Bob Control Board Latest change 2017-09-06
Discussion of the electronic circuits involved.
Download the zipped Eagle files for the BobControlBoard . Or the PDF's:
Schematics and board layouts are
drawn with Eagle 7.7.0 Premium. Although I have a licence for larger
boards I tried to limit the boards such that they can be handled by the
free version of Eagle. As I did not intend to make real PCB's for this
project I did not route the boards, I used Eagle just to place the
components in a more or less logical way and did the wiring by hand.
The Eagle files are likely to contain inapropriate footprints.
Bob Control Board.
The functions on this board are
- Amplifiying the signal from the Center detect coil.
- Amplifying the signal from the Rim coil.
- Amplifying the signals from the Hall sensors.
- Pulsing the current through the Drive coil.
- Pulsing a ring of Led's on the body of the Rim coil.
- Pulsing an output to trigger a camera for the time-lapse video.
Fig 1. Bob Control Board.
and Rim detection.
Fig 2. Amplifier for Center signal.
The amplifiers for the Center and Rim detection signals are identical,
IC2 D and C. The signal is low-pass filtered with R9/C7, a timeconstant
of 1 msec, or 160 Hz -3dB. The gain can be adjusted from 1 to
101 with pot R17. In case the gain is set to maximum there will be an
extra low-pass timeconstant of 10 ms or 16 Hz -3dB. At lower gains this
timeconstant will be shorter. The amplified signals are AC-coupled to
the Arduino's analoge inputs with a time constant of 60 ms, or 2.6 Hz
-3dB, and clamped against GND and +5VD with the diodes. The DC levels
are halfway the +5VA voltage.
The firmware will sample the signals at a 3.3 kHz rate.The Center Pass
is detected in two steps, first when the bob approaches the Center Coil
the signal will rise, and when the bob is excatly over the coil the
signal goes downwards through the midscale level.
The Rim passage is detected by finding the peak value, that is when the
signal is falling after going up. The Rim passage is only detected when
the bob goes out from the center. The return passage is ignored.
Note that the coils must be connected such that the signals are
positive going when the bob approaches the coil.
The choise of the opamps is very uncritical. You can use any quad 741
[Here we see the signals on the oscilloscope]
Fig 4. Amplifier for Hall signal.
The signal from the Hall sensors are amplified by IC1, much the same as
Center and Rim, although now inverting. The signals are DC coupled to
the Arduino because they are changing very slowly, and not at all during
certain callibration procedures.
For each channel the gain and offset can be adjusted individually. The
adjustment procedure is described here.
Be sure to mount the Hall sensors and the magnet such that the signal
on the output of these amplifiers rises when the magnet approaches the
Fig 5. Pulsing the Drive Coil.
For this I chose a bipolar darlington power transistor, but a N-channel
power fet can be used too. D9 protects the transistor from a possible
high flyback voltage from the coil. We talk about pulses of a few
amperes during a quite short time.
Be sure to connect the Drive coil sucht that there will be a repelling force on the Bob.
On the body of the Rim coil are 12 Led's, which will be pulsed when the
Drive coil is engaged. I made a separate circuit for this because it
may turn out that using the Drive-pulse itself will not produce a
sufficiently visible effect.
Each time a center pass has been detected the state of this led is toggled.
It is primary a diagnostic tool to see that center passages are
The intension is to have a video camera looking down on the Floor Unit
for a real time streaming video. This trigger signal is provided to
allow time-lapse storage of one image, each time the bob is at the
maximum of its swing. When these images are played at a normal rate you
will see the bob rotating around the Floor unit.
When the settings are updated the trigger is given at the end of that
half period, and from that moment at the end of each other half
Connections are provided to use a 'Precise Clock' signal for the timing
of the Center and Rim detect and the Drive coil.
This is a future option where the clock signal will be phase-locked to
a Standard Time Transmitter, in my case the German DCF77.