TEMPEDOCLES
Temperature LCD readout & Strip Chart Plotter

The project

tempedocles - The name is derived from Empedocles of Akragas (ca. 492-ca. 432 BC):
    A Greek philosopher who was a subscriber to the Pythagorean school of natural philosophers, but without the mysticism of his contemporaries. He felt that we must use the data from our senses, even though they could not be completely trusted. He believed the Moon shone by reflected light. He postulated that all substances were made up of air, earth, fire, and water combined in different proportions. He is the first to have suggested the humoral theory, later adopted by the Hippocratics. He also proposed a crude theory of evolution. (Wolfram)
   

Schematics

Overview

This project monitors the evolution of the temperature during the day and plots it on a sheet of paper.

The project reads out the temperature from the Dallas DS1820 digital Temp. IC, displays it every 14 seconds on an LCD and plots the graph on a roll of paper. Feel free to look at the schematics, but pls.,
DO NOT USE FOR COMMERCIAL PURPOSES!!
The last section of this document contains some tips to make construction better & easier.

All parts of the schematics are discussed below.

Processor

The processor is an ATMEL AT90S1200 (or higher). Programming this one is easy with the free debugger and simulator from ATMEL. The processor controls pen motion, sensors and LCD. When a temperature-measurement is triggered (via timer or button) then the DS18B20 is read out and the info is sent to the LCD and the pen-drive mechanism (via the H-BRIDGE).

Assembly source code and binaries can be downloaded below.

Ultra-long timer

This is the Vital part of the system. It provides the heart-beat; A pulse every X seconds to processor -to update the temperature- and gives a pulse to the paper-feed mechanism -to advance the paper-.

H-BRIDGE

The H-BRIDGE is a way of controlling the direction of a DC motor (the TIP darlingtons here are a but over-dimensioned, but hey... :-) Use a simple two-line input to control the flow of the current through the DC motor (lead 1 and lead 2).
Using the motor in full stop (both input high) is possible but not recommendable (as it draws much current)...

Pen mechanism

The pen mechanism is pretty powerfull (but required lots of experimentation before I could come this far).
Disadvantage: It tends to 'drift' off if the system is turned on too long.. (this should/could be solved with an encoder & feedback to the processor so the processor knows the drift of the pen and could correct it, but I didn't implement this)

Paper feed

Simple mechanism, based on the motor from a remote-controlled car. I added a pressure wheel and some 'obstacles' to guide the paper straight.

Software

Following package contains .asm source- and .hex binary files for the AT90S1200. The .apr is the project file, needed for AVRStudio v3.

The software is based on the original code from Wayne Peacock:
DS1820 Temperature Sensor High-resolution Readout Demo
DO NOT USE FOR COMMERCIAL PURPOSES!!

The package contains changes for the clock (this is 10MHz) on my board (iso 4MHz in the original version) and extra routines for the motor control.

The original DS1820 readout code has been changed for my DS18B20-chip (this is a variant with higher accuracy.)

The software starts with a calibrartion (sets the pen of the LBOUND position, or, at -10ºC. For each measurement, it writes on the LCD and sets the delta, multiplied with a time constant on the H-Bridge. When the pen reaches the lbound or rbound boundaries, the system will re-calibrate and start over.

package - (src&hex 12kB)

Conclusion

Micro-processor, software and hardware are easy to build and maintain. One thing is not: the mechanics. This was the first time I was experimenting with robotics, but it's really a speciality that requires special tools.
Two tips that I've learnd when dealing with mechanics:

  1. Avoid belt transmission at all cost -they tend to slip
  2. DC motors are too imprecise to use in a plotter.
All is true. Sadly, I HAD to use both DC motor and belts, bcs my original plan to use a stepper motor fell apart: The motor was too slow, not enough powerfull and used to skip some steps. The drives both came from spare parts: a cassette drive for the pen and the motor came from a remote-control car. The latter already had all the gears in place, but for the first I used a flywheel from a cassetteplayer and some LEGO parts (see pictures above).

A alternative and prob. better solution if you want to create a similar project: Use the mechanism of a simple ink-jet printer. This machine has both a pen-mechanism and a paper-feed motor, plus all the accesoires (paper guiders,gears,…). Moreover, It's very low budget.


OH - jun 2003
If there are any questions? .... Mail me.