Norcim rc electronics club page 4…….


flight line3 





 SOME USEFUL LINKS    image001 HobbyCity        image001 GreatHobbies         image001 BangGood        image001 ModelFlight RC       image001 HorizonHobby




ECONOMY ELECTRIC FLIGHT was outlined in Radio1 and Radio2 pages.


Well the picture shows one of the test models, a Biene2Graupner Biene’, using the ‘Sun’ motor, on a typical climb to thermal height, and it will do this twice on seven GPâ AA nickel hydride cells from CPC at just £1.00 per cell. With the motor costing just £2.30 from JPR electronics, this shows just how inexpensive good electric flight can be. The AA cells give around ten minutes of full power in the air using a Perkins Distribution (most model shops) 8x5.5 electric flight prop. Increased climb rate can be had, by using a Graupner 9x5 (most model shops) electric flight prop, but this reduces ‘full power’ in the air to around 8 minutes.  Increased performance can be had with a Groaupner speed 600 motor and a standard 6 cell sub C battery pack. (see page 2)

The following circuit shows a simple speed controller. Many of these controllers have now been used with test models using Micron receivers and all have Image1bienne 50performed well. The few components of the controller allows home construction for less than £3.00! The controller is based around a CD4001N quad 2-input NOR gate IC which can cost as little as 30p from your electronics store! The real expensive bit is the MOSFET output transistor which could lay you back around £2.50. The single MOSFET drives the SUN motor (including standard 380 and 540 motors) using seven cells (8.4volts) with no problem. You can even cut the metal tag off the transistor to make the controller more compact, without overheating of the MOSFET.

image011The input circuitry produces a saw-tooth waveform with each servo pulse. This can be adjusted via trimmer R5 to trigger IC1A gate to produce a 50/50 mark space output thus driving the motor at half speed. Changes in the pulse width cause the saw-tooth waveform level to change giving variable speed from fully ON through variable to fully OFF condition. The circuit without the optional 5v regulator is intended for use with a separate receiver battery. If the LM2940 is fitted then the receiver, servos and controller are supplied by the flight battery.

This simple circuit is more suited to receivers with full rail voltage output pulses (usually Cmos decoder chips) and has been tested using  Micron, Fleet and RCM+E receivers.

Using the three remaining gates of the CD4001BCN in parallel gives adequate drive for up to five MOSFETs in parallel for ‘hot’ motors. (at the relatively low switching speeds!).

D2 and C5 provide good suppression of motor interference for the Sun motor, which has no internal brush suppression capacitors. For ‘Hot’ genuine electric flight motors just follow the suppression instructions supplied with the motor.





 The following circuit offers manual control of almost any make of R/C servo, using just a receiver battery  (4.8 nicad) for power. The components should cost you no more than a £1.50 or just over 2 Euros! and it should work with all makes of servo. The circuit is a text-book ‘two transistor multivibrator’ providing an 18 to 20ms cycle time which suits R/C servos. The circuit is however extremely lob-sided! In that one side produces around 17ms delay before triggering the other side, which only stays on for 1.5ms! It’s this side that the servo output is taken from. The 4k7 pot with knob, produces a pulse variation from 1ms to 2ms at the output which can be used to test most known servos. C3 provides sufficient damping of servo noise to let the circuit work correctly. It’s worth mentioning that the hfe of the transistor used must be high, better than 250 @ 1ma. So if substituting, then pick a high gain device.








image016There are several methods of fast charging model car and ‘electric flight‘ batteries but the most popular system, is called the ‘Voltage Peak Detect Method’. When NC or NH batteries are charged, their voltage slowly rises to a peak (when fully charged) but this is quickly followed by a small but significant dip in the battery voltage. ‘Peak detect’ fast chargers, rely on their ability to detect this small voltage drop when the nicad is full and terminate the charge before overcharge damage is done. A simple ‘peak detect’ method for six or seven cell sub-C batteries is shown. At switch-on, (connection to the 12v supply), the forward volts of D1 holds the inputs of the OP AMP apart, preventing any switching of the output. Initially, the voltage across D1 (which is a Schottky RF diode) can be as high as 0.3v but this quickly reduces within minutes to less than 0.05 volts because of the extremely low current flow through the diode. C1 then, follows the increasing nicad voltage (which can be as close as 0.02 volts when near to full charge). At full charge, the nicad voltage begins to fall…..D1 soon becomes reversed, (also IC1A inputs!) causing IC1A output to ‘flip’. This terminates the charge current. The result is a fast, half to one hour, automatic cut-off charge! There is no need to discharge the nicad first. Full charge is detected even with a partly charged nicad. It is important that C1 is one of the recent Aluminium Electrolytic low leakage type capacitors. (Rubicon TWL series 47UF16V-16TWL47MO811 Farnell order code 499-067). It’s worth mentioning that the 1N6263 has a lower reverse leakage current than general purpose schottky diodes. This helps at the peak detect point.




