The latest addition to my collection of Infra Red (IR) Repeater circuits. The Mark 5 is a much improved version
of the Mark 1 circuit and has increased range and sensitivity. It is also immune to the effects of ambient light,
daylight and other forms of interference and has a wider frequency response for remote handsets using modulation
up to 120kHz.
R1,R2: 5M6 RESISTOR (2)
R3,R5: 3k3 RESISTOR (2)
R4: 120k RESISTOR (1)
R6: 220R RESISTOR (1)
R7: 47k RESISTOR (1)
R8: 120R RESISTOR (1)
R9: 10k RESISTOR (1)
R10: 2K2 RESISTOR (1)
R11: 100R RESISTOR 1 W (1)
C1,C3,C4: 22n polyester CAP (3)
C2: 100u electrolytic 25V(1)
C5: 100u electrolytic 25V(1)
Q1 BC107 (1) alternatives, BC107A, 2N2222, 2N2222A
Q2 BC109C (1) alternatives, BC109, BC549
D1: 1N4148 DIODE (1)
D2: Red LED (1)
IC1,IC2 CA3140E opamp (1)
IR1: SFH2030: (1)
IR2,3: TIL38 (2) or similar.
Alternatives to SFH2030
In some common Countries the photodiode SFH2030 may not be available. I have recently heard from a viewer in the Ukraine
using an alternate photodiode FYL-5012PD. The FYL-5012PD datasheet is on the link below (in Russian only):
The FYL-5012PD is available from Elcom Electronics
This circuit has been designed from the ground up, and contains a current to voltage converter, a wideband infra red (IR)
preamp and photo emitter driver stage. The circuit receives and re-transmits the entire baseband signal from a remote handset,
and due to its wider filtering of 30-120kHz should work with most handsets.
Static light levels such as daylight that do not change are filtered. and not passed by this circuit. A major problem
with the early Mark 1 circuit is that it reacts to all light sources; ambient light therefore produced a continuous signal
from the IR photo diode and was amplified by the rest of the Mark1 circuit.
I have published a modification to the original Mark 1 circuit, click here to view
. The Mark 1A
filtered out ambient light, this Mark 5 circuit does the same but with extra amplification.
It is difficult working with signals you cannot see. In designing the Mark 5, one problem was how to
differentiate between daylight and an IR signal. Ambient light produces a steady continuous signal, changing very little
over several hours. A signal from an IR handset contains control pulses modulated with a carrier frequency (typically
36kHz). My solution used here, is a simple RC filter formed by C1 and R3, which blocks the steady ambient light
but allows modulated light to pass.
At low frequency e.g. 50Hz the impedance of C1 is high, around 144k. The voltage gain of
inverting op-amp IC2 is approximately R4 / R3, but at low frequency C1 is in series with
R3 so the gain is now 120k / (3.3k + 144k) or less than unity. Daylight or ambient light will
change slowly over several hours, in frequency terms this is almost like a DC signal and blocked by the
high impedance of C1.
Now imagine a signal from a remote modulated at 36KHz. At this frequency the impedance
of C1 is very low, around 200 ohms. This has little effect on the input impedance of the
op-amp stage and voltage gain will now be R4 / R3 or about 34 times. The impedance of capacitor
C4 also supplements noise rejection, allowing high frequencies to pass and impeding low
frequency signals at Q1 base.
Light photons are received at IR1, this is an IR photo diode type SFH2030. A SFH2030F, which
contains a daylight filter,may also be used instead of the SFH2030. The photo diode is reverse
biased and when light strikes it, the energy of the IR signal releases additional charge carriers
within the diode, allowing more current to flow. This current is amplified and converted to a
voltage by the first CA3140 opamp, IC1. IC1 is wired as a current to voltage converter and amplifies
the diode current, which is converted to a voltage at the op-amp output, Pin 6. see diagram below.
