Cornell University  Spring 2002 
Professor Bruce Land
 
 
Digital Hourglass
 
Akosua Kyereme-Tuah     Andrew K. Lee     Susan Li-Wen Lin
 
[Home]     [Hi-Level] [Hardware] [Software] [Results]
 
 
  The hardware used for this project include:
  • STK200 programming board
  • ATmega163 MCU chip
  • Hyundai (model HG25504NG-01) 256x128 LCD display
  • Analog Device ADXL202EX accelerometer
  • Speaker (small one from cricket call lab)
  
  STK200 Programming Board & ATmega163 MCU  
 
  • The STK200 programming board was used so that we could take advantage of its A/D converter rather than building our own. The A/D converter worked well, except that we had to check the internal analog reference voltage (Aref) each time we used a different STK200 board and adjust the offset voltage in our code.

  • Atmel's STK200 can be configured to use either the internal reference voltage or an external one. It has an analog port that can be used for A/D purposes.

  • The user can configure the MCU to take in voltage at any one of the 8 pins on the analog port (ADC0 - ADC7) by changing the ADMUX control register. The ADMUX register also controls the alignment of the 8-bit sample in the 2-byte ADC data register (ADCH and ADCL). Once configured properly, the user needs to put the MCU into sleep mode, and use the [ADC_INT] interrupt to put the 8-bit sample into the one of the two ADC data registers.  The 8-bit sample should then be a digital representation of the analog input voltage. Please refer to the software section on more details.

  • We incorporated the push buttons on the STK200 board to allow users to dynamically increase/decrease the amount of sand in the hourglass. There was no need to debounce the buttons because we did not care if a button was held down or not.
Hyundai 256x128 LCD
 
  • We purchased a 256x128 LCD from All Electronics Corp. (CAT# LCD-58). We chose that particular LCD because it uses the SED1330F display controller.  This is the same controller as the one in all of the LCDs that were used for last year's final projects (Spring 2001). We felt that the reports of those projects could provide additional help in case the documentation of the LCD was not enough.
  • The LCD display was used to display the hourglass and sand. The documentation was complete and provided us all the details we needed to hook up the LCD to the board. The LCD required a -24 V power supply and a potentiometer for adjusting contrast. At times, the connection of the potentiometer to the breadboard would not be good enough for the LCD to display properly. We then soldered wires to the potentiometer and this worked a little better.
  • Since LCDs with the SED1330F display controller was used extensively for last year final projects, it wasn't difficult to obtain usable code to output something on the screen.  Here is the basic methodology in controlling the LCD, for anyone who is trying to use this LCD, but has no idea what the C-code means.  Of the 17 pins on this Hyundai model, aside from the normal supply pins and display contrast pin, there are 8 data pins and 5 control pins.  Whenever the user wants to talk to the LCD, he needs to send the "content" as an 8-bit word to the data pins, and force the LCD to accept it by adjusting the control pins (such as toggling the NOT write pin, /WR). Now there are two types of "content" words that the user can send, one being command codes and the other data codes. The status of the control pins tell the LCD whether the "content" sent on the data pins constitute a command code or data code. There are many command codes (refer to documentation for listing), each particular one puts the LCD in a particular mode, such as the system set mode, or the writing mode, etc. For example, if the user wishes to write the letter "A" on the screen, here are the steps (NOT CODE):

    1. Send command code for writing character on screen on data pins
    2. Put LCD in character display mode by forcing it to accept that command code using control pins
    3. Send data code ("A") on data pins
    4. Tell LCD to intepret word on data pins as data code through control pins

    So from the above steps, it would be correct to say that the LCD stays in one mode until it accepts another command code. Note that as written on the documentation, certain modes can accept more than one data code, and the controller expects particular sequences of data code.  So it wouldn't be wise to send a command code every time before a data code is passed in.

  • We were pleased to have found a program on Mike Rankin's website to test our LCD.  It proved to be invaluable for this project.
Analog Devices ADX202EX
 
  • We were able to obtain free samples of Analog Devices' ADX202EX accelerometer. The chip did not come in a DIP package, so Professor Land helped us solder the accelerometer onto one of his circuit boards to make it more accessible.
  • As described in the data sheet, ADXL202EX outputs both duty cycle modulated signals (DCM) and DC voltage, either of which alone is enough to gather the tilt angle. Since the circuit board Professor Land provided us only wired out the voltage outputs, we used that to get the tilt angle. The accelerometer outputs different voltages for each of the different positions that it is in (0 to 180 degrees). The raw output voltage range for our particular chip was between 2.15 and 2.7V, approximately.
  • We felt that the approximate 500mV range was enough to implement a 180-degree range because the A/D converter has a resolution of about 10mV ((1/256)*Aref, where Aref is about 2.4V). It also didn't seem necessary to implement the range in steps of 1 degree. Although the upper limit of the raw voltage range from the accelerometer is higher than Aref, we simply resistor-divide the output down to a range that is acceptable to the A/D converter.
  • In order to implement the whole 360-degree range, it was necessary to use two accelerometers. We put one accelerometer perpendicular to the other on the same plane, like the following diagram:

  • If we assume that the plane is oriented with 0 degree on the left, then as the diagram above suggests, when the voltages from both accelerometers are small, the tilt angle should be between 0 and 90 degrees. When both voltages are large, it should be between 180 and 270 degrees, etc.
  • The voltages enter an A/D converter. To use the on-board A/D converter, we had to measure Aref each time, since it is unique for each individual board. Aref is the maximum voltage that the A/D can convert.  Please refer to the software section for more details.
Speaker
 
  • We used the speaker from the one of the lab assignments of the course (cricket call generator lab). It is a tiny speaker that has no built-in amplifier. Hence we simply connected the output signal from the board to one of the pins, and connected one of the supply rails (Gnd or Vdd) to the other. Because of a shortage of ports, we put the speaker onto PORTD, which is also used for the buttons. We had to just be careful when setting the input and output bits of PORTD.

Schematic of hardware setup