image018This next circuit lends itself not only for home checking but also club and quick model shop checks. The circuit checks for correct power output of any 35 or 40 MHz radio control transmitter is shown. These things are called ‘field strength meters’ and are a standard piece of electronic equipment in the service workshop to check the output power of R/C transmitters. ‘Field Strength Meters’ (as they are called) are usually based around a reasonable size sensitive 50uA moving coil panel meter. These are now listed (Farnell) between £20 and £30 each (before circuitry!). This circuit is based around the National Semiconductor LM661CN Cmos quad op-amp IC. The circuitry components should cost no more than £4.00! and it has greater sensitivity than the standard meter type. Transmitter output strength is shown by four Superbright red light emitting diodes. A correctly functioning R/C transmitter, will illuminate three to four LEDs at a distance of 10 metres away. Adjusting the length of the short telescopic aerial will allow all LEDs to operate at a shorter distance for indoor checking. With occasional use, a four AA alkaline battery lasts over a year norcim website 001A(even occasionally leaving the thing switched on)

The OA47 diode seems to work best but more difficult to get. L1 needs to be initially adjusted to illuminate the maximum number of LEDs at a range of 10 metres or so. Once set that’s it. The Toko coil used is no longer manufactured but many are still in the pipeline and there are alternatives. Remember, if you set L1 using a 35MHz Tx then the unit will only check other 35MHz transmitters. If 40MHz Txs are to be checked, set L1 using a 40MHz Tx. L1/C1 form a tuned circuit at 35MHz. A 35MHz Tx will excite this coil and cause a resonance of L1. D1 detects this and a little current flows at 35 million times a second! into C2. This increases the voltage across C2 (slightly) in proportion to the power of the transmitter signal. The LMC660CN is a Cmos op-amp and has little effect on the input circuit. The op-amps are arranged as voltage comparators using the potential divider R1-R5. The resistor values are selected to give a 3dB step between op-amps flipping on. (each one showing twice the transmitter power output) So with a weak signal, IC1D output will illuminate LED4. As the received signal gets stronger, the remaining LEDs will illuminate in turn, until all four are illuminated.




The above Field Strength Meter based on the LMC660 rapidly runs out of Puff after the 40MHz band. This leaves the 72MHz and the now popular 2.4Gigahertz bands not detectable. DAVE1993 of has done some work producing a really interesting Broad-Band Field Strength Meter. This detects from 27MHz to around 5 GIGAHERTZ !  Dave’s posting in the website is  His posting links to other threads that started and compliment the discussion.

FIELD STREGTH METERS are essentially comparative devices. They simply compare ‘like with like’. The above LMC660 FSM compares 35 and 40MHz transmitter output at a distance of 10 metres away. The DAVE1993 transmitter Field Strength Meter compares transmitter output at a much closer range (as shown in the picture). However the result of testing is the same. If a known ‘good’ Tx gives a reading of ‘X X . XX’ when positioned as shown in the picture….then a ‘suspect’ transmitter should give a similar reading when positioned in the same place. The following pictures are courtesy of the man himself…DAVE1993. Note that there is much more detail using the link above.


rf_meter02Dave RF meter3  Dave RF meter1 Dave RF meter2


The possibility of using DAVE1993’s circuit in place of the TOKO coil version in the LMC660 circuit above,

 does come to mind !




