This hybrid circuit uses a mixture of transistors, an IC and a relay and is used to automatically
open or close a pair of curtains. Using switch S3 also allows manual control, allowing for
curtains to be left only partially open or closed. The circuit controls a motor which is
attached to a simple pulley mechanism, to move the curtains. I first started this circuit over
20 years ago and apart from now using metal gears, very little has changed.
The circuit can be broken down into three main parts; a bistable latch, a timer and a
reversing circuit. Toggle switch S3 determines manual or automatic mode. The circuit as
shown above is drawn in the automatic position and operation is as follows. The bistable
is built around Q1 and Q2 and associated circuitry and
controls relay A/2. S1 is used to open the curtains and S2 to close the curtains. At power
on, a brief positive pulse is applied to the base of Q2 via C2. Q2 will be on, and activate
The network of C3 and R4 form a low current holding circuit for the relay. Relay
A/2 is a 12V relay with a 500 ohm coil. It requires slightly less current to keep a relay
energized than it does to operate it. Once the relay has operated, the current through the
coil is reduced by R4, saving power consumption. When Q2 is off, C3 will be discharged, but
when Q2 becomes active (either at switch on or by pressing S1) capacitor C3 will charge very
quickly via the relay coil. The initial charging current is sufficient to energize the relay
and current flow through R4 sufficient to keep it energized.
Q1 bias is applied via R3 which is tied to Q2 collector. As Q2 is on, the
collector voltage will be low, close to 0v and therefore Q1 and LED L1 will be off. As Q1 is
off, its collector voltage will be high, and Q2 bias voltage is applied via the chain L1, R1
and R2. The curtains should already be fully open.
If now S2 is pressed, the base voltage of Q2 will become 0 and Q2 will switch off. In
switching off, its collector voltage will rise to the supply voltage and Q1 will now be
forward biased via the relay coil A/2, R4 and R3. LED L1 will now be lit, relay A/2 will be
de-energized and as Q1 collector will be low, Q2 will be off and the circuit latched in this
At the same time as S2 is pressed, the trigger input of IC1, a 555 timer (normally held high
via R7 will be taken low. A timing sequence now commences. Duration is controlled by preset
P1 and C6 and the timing is adjustable between about 1 and 12 seconds. This delay is adjusted
so that the motor will run for sufficient time to fully open or close the curtains. The output
of the 555 turns on Q3, fed via R8 which now applies power to the motor via relay contacts A1
At any time the motor is in operation, and for any direction, LED L2 will always be lit.
Contacts A1 and A2 reverse the polarity of the voltage appearing at the motor terminals,
for more help on relays and switch contacts, visit this page
in my practical section. A running motor generates a back EMF and D4 and D5 prevent this
voltage from destroying the IC and transistors.
The medium power transistor BD139, Q3 is the workhorse in this circuit providing drive current to the motor. The case style is TO126,
which is shown left. Note that the terminals are if E, C, B order.
All transistors are like having two diodes back to back. The BD139 can be tested with a digital meter set to read resistance in range 2k.
Between E-B in forward bias (red lead base, black lead emitter) you will get a reading, my test BD139 showed 1.284k. The opposite way should
be a very high reading, greater than 2Meg. Testing between C-B also gave a reading of 1.274k, testing between B-C gave no reading. A shorted
transistor will be approximately the same reading across the E-B or C-B junctions.
If the toggle switch S3 is changed to manual mode, operation is slightly different as
outlined below. The bistable latch formed around S1, S2, Q1, Q2 and associated circuitry
operates the same way as in automatic mode.
S1 and S2 set or unset the bistable circuit which control relay A/2 and determine the
direction of the motor. In addition, as long as either S1 or S2 is held pressed, a bias
current will flow through either D1 or D2 and R6 into the base of PNP transistor Q4. This
small base current results in a larger collector current flowing via R9 into the base of
Q3. The BD139 will now be fully switched on and drives the motor as long as either S1 or
S2 is pressed. Hence it is now possible to partially open or partially close the curtains.
If you prefer a manual control then the following simpler electrical circuit is available.
