Nick Pennella
Project Write-up
Description:
For
this project I have chosen to implement Darrell Bishop’s Marble answering
machine. This idea was first brought up
in the early 90’s while Bishop was working on his graduate work. His idea was to have a tangible interface to
a very standard device, an answering machine.
The idea for this was fairly simple; represent each message that has
been left on a simple telephone answering machine by a tangible object, in this
case a marble. Once a message was left
on the machine, a marble would drop out of the machine into a tray. In order to play back your message, all you
had to do was simply re-insert the marble back into the machine and the marble
would trigger some mechanism that would play back the message. Unfortunately, Bishop never got around to
implementing his idea in any practical way.
He left of the idea and concept as an exercise for a grad student to
implement (someone like me). Using some
basic knowledge of electronics and some help from the logochip, I hope to build
the same, or similar interface that Bishop had originally perceived.
Subcomponents:
This
project can be broken down into a number of subcomponents that will make up the
total structure. The components for
this system would include a color sensor (for detecting which marble matches
which message), a Hopper system (to dispense the marbles), a digital recorder
(to record the messages), and a telephone interface circuit.
One of the major
sub-components to this system is a color sensor. Since color sensors are traditionally very expensive and are not
all that widespread, it would be my first and most daunting task to try to come
up with a design to implement one. The
design I have chosen for this circuit was based on a couple of designs that
used LEDs and refraction to take a measurement. The basic idea is to have a set of 3 LEDs shine on a given
target, one blue, one red, and one green.
Then by using a phototransistor you can take readings from the intensity
the amount of light that is reflected off for each one of the colors, and
deduce what color it is. The layout for
each of the items would look something similar to the following:

Where R is the Red LED, B is the blue LED, G
is the green LED and P is the phototransistor.
Once wired up, I would simply have to apply a current to each one of the
LEDs in sequence using the logochip and record each one of their
measurements. Certain “calibration” may
have to be done in order to get the type of response that I am looking
for. One choice that I did make when
making this circuit was to use a phototransistor rather than a more standard
approach of using a photo resistor. I
made this choice due to the fact that photo resistors need time to clear before
applying power a second time, otherwise you would only be able to take
approximately 2 readings in a second because of the average amount of time it
would take to clear the cell.
The
second component would be a drive system to get the marble from a storage
location to a tray to display for the user.
In order to do this I have created a new and original design that was
inspired by a trip to an average vending machine. The idea is to use a spiral from a notebook to drive the marble
from a marble storage location to a marble display tray, much like the way
potato chips are dispensed from a vending machine. This design involves using a length of PVC pipe along with a
spiral from a used notebook along with a motor. The basic design is illustrated below:

Using
either a handy cricket, or a logochip I would drive the motor to pull out a
marble from the funnel like hopper down the PVC tube and deposit the marble
into the tray. In order to be
efficient, right before the marble is deposited, I would take a measurement
from the color sensor.

A
piece that would be important to have would be the subcomponent that plays back
the messages. This is rather easy to
implement once I conquer the color sensor.
This would have to be nothing more than a simple ring of PVC cut about ¼
inch long with one of my custom designed color sensors at the bottom. The PVC pipe (the diameter of which is
smaller than the diameter of the marble) would be no more than a holder, which
the marble can be placed on so that the message could be played back.
![]()

![]()
The most important
piece to this puzzle would be digital device in which I would record messages
with. This device would probably have
to be some sort of purchased device, seeing I can’t really think of a way to implement
a cost-effective part with my limited knowledge of electronics.
The final and most worthwhile piece
is the telephone interface. Since the
telephone standards are widely published, I would implement a set of code to
deliver the required signals to the telephone line through the use of an RJ11
input. A cricket or logochip may be
required to create this as a bus enabled device to have the extra feature of
interfacing this with either a logochip or the device that will do the digital
recording.
Milestones:
|
Goal |
Date Due |
|
Come up with a design to move the marbles
from some sort of storage area into a tray. |
Completed |
|
Research a way to detect color using some
sort of photo sensing or reflecting or LED device |
Completed |
|
Implement a cricket bus enabled device to perform
color sensing, which will deliver a number (corresponding to a color) as it’s
output. |
3/22 |
|
Create the PVC tube structure that will
hold and move the marbles, and hook up a notebook spiral to a motor. |
3/23 |
|
Hook up the structure to a logochip, which
will act as a master to drive the contraption and take sensor inputs. |
3/24 |
|
Hook up a digital recording circuit that
will deliver a signal to the logochip once a message is received to spit out
a marble |
3/26 |
|
Implement the playback feature using a
piece of PVC and a color sensor and deliver its outputs to the master
logochip. |
Time Permitting |
|
Implement a telephone circuit to answer a
phone call and then deliver the output to the digital recorder |
Time Permitting |
Parts and Cost:
|
Part |
Quantity
|
Cost |
Description |
|
Jameco 120221 Phototransistor |
2 |
$1.95 each |
The phototransistor needed to implement the
color sensor |
|
Ultra bright Red LED |
2 |
$0.99 each |
Color Sensor part |
|
Ultra bright Blue LED |
2 |
$0.99 each |
Color Sensor part |
|
Ultra bright Green LED |
2 |
$0.99 each |
Color Sensor part |
|
Marbles |
1 |
??? |
The marbles needed for the device |
|
PVC piping |
1 |
< $3.00 |
Miscellaneous PVC needed (less than a foot
total) |
|
Motor |
1 |
$1.50 |
Motor needed to drive the spiral |
|
Digital Recorder |
1 |
??? |
|
|
RJ11 Telephone interface |
1 |
??? |
|
|
|
|
|
|
Integration Plan:
Now that I’ve described the total design, I need to talk about what I
need to do to put it together, and what I hope to achieve by the time Bot-Fest
rolls around. Although this task seems
a little bit daunting, I hope by the time I need to perform a demo that I have
at least a working ball dispenser, and a set of color sensors. Ideally, I would like to push a little bit
further ahead and get a digital recorder up and running but I don’t think there
is enough time (at least for this first demo).
In
order to piece the individual subcomponents together I was picturing myself
using a single Cricket as a master controller.
The cricket (through the use of the cricket bus) would drive each one of
the color sensors as well as drive the mechanism that dispenses the
marbles. The color sensors for this
project are going to be no more than a simple logochip bus device that will
take a reading when sent a signal and deliver a result. The result would then be sent to the Cricket
who would save in an array (indexed by color) a number. For Bot-Fest, I would like to have this
mechanism rigged in such a way that when the user presses a button a marble
would be dispensed, and the color would be displayed or beeped so that the user
knows what type of color was dispensed.
This would a huge step in making the marble answering machine a
reality.
The
next step once the Bot-Fest portion is complete is to handle the digital
recording of a telephone line. This
would be a multi-step process in which I would have to buy/integrate a circuit
that would either have or would have to be driven by a simple RJ11 interface
(possibly through the use of a phone device as a device on the cricket
bus). This functionality would have to
be added time permitting.