In Chemnitz we have a planetarium for educational purposes
(Schulplanetarium Chemnitz).
The planetarium hardware itself is a rather old Carl Zeiss Jena ZKP2
model. This planetarium is more or less completely manually controlled.
It provides control panels with various switches and potentiometers
which control things like lamps and speed of motors.
The goal of this project was the development of a small interface that
allows to control some of the planetarium functions through a computer.
Of course, the purpose of this is not just to press a "virtual" button
or adjust a "virtual" potentiometer instead of manipulating the real
controls. Rather it is intended in order to prepare complete sequences
in advance to a presentation. So the demonstrator has his hands more
free during a presentation.
If all this control stuff would have been built from scratch, the best
solution would probably be to use one (or a few) micro controllers that
are controlled by the host computer via a standard RS232 interface.
So one would send commands to this micro controller that change the
state of a switch function or the speed of a motor.
However, that would require to write a corresponding piece of software
for the host computer.
Writing some simple graphical front-end would not be a big deal, actually.
More complicated is an appropriate sequence flow control.
But the problem is that it would be some stand-alone solution.
At the Schulplanetarium Chemnitz is already a system in use that appears
to be a de-facto standard system for audio-visual control systems.
This system is provided by the Swedish company
Dataton.
One subsystem is the so-called ANALOG SMARTLINK. It provides 32 analog
channels that have a voltage range of 0V - 10V (12Bit resolution).
There is also a sophisticated software that allows the generation of
complex sequences. So it was a natural conclusion to use the
analog channels as a basis and the planetarium control
can be nicely integrated into the system that is already in use for
other things (such as additional projectors).
Apart from the Dataton ANALOG SMARTLINK system it should be no problem
to attach almost every other DA-converter system - provided that it has an output range of 0V - 10V.
The planetarium is being manipulated by the control interface directly
at the control panels. That is, switches and potentiometers can be
bypassed by this control interface. The advantage of this strategy is
that the manual controls remain intact and can be used in parallel to
the computer control.
The fact that the interface connects to the control panels also makes
it very easy to integrate into the ZKP2 system. The only thing that has
to be done is to solder a bunch of wires to the planetariums' control
panels.
The hardware is more or less completely analog (i.e. without any
micro controller etc.).
All analog channels coming from the ANALOG SMARTLINK are isolated
from the planetariums' electrical circuit using opto couplers.
That provides a good protection of the Digital-to-Analog conversion
hardware for the case of some "events" inside the planetarium.
Switch Control
The switch functions are implemented rather conventional using Reed-Relays.
A relays can be activated with a high analog voltage and deactivated by
applying a low analog voltage. There's also an hysteresis implemented
that avoids oscillation of the relays in case of critical analog levels.
The interface can handle a total of 14 selected switch functions.
Motor Control
Controlling motor speed is not done by applying a desired voltage to the
motor (as done normally via the potentiometers), but by driving them
in a pulse-width-modulation scheme (PWM). MOSFETs are used as switching
devices. The higher the analog voltage the longer is the pulse width and
the faster is the motor speed.
There were some difficulties with a few motors that are bidirectional
and require to switch either +12V or -12V (the use of an H-bridge was
not an option there).
Five selected motors can be controlled by this interface.
As an additional feature, a threshold can be set for each motor
individually. It can be used to prevent the motors from being operated
in an illegal state (i.e. too short impulses that are insufficient to
get the motor rotating).
When treated just as a manual control panel replacement, there will appear basically
no problems or changes in behavior compared to conventional manual control.
As for the switch functions, there is no question at all in any case.
The behavior of the motor controls is not completely linear. That is, the motor
speed does not scale linearly with analog voltages from 0V to 10V. In fact, the motors start rotating
at around 4-5V. (Note: A design that stretches the usable range from 0V to 10V
with an ultra-linear behavior would be well possible, but it requires more active
circuitry and more expensive opto couplers.)
But that's actually not an issue. What appears to become a potential problem
is repeatability. The motors inside the planetarium are neither stepper motors,
nor does the planetarium return positional information to be evaluated in order
to implement a closed-loop control with feed-back information.
The only information given is the final projection. That is, when the demonstrator
wants to have a certain element at a certain position, he manipulates the
appropriate potentiometer until the element has its desired position. So to speak,
it is the human being that closes the loop (eye-brain-hand).
When playing back complete sequences via control of a computer, the demonstrator has
to hope that an object on the sky moves always for the same angle given
a certain motor speed and motor activation time. Whether this hope
can be fulfilled depends on the quality of the motors and mechanics inside the
planetarium, finally.
Practice will show to what extend these potential problems attribute to the
handling.
The following picture shows the top side of the board.
At the left side is the 37-position SUB-D connector for the Dataton ANALOG SMARTLINK.
Clearly visible is the bank of opto couplers right next to this connector.
At the right side there are two connectors that connect the interface to the planetarium.
Left to these connectors are some small fuses visible (blue parts). These are intended
in order to provide protection for the case of wrong installation.
The Reed-Relays can be found at the upper part of the board.
The radiators that can be seen
cover the MOSFETs that are used for the motor control.
Because the motors inside the planetarium do not draw a vast amount of current
(currents are in hundreads of milliamperes area),
MOSFET heating is not very high. The radiators are mounted nevertheless for
safety reasons.
The circuitry in the middle of the board is the actual motor PWM circuitry.
The PWM signals that are used for switching the MOSFETs are generated there.
At the lower left corner some voltage regulation circuitry can be seen that derives
the supply voltage from the planetariums' power rail. Modern switching-mode
regulator technology is used here instead of old and power-wasting linear
regulators.
The next picture shows the board from the bottom side.
There's not much left on this side. The most important parts are these visible
in the upper area. This circuitry drives the Reed-Relays.