The close switch applies power to the motor via the relay contacts. The 1N4001 diode prevents
the relay from operating. When the open switch is pressed, the relay is operated and power
is again applied to the motor, though this time the contacts have changed and the motor
will turn in the opposite direction.
This is not the best mechanical system and I am always open to new ideas and suggestions.
Originally I started out with a plastic worm gear and plastic 50 tooth gear, always in
contact with each other. Then one day I forgot the gears were engaged and manually closed
the curtains ruining the gears! I now use metal gears though plastic gears are perfectly
suitable for this project.
This mechanism is suitable for the plastic or metal flat strip curtain rails only. Two
pulleys are used at each end of the rail and a loop of string is passed around the pulleys
and kept in place by a tension spring, see below. To get a better "grip" on the string metal
pulleys with serrated flanges can be used; alternatively wooden pulleys may be used. The grooves
may be slightly ruffled with a file to aid grip.
The string will always travel in opposite directions and a small hook or piece of wire is
attached to each end fastener of each curtain and also to opposite sides of the string loop,
Each pulley is spaced from the wall with a bracket or small piece of wood. This is supported
by a bracket or block of wood. A small Axel passes through each bracket, one end will have a
collar, see below, the other end will have a gear that is driven by the motor.
At the motor end, I used two pieces of wood screwed together for the bracket. This now allows
the motor to be moved away from the driven gear and un-mesh the gears. As each curtain is
attached to the loop, moving just one curtain also moves the other curtain on the loop.
Friction - Friend or Foe ?
This mechanical design relies on one important property and that is friction. If there is too
much friction the motor may not move the curtains at all and the pulleys may just slip. If
there is not enough friction in the loop applied by the tension spring, the motor will drive
the pulleys which will just turn and not move the loop at all. To overcome this, I use a
silicone based furniture polish on the plastic rail, this reduces friction greatly and allows
easy movement of the curtains along the loop. To tension the loop of string, fasten one end
first to the spring then pull the free end of the string to tension the spring and fix it
with cable fasteners or glue, see below.
This may require some adjustment to get right. To answer the heading friction in this case
is both friend and foe, as the spring requires friction, the motor does not. The motor with
drive engaged is shown below.
Depending on the length and weight of the curtains, the motor may have to be changed. I used
a 12V hobby motor from Maplin electronics, I had to slightly enlarge the motor shaft with
some brass tubing available from most hobby shops. The torque of the motor was not great,
but if the output speed is reduced with gears, the torque (twisting force) is increased
by the same amount. A worm gear has 1 tooth, and I used a 57 tooth gear, giving a reduction
speed of 57:1. The torque of the motor (at the 57 tooth gear) is now increased 57 times. A
light grease or machine oil may be applied to the gears, too much and it will splatter all
over the walls and curtains!
This is best done with the curtains open, and motor gear un-meshed. Move one of the curtains
by hand. They should move easily and meet in the centre of the rail, if not apply some
silicone polish to the rail and alter the fastening on the wire.
Next switch the circuit to manual. With the curtains open, press the close switch. The
curtains should start to close as long as the switch is pressed and stop moving when the
switch is released. Then press the open switch. The curtains should now move as before but
in the opposite direction. If all is well, open the curtains with the switch and then
fully close them and use a watch to time this. The motor should be sufficiently slow and
take a few seconds (about 3 in my case but my room is small).
Finally open the curtains, adjust the preset P1 to minimum resistance and set S3 to automatic.
Press the close switch, the motor will run for a second or so and curtains will start to close.
Switch back to manual and open the curtains, increase P1 slightly and switch back to auto and
press close again. Repeat until the timing is sufficient for the curtains to close. Now press
open (with S3 still in manual) the curtains should be timed to open fully.
Should you have problems with this circuit, you first need to determine if its mechanical or
electrical. Mechanical problems will happen on both manual and automatic settings, and
be related to the opening or closing mechanism in general.
If electrical, check the power supply first, then L1 and L2 indications. If nothing works
at all build the single manual relay circuit above and once perfected, return to the
I am not mechanically minded so any suggestions or improvements towards a better mechanism
can be included here; or if any of you have also made electric curtains, I will be happy to
display your work.