Glowing Cocktail Stick

Description

A stick that glows when one end is put in water
(Implemented and put online in half a day).

Introduction

Two years ago Petter (old friend of mine), and I was talking design over a drink. He had a faschinating Cocktail Stick that looked hi-tech, and was made from a harddrive motor rotor and an steel axle. I was joking about Petter beeing an landscape architekt usually designing much bigger stuff then I do ;-). So we started to talk about intelligent cocktail sticks with LED's.

This summer Petter is finally back from a year in Canada, and since the brain works in mysterious ways:
Here is the first implementation!

TODO: Add an SMD Atmel 3v processor and change to a 3-color LED. Use pulse-width modulation to produce any color. Use a g-sensor and a crazy algorithm to produce strange color transformations depending on what you do wiht the cocktail pin.

We talked about making it from glass or lexan, but I guessed a new probe would be a little nerdier :-)

I guess an other reason that I made this stick is that I just started to solder surface mount components, and anyone has to understand the necessity of implementing any Cocktail Stick with SMT ;-).

Desing

Which is the limiting factors? 
Li Cell CR2032 3.005v
LED SMD Green 1.82v forward, or Orange LED 10mm ~2v. max 20mA
SMT BC817-25 NPN 38 mV forward V only!! 0.25W max =>
0.25W/1.2v = Max 0.2 A through the transistor, LED is limiting.
1.2V/0.02mA =60Ohm => 0.024W => any SMT resistor can handle this.
Wather resistance was today 0.3-0.7 M Ohm.
A test reveals that the BC817-25 amplifies enough but that we get poor results without any transistor. When I changed from 100 Ohm to 50 for the LED, I noticed that the difference in brightness was not as big as I would have guessed. ---> How much current will a Li cell deliver?

Ooops! 100 to 1000 Ohm is considered a pulse load. The lowest load mentioned in the tables is 3.5 k Ohm. If the LED and transistor uses up 2v, then 1 k Ohm in our circuit will result in total circuit resistance approx 3 k. We can of course use 3 parallell coupled cells, or CR2032. But it will be much bulkier, so lets try to avoid that for now.


This is the final circuit with designlayout, sure it looks professional ;-)
The Li Cell is keept in place by the LED-legs, and a 2-pin "Straight Female Header".


The negative side is insulated from the right leg of the LED by Tape (or swe: bokplast). First a small piece to cover the + terminal closest to where the thin wire will go. then add the thin wire, then a piece to cover the entire negative part of the cell.



The SMT transistor and resistors are soldered together. Very thin copperwire is then soldered to the leads, then they are glued to the Female Header, the wire from the 1k resistor is soldered to the right leg of the Female Header.

The wirefrom the 10k resistor is routed along the side of the probe, down towards the probe-tip. It is melted into the probe, by using any part of a soldering iron that is not too hot. The wire marked with the Ground symbol is taped to the negative battery terminal without any part of it or the battery tuching the LED's negative terminal = right lead.

Result

Cost is not an issue in this project
Last modified: Mon Jul 18 00:51:08 W. Europe Daylight Time 2005