In an ideal current to voltage converter the output voltage would be the product Rf multiplied by the
input current. The non-inverting input would be tied to ground. In the Mark 5 circuit the output
voltage is iR1 or about 5.6 Volts/uA appearing at pin 6 of IC1. The current generated by the SFH2030 photo diode
when receiving a signal from a handset several metres away is less than 50 nA and requires the
extreme high input impedance to avoid shunting the signal. There are two reasons for using the CA3140,
the first is its high input impedance, over 1000G. The second reason is that normally the non-inverting
input would be at 0V when working from split + and - supplies. In this single supply version the non-inverting
input is returned to negative supply via R2. This can only be done with a MOSFET input, hence
the choice for using the CA3140.
IC1 converts all current from the photo diode IR1 into a voltage, appearing at Pin 6 the op-amp output.
A DC voltage will always be present at the output, even in constant daylight. In daylight this can be as high
as 11V DC (if working from a 12V supply) and as little as 0.5V DC in a dark room with a 40 Watt desk lamp. This
DC voltage is not passed to the next stage. Although the SFH2030 is most sensitive at infra red wavelengths,
it will produce tiny currents from daylight and also 50/60Hz noise fields from fluorescent and mains lighting.
To minimize this, C1 and R3 form a high pass filter, allowing
30kHz and higher signals to pass but blocking low frequencies. The impedance of C1 increases with
decreasing frequency being 31kohms at 50Hz. Daylight for example, produces a constant luminance, changing
slowly over several hours, to which the impedance of C1 is effectively infinite.
The signal voltage from IC1 is now further amplified by IC2, gain being the ratio R4/R3 or 31dB. All
op-amps have a limit called the gain bandwidth product. The gain will fall to unity at the highest
usable frequency and be a maximum value at dc. Between these limits the gain falls with increasing
frequency as shown in the bode plot for the CA3140 below:
Looking at the chart above, at 100kHz the maximum gain can only be about 30dB. However this is ample
and boosts the received range of signals from a remote handset to the photo diode which have worked well
up to 4 metres apart. Because R5 is returned to the negative supply a MOSFET input op-amp must again
be used. The output is again filtered by a high pass filter comprising C4 and the associated input
impedance of Q1. R6, C2 and C3 provide decoupling for the IR pre-amplifier, C3 is in parallel with C2
because an electrolytic is not always a low impedance at high frequencies.
The IR driver stage is comprised of Q1 and Q2 and associated components. The output is arranged so
that with no input signal, Q1 is on and Q2 off; the visible LED, D2 will also be off. With no
signal the 47k resistor biases the driver transistor, Q1 into full conduction. Its collector voltage
will be near zero volts and the output transistor Q2, which is direct coupled to Q1 collector will
therefore be fully off. Power drain will be minimal.
Resistor R11 limits the current through the photo emitter diodes. The range
from photo emitter diode to the equipment to be controlled has proved successful at over 4 metres when
powered from a 12 Volt supply. D1 quickens the turn off time of Q1, improving the output fall time.
D1 can be omitted but the circuit performs better with D1 included. A simulated transfer characteristic
is shown below:
AC Transfer Charcteristic
The output is measured between Q2 emitter and ground. A simulated transient response is shown below.
Three graphs are produced with input waveforms of 40,80 and 120kHz.
Please note that the above waveforms are simulated using a perfect square wave input, with rise and
fall times of zero seconds. The output is measured between Q2 emitter and ground with a 200 ohm resistive
load. In the real world, the cable to the remote photo emitter LED's will contain both capacitance
and inductance. This will increase both rise and fall times of the output signal. As with the Mark 1
circuit I recommend using speaker wire or bell wire to be used to cable the remote photo emitters. You
can also use twisted pair telephone or CAT5 cable. Twisted pair cable also has the advantage that the
successive twists will also minimize interference.
Note that the veroboard layout below only includes the components from the left of the schematic to C4, I had
Q1 and Q2 on breadboard during this testing phase.
Setup and Testing:
Click both images for a larger view.
There is little to adjust in this circuit. First I suggest disconnecting the wiring to the emitters
IR2 and IR3. Try and test first in daylight with no mains powered lighting. Apply power to the circuit
and red LED, D2 should be off. Aim a remote in the direction of IR1 and press any button on the handset.
D2 should light and be seen to flash when a button is held on the handset and go off when released.