image020Fig 1 shows a possible output stage for charging a 7v2. nicad. Note that this is not a ‘constant current’ circuit but should work OK. Note that the 1R resistor will initially let through a charge current of around 3 Amps which will slowly fall to 1 Amp or less at peak detection. The MOSFET in this version would not need to be fastened to a heat sink as it is in a fully ‘on’ condition when charging and will need to dissipate little heat. The 1R or 2R resistor however will get very hot and would benefit from a heat sink. The 1R resistor value should be used for sub-C nicads. The 2R2 resistor is a better value with less current flow for 7v2 AA nicads. Fig 2 shows another possible output circuit in which the charge current through the MOSFET (Q2) can be set by adjustment of the 22K trimpot. A charge rate of 2 Amps is suggested for 7v2 nicad batteries. Some care will be required with the setting of the 22K pot and best to start with the wiper at the negative end of the pot.  The charge rate should be reasonably constant with this circuit, owing to the bias characteristics of MOSFETs. Note that this circuit uses the MOSFET (Q2) to restrict the charge current flow and therefore Q2 will get hot! AND will need to be clamped to a metal heat sink. IT MUST BE POINTED OUT THAT ALTHOUGH THE ‘PEAK VOLTAGE DETECTION CIRCUIT’ HAS BEEN WELL TRIED USED TO DATE, THE TWO ABOVE CIRCUITS FIG1 & FIG2 ARE ‘POSSIBLE OUTPUT’ CIRCUITS AND ARE INCLUDED FOR EXPERIMENT.





Fig 3 shows a simple circuit that can be plugged into one of the receiver servo outputs to switch on and off an external load of up to 15 Amps. The circuit is shown driving an electric flight motor but for boat people the relay could switch several torch bulbs, or a high power halogen lamp, or winch, or sound circuit. The dotted line shown from Power switch3athe relay back contact to ground provides dynamic braking of the motor when using folding props with electric flight. The input components form a text-book ‘diode pump’ circuit. The trim-pot is set so that thin pulses from the transmitter pump up C1 voltage to just get the Logic Level MOSFET, conducting. The current flow through the relay however, is very small and well below that necessary to close the relay contacts. But throwing the stick at the transmitter to get fat pulses, increases the voltage across C1, allowing much more current flow through the MOSFET and the relay closes. Both 5 volt and 6 volt relays have been used in this circuit with windings of 55R to 90R. Notice that only the Rx pulse input and the negative input is required from the receiver (The red positive input is simply left unconnected). The current drain of the circuit on the receiver battery is almost undetectable. At higher voltages of the external circuit, the MOSFET (Q1) tends to limit the current flow through the relay to acceptable levels. The MOSFET, RFD4N06L (or similar) is a TO 251 (I-Pak) device and may get warm with a 12 volt external supply. This is normal. Several relays are suitable…finder type SPCO 16A Farnell code 431-308…or smaller 12Amp version finder type Low profile 21 series Farnell code 321-0224. Smaller low profile relays are now available and should be suitable.




EVER LOST A MODEL IN A CORN FIELD??? Well this next 555 circuit will detect its position!


image024PIEZO SOUNDER DEVICES ARE SMALL AND WEIGH JUST 5gms! These small units can be heard up to 100 metres away in quiet countryside conditions, particularly if they are switched on and off! The following circuit uses a 30p 555 timer IC, to drive one of these Piezo sounders. The 555 is configured as a 50/50 astable pulse generator which sounds the piezo thing around once a second. Transistor TR1 keeps shorting C2 with each servo pulse, so that the 555 IC can’t get going. The piezo sounder remains quiet during normal flight conditions. Should the model decide to land in a corn or rapeseed field, then switching off the Tx allows TR1 to go open circuit and C2 to begin its charge/discharge cycle enabling the 555 to switch the piezo sounder on and off at around once per second. With careful listening, the model can be located! Suitable piezo sounders are Farnell order code 927-181 or 927-119. The receiver input stage should suit most makes of receiver including those with low voltage servo pulse output.





Did you know that one of the first offerings from the ‘Micron people’ was a pulse proportional system?  Based around a ‘Mighty Midget’ electric motor? Full kits were available from around 1962! The system used a single electric motor to pulse both rudder and elevator control surfaces. Surprising proportional flight could be achieved with this system, which was termed ‘Galloping Ghost’ (because of the noise of the flapping control surfaces!). One of the flight demonstrations of the system, involved several passes through soccer goalposts (at a medium throttle setting, as there was no throttle control), which was quite an achievement alongside the standard ‘reed’ equipment of the day. The original Gallatrol transmitter circuit is shown in page9.

STICK GG JPEGAAThe receiver (shown bottom left) used a super-regen front end (only allowing one model to fly at a time!) was designed by early R/C pioneer, Doug Bolton. Other R/C equipment of the day used self-centre toggle switches on the front of the transmitter for full throw of the control surfaces Joystick units had not been invented! So joysticks for the ‘Micron people’ transmitter kits had to be individually made from brass stock with piano wire wound springs! Two external pots provided the in-flight trims. Early transmitter circuitry used a variable rate saw-tooth generator (unijunction) followed by a varying level Schmitt trigger to generate the mark/space and pulse/rate, pulses for the ‘mighty midget’ actuator. Later circuitry used a curious diode gated two-transistor astable-multivivrator, which gave a logarithmic effect to the rate channel image031(elevator). This improved the elevator effect when using the ‘mighty midget’ actuator. Unfortunately this effect could not be used with the later more popular American Rand actuator.


Rapidly advancing multi-channel analogue and digital technology saw the demise of this kind of simple pulse proportional control but there is little doubt that these early systems began the trend into true proportional control of model aircraft.


The photos show one of the early Micron People ‘Galloping Ghost’ proportional systems, which is now fifty years old! When last tested with replaced nicad batteries…….It still worked!….and without any doubt, it would still fly a model through those soccer goalposts! Unfortunately my grey matter has slowed down somewhat and there is no way now of proving this!


A photo of one of the models that flew with this system is shown and probably dates around 1962 ish. Although looking good, it was found that the pulsing of the radio system was clearly seen whenGG model  flying. Other models used at the time were the ‘Gasser’ and the Charles Rail design called ‘Galloper’ both were plans from the Aeromodeller magazine. The conclusion was that the greater length of the fuselage from the wing to the tail actually helped to damp out any visual ‘hopping’ effect of the elevator. The use of a heavy motor, prop and motor mounting plate also helped. This produced a situation of having to add lead to the tail of the model to get the correct CG. This extra weight at the tail end dampened out the pulsing elevator effect. The pulsing rudder never seemed a problem and under all flying conditions could not be noticed when flying.

Memories of early galloping ghost models that I flew, showed that they were extremely good at slow flying during the glide after the motor had cut. In fact I remember that with one model, I could often bring round into wind with the intention of landing several yards in front of me, but Full Up elevator at that point produced a slow fully controlled rise into my chest where the model was ‘caught’. I never achieved this with later models fitted with proportional gear. One reason could be that GG models were built of balsa and tissue covered (they were light!). On reflection an additional reason for this uncanny control could be the fact that at extremely low speeds, the wing was continually stalling and uninstalling with the ‘hopping’ effect of the elevator at full UP of the system. (generally the elevator is kicked UP around twice a second with full UP joystick and hopping is noticeable). The GG actuator (or servo) designs of the time mostly used a Mighty Midget electric motor with its nylon top gear. This allowed light spring centring. The drawing shows a typical modified version by Terry Tippett. The centring used a spring similar to that seen on joysticks today but instead of stretching the spring, it was simply twisted clock and anti-clockwise of its natural un loaded state. The nylon gearwheel had a 6BA nut bolt fitted at the 12 o’clock position with a short bit of fuel tube which stopped rotation of the motor in either direction by resting on the motor body. At slow pulse rates this brief ‘resting time’ on the motor body lengthened the time that the elevator stayed in the ‘UP’ position. This allowed powered loops with the system.




The hole at the back of the fuselage was purposely made large to aid the final adjustments of the elevator angles. When correct the brass tube was positioned with balsa inserts and cemented in place. This now historic radio control system flew alongside ‘Reed Systems’ of the time. On many occasions the ‘Gallatrol’ system would be demonstrated flying backwards and forwards through Soccer goalposts! The following drawings show the typical oscillation of the 6BA nut bolt fitted at the 12 o’clock position of the nylon gearwheel, alongside transmitter joystick positions. The bolt had fuel tube pushed on and acted as a rotation ‘stop’ when reaching the motor body during FULL ‘UP’ condition. (see bottom left drg).






Interested?......then here are some links to the historic Galloping Ghost system with even a video !




In this system the pulse rate from the transmitter was usually fixed around 10 cycles per second of a ‘Tone’ followed by an equal time of carrier (27MHz).The tone was around 1000 cycles per second. (a kind of Low whistle!) This ‘equal’ Tone/Carrier would GG Driver2oscillate the motor actuator (and the rudder) around 10 degrees about centre position (centring was caused by a light spring). Left/Right Movement of the single joystick at the transmitter, altered the length of the Tone compared with the carrier. This was called proportional mark/space transmission. The motor actuator responded by pulsing proportionally left or right.

