I could not help it, I loved the VBPW77 phototransistor
the first time I had the chance to use it. Housed in a high reliability
hermetically sealed TO-18 metal can package with gold plated lead and a cute curvy glass lens
protruding out of its cap, its appearance alone gives me a lasting impression of its superiority compared to its cheaper plastic cased
Its glass lens limits its field of vision to about 10
degree range, a useful characteristic because in my application, I want
the sensor to keep its focus on the job in front of it, and not be
distracted by stray lights that may come in elsewhere. It also has
respectable sensitivity, in fact, in most cases; I can use it without the
aid of an amplifier. I know right that instant that this is the kind of
phototransistor I want to use in my projects whenever and wherever it is
Line Follower Sensor using VBPW77 phototransistor
I got immediate inquiries to people wanting to use
this device. Apparently, everybody is building a line following mobot, and
want me to give an application circuit using the BPW77 as line sensor.
To get the job done quickly, I assembled my first
prototype on a prototyping PCB. I connected a 10K resistor in series with
the collector terminal of the phototransistor, and used a 5mm LED with a
100 ohm series resistor as the illuminator, as shown in the figures below.
Figure 1. Sensor Test Circuit. Circuit too simple? Donít worry, we will complicate the circuit soon
Figure 2. The phototransistor sight is
directed towards the light spot emitted by the LED.
The VBPW77 transistor and the LED are physically arranged such
that the phototransistor is directed towards the center of the light spot
on the reflecting surface being emitted by the LED.
Figure 3. Sensor Test Setup-
The sensor is tested by swiping it over a white bond paper overlaid
with black electrical tape strip of varying width. The light meter
measures the light ambience.
Figure 4. Close up view of the sensor circuit. I borrowed a prototyping
board being used in another (different) project; all other components not
shown in this picture are not part of the sensors and should be ignored.
A white bond paper overlaid with a short strip of black electrical tape
was used to test the line sensor. Three tapes, each with widths measuring 18, 9, and 4.5mm,
respectively, are used as test tracks. The output is monitored by Tektronix
TDS744A oscilloscope. Tested at a distance of about 5 mm from the surface,
and with a light ambience of 100 foot-candles, the sensor yanked out an
output waveform as it is swiped from right to left as shown in the figure
Figure 5. Sensor response at 100 foot -candles
The first pulse is the sensor response for the
4.5mm width line, the second pulse for the 9mm line, and the third and
widest peak for the 18mm line. I increased the light ambience to about 1000
foot candles and repeated the test, resulting in a waveform as shown below:
6. Sensor response at 1000 foot candles ambience.
The first pulse is generated by the passing of the sensor over the 9mm
width line, the second pulse over the 18mm width line. The pulse for the
4.5mm line is altogether gone!
Recommended MOBOT Line Sensor Circuit
A well-lighted office is about 25 foot-candles bright.
Under a shade on a mid afternoon sun is about 1000 foot candles bright. The
oscilloscope waveforms shown in figure 5 and 6 tells me that, under these
two conditions, the sensor can be used to reliably detect line widths as
small as 8-9mm. Hence with just the addition of a Schmitt trigger buffer as
shown in figure 7, we can complete a working line sensor!
Figure 7. The recommended line sensor
circuit. Not shown, for circuit clarity, are the IC power supply
connections. Do not forget to connect them. It will also do much good if you
connect a 0.1uF capacitor as close as possible to the Vcc-GND pin of your
A 74HCT14 sensor is chosen because of its relatively
low trigger point compared to the other Schmitt triggers in the high speed
CMOS family. Note that you cannot use a TTL logic here! Also note that the
sensor output will be on low state when it is over the black line.
Due to component value tolerance (or some other
reasons), it may sometimes be necessary to tweak the value of R1 to obtain
better results. Experiment with the value of R1. Higher resistance results
in greater sensitivity but makes it more affected by stray unwanted light.
Lower the resistance value if your sensor is to work in very brightly
A camera flash may momentarily trigger the sensor. In
our experiments, we measured a camera flash triggered output of as long as
10mS wide. Hence, your mobot software must take this into account Ė it
should ignore pulse detection of less than 10mS. Otherwise, the flash may
dazzle your mobot and send it off course.
The line sensor will not work under direct sunlight! (July 29,2003)
I got news that IC stores ran out of 74HCT14 stock.
You can use CD4093 or MC14903 schmitt trigger NAND gate as a substitute,
noting the consequence that the circuit may not work well under very bright
ambience. You can experiment to improve the circuit by using a super bright
LED for D1 and lowering the value of R1 to, say, 2k2 ohms. Select the final
value of R1 so that the VBPW77 fully saturates when the sensor points to the
non-black portion of the track. (Aug. 5,2003)
By: Henry Chua
mail me firstname.lastname@example.org