When testing in the evening or with any mains powered lighting, the light will always contain noise
at line frequency 50 or 60Hz. While the response at line frequencies is reduced you may notice that
D2 is lit faintly. This is normal, and when testing with a remote, D2 should flicker brightly.
If all is well reconnect the wiring to emitters IR2 and IR3. Without lenses, the light is quite
directional and so you will need to aim it carefully at the remote equipment you are controlling.
A digital camera, or camcorder can "see" into the Infra red range. This is useful to prove that
IR2 and IR3 are producing output.
Below is a picture of my veroboard layout for the Mk 5 IR extender using Ron J's excellent
veroboard images. A tutorial on how to use the image is available in my practical section,
click here for the article
. Special thanks to Derek
Smith for checking the veroboard layout and pointing out one small error (which is corrected
I have omitted Diode D1 in my prototype and also the
veroboard layout above, and the two images below.
Click the links below to view the actual veroboard layouts. The veroboard drawing above shows
the component site, the yellow circles represent the breaks on the bottom (track side).
Component side (106k)
Track side (97k)
Note that this is reversed from component side.
For more help on vero layouts see this Practical Page
If your circuit does not work, first check that your circuit is receiving power. Next compare
the voltages to my prototype below. These checks are all made with a digital multimeter with
a supply voltage of 12V DC. All checks are made with respect to ground (i.e. the back or negative
meter probe is always connected to the negative or 0V power rail).
With no input signal:
IC1 Pin6 1.15V (Voltage proportional to ambient light, can be as high as 9V in daylight.)
IC2 Pin6 0V
Q1 base 0.8V
Q1 collector 0.13V
Q2 emitter 0V
With a strong input signal (handset same room less than 2 meters away):
IC1 Pin6 1.15V (Voltage proportional to ambient light, can be as high as 9V in daylight.)
IC2 Pin6 0.15V
Q1 base 0.65V
Q1 collector 3.16V
Q2 emitter 2.79V
A good tip from Derek Smith (UK, who had problems with poor noise immunity in this circuit.
Derek cured his fault by replacing the SFH2030 photo diode, the new SFH2030 provided much
better noise immunity. So, if your voltage levels are similar to my prototype above then
try replacing IR1.
If you still have problems with noise immunity check the supply voltage. Special thanks to Roch
who found out that his 12V power supply was actually running at 16V. After reducing the voltage
to 9V the problems disappeared for him. My original circuit ran happily from a 12V regulated supply.
If you build the mark 5 circuit please let me know the make and model of your remote
control. I will add it to the list of compatible handsets below:-
Kameleon One for all remote (URC-8060)
Logitech Universal Remote control (Harmony 600)
Maplins 6 way Audio/Video Switcher Hub order code L63AB
One for all remote
Panasonic DVD player model no N2OAHC000012
Pioneer DVD remote
RCA systemlink 8 A-V
Sanyo vhs remote
Sony Camcorder HDR-CX11
Sony RM1- V141A VTR/TV
Std Sky digi box handset
The above PCB layout was kindly designed for this project by Lubo Veselsky.
Modification for 12V Vehicle Use:
The modifications below were kindly suggested by Allan Popplewell.
Allan writes: "I placed a TS7812 voltage stabilizer 2 columns over from Q2, across rows 2,3,4 with the heatsink to the right looking down (painted purple and called IC3 on the attached picture). I then moved the power input connector to the second (previously unused) row. The circuit will now receive only 12V, even though the car supply ranges from 13-15V depending on the charge state of the battery.
The circuit seems to work equally well at 9V, so a TS7809 would probably do just as well. I also replaced R1 with a 2M7 resistor seems to greatly improve noise rejection, especially from fluorescent or low-energy bulbs.
Alternative Mark5 PCB
Below is a single sided PCB that I have made using kicad. The first (enlarged view) is the 3D PCB layout showing component outlines. Next is an
actual size view of the component side, the last black and white image is the copper side. Black areas are copper, white areas are where copper
has been removed.
Page Last Modified 23rd August 2011 ( Added alternate PCB layout)