A DOWNSIDE of this simple system was Battery drain. Not only was the electric motor continually running but its direction was changing ten times a second ! The current required to reverse an electric motor that is going in one direction is many times its normal running current. A 500mah DEAC rechargeable battery of the day would only last a couple of flights. A typical receiver output circuit of the day is show left. (note without the capacitor C1) During the writing of this history, DAVID and I began to think what would happen if a capacitor (C1) was added to the circuit. By careful value selection, a saw-tooth waveform would be generated at the input to the Motor amplifier. With 50/50% mark space, ZERO current would result at the motor as the drive transistors need at least half a volt to activate . The rudder would simply centre itself with the centring spring. As the mark/space was slowly altered, the motor drive circuit would begin to drive in one direction only giving a proportional drive against the centring spring. So motor current would only take place (and at a relatively very low level) by moving the transmitter joystick left or right. If this was done for mild corrections of flight only………the receiver battery would last for up to twenty flights.!!


A CIRCUIT COMES NEXT from David Caudrey which would produce the Mark/Space pulsing for the transmitter. This modulator works at 50 cycles per second so control to the motor would be proportional against its spring centring without any apparent pulsing. For the moment ignore the 10K ‘elevator’ pot as this will be covered later.

The 100K Rudder pot produces the variable Mark/Space. As a normal joystick pot only moves around 30 degrees each way of centre, this would produce only a small variation of Mark/Space. However a linear slider pot could be used or a rotary wheel on a normal 270 degree pot. The desired 80/20 Mark/Space variation would then be possible. It would also be possible to use a 500K pot in a Joystick unit suitably modified with conductive paint to reduce the wiper part of the track to 100K. The modulator circuit would suit a 35MHz FM transmitter output section coupled with a 35MHz receiver. It may also be possible to use the circuit to drive a  FrSky 2.4GIG DIY transmit module.


Modulator2             C1 votages


The circuit above left is designed for pulsing the transmitter RF stage.  The FrSky 2waveforms picture (right) refers to the voltage ripple across C1 in Fig 7a.  produced by different Mark/Space ratios arriving at the receiver.








The picture left, shows the FrSky Do It Yourself 2.4 Gigahertz transmit module. This unit can be used to convert Vintage R/C transmitters or home assembled Model Radio Control equipment to a  UK acceptable form for club and organized flying/boating events.



GG Elevator Ratemeter 





Regarding the 10K elevator pot in the pulse circuit above, it may be possible to use one of the two circuits (right) to convert the variable pulse rate into a varying voltage to drive a second Actuator (servo). This could be used for an elevator control.











(the following circuits relating to Analogue servos are for discussion only. They are not intended as construction projects)

Mark/Space pulsing at the transmitter began to get faster. This smoothed out the flapping effect of the actuators. It became apparent that the receiver output(s) were beginning to look like a simple moving voltage with some ripple on it. Move the joystick to the right and the Rx output would rise by more than a volt. Move the joystick to the left and the Rx output voltage would fall by more than a volt.

The first feedback servos were borne!

Surprisingly the ‘servo mechanics’ were very similar to modern day ‘digital’ versions. A small 20mm electric motor driving a set of gears to a rotary push rod output arm. Gear ratio was around 270:1 which gave lots of push and quick transit time over the 90 degrees of the output arm. The output arm shaft internally also had a bronze wiper attached which was in contact with a 270 degree potentiometer. Analogue servo analogue servo3Aamplifiers were very simple and often used only four transistors as in the possible circuit to the left. The output voltage of the receiver, which moved up and down with joystick movement was applied to the input of the servo amplifier. The servo amp would drive the electric motor until the feedback voltage from the pot was the same as the input voltage, resulting in motor drive switching off. The servo output arm faithfully followed the joystick position.


Analogue Servos were a product of the late 1950s. Transistors were predominately Germanium types. Perhaps present day Silicon transistors would Idris Francis Servo Ampimprove performance of this simple amplifier. A pioneer of analogue servos was Doug Bolton. He designed and used a servo with only three transistors ! These were the days of simplicity !


So What went wrong with analogue servos ?

The analogue servos themselves were simple and cost effective. The circuitry was well within the home build category at the time. The real problem was with the rest of the analogue R/C system. Generating up to four analogue servo channels at the transmitter with complete isolation from each other proved difficult. Invariably, one servo movement would ‘slightly’ effect one or more positions of the other servos. Producing any more than four servo channels would have been technically challenging. (I don’t recall an analogue R/C system with more than four servos).

In order to produce ‘smooth’ voltage outputs at the receiver, large value capacitors were used at the servo outputs in order to produce a level voltage. Unfortunately these smoothing capacitors also slowed down the inputs to the servos. This resulted in a fractional delay between moving the joystick at the transmitter and the servos moving.


Analogue Radio Control lived a short life (perhaps only two years) but this was an important life in the history of Radio Control Modelling.




The following circuit (right) may have helped to provide a more instantaneous servo response to Ripple Filter2transmitter joystick movement. The circuit would allow much lower values of smoothing capacitor at the receiver to be 2v4 supplyused. This would have minimised the joystick to servo response. Voltage ‘ripple’ would be present at the receiver outputs but lost before input to the servo amplifier via this in-line circuit.



The circuit left may also have helped giving an electronic 2.4 volt supply for the servo motor. These early servo amps required four wires. (the extra wire was from the nicad battery centre tap.) It became very fashionable to have only three wire servos.

This circuit would have produced three wire Analogue Servos.


The short life of Analogue Radio Control Systems was abruptly brought to an end by the inventiveness of two NASA space engineers. Doug Spreng and Don Mathes. They invented and developed the Digital R/C system which is now 50 years old and to this day is still used by many professional R/C manufacturers.

The Digital R/C system solved all of the problems of analogue proportional Radio Control. Servo channels were transmitted and received sequentially which completely isolated each servo control input. Up to nine servos could be used with digital systems. There was no time delay between joystick movement and servo movement. All of the problems of the Analogue Radio Control Systems had been solved by this new technology and manufacturers dropped Analogue Radio  Control Systems, like a ‘Hot Brick’ !

It is interesting that a small UK electronic kit supplier called Aercon Developments listed a servo kit suitable for Digital Radio Control systems in the early 1960s. It used SLM mechanical parts (UK) and after assembly, compared very well alongside commercial servos of the time. Visually there was no difference in the servo movement and operation except…… with the loss of transmitter signal… The servo self centred itself !  There was no schematic included with the kit instructions but looking at the PCB layout. It was quite definite that the amplifier was of the analogue type shown above.








This simple circuit charges ‘part-charged’ two-volt glow batteries in an hour or two, with auto slow-down to trickle charge. (The charge rate gradually reduces toward the end of charge, to minimise battery ‘gassing’). The power input is taken from the 12-volt field starter battery  (or car battery). Both input and output are polarity reverse ‘safe’ and the green LED will only illuminate with correct input and output connections. The Car Bulb LMP1 indicates charge, taking place and its illumination intensity shows the level of charge. Bright is fast charge, Dull is slow charge and No illumination is charge complete with ‘Trickle charge’. The glow battery can be left in the trickle charge condition indefinitely (for a convenient time to unplug!) as the trickle charge eventually drops to less than 25 milliamps. The LED US1 is used as a voltage source for the base of T1, so care must be taken with substitutes. The one used in the prototype was a ‘Kingbrightsuperbright 5mm green! LED, which gives a foreword voltage (Vf) of 2.3 volts (in this circuit). D1 adds another 0.75 volts to the base of T1. The voltage across US1 can easily be measured after connecting the input leads of the circuit. (red LEDs give a lower Vf so don’t use). All of the other components are non-critical. NOTE….if the glow battery is severely discharged, (to be avoided for long battery life!) it may take a minute or two for the green LED to illuminate. The ‘Hawker Cyclonâ’ glow-battery quotes 2000 charge cycles and 10 years life with sensible charging!








OK! Not directly associated with model electronics but as many modellers have a caravan, then this little circuit could be of interest. The prototype was assembled on Veroboard and inserted in place of the bulb in the awning light. It gives a bright looking light that can be left switched ON for up to six months! Using just the internal caravan battery. It gives the impression that the caravan is occupied which adds to security. The prototype used 5mm Ulta-bright yellow LEDs from Rapid Electronics at 15p each. They drop 1.85 volts each at 15 milliamps. The resistor is a 0.25 Watt type. There are now Ulta-bright white and blue LEDs which are being used for front lights of bicycles! Providing the volt drop is similar then these could be used or a resistor value change to compensate. Another idea was to place the device in a small plastic box with wire connector, to plug into a 12 volt interior socket in the caravan. The unit could then be placed near the front blinds or curtains to suggest there was an interior light on. Wherever you park your caravan, this device will help with unwanted callers! For Months!