driving an LED 
 -Software scheme:
  -User points remote at IR receiver and presses a buton on it.
  -MCU remembers data received from button press (probably store a few labeled commands)
  -Transmit stored commands through IR LED.
 e.g.: 1. Select 'learn command' on device.
       2. Point your stereo remote at device and press 'play/pause' button.
       3. Use device to play/pause your music


9/15:
-Simplified analog input scheme:
 -One analog channel for voltage measuring through zener protected divider
   (digitally selectable range?)
 -One analog channel for VLF/audio using FSA3157 digital SPDT as such:
       
                                /---[VLF ant.]
   [MCU]<---[NPN amp]---[FS4157]
                                \---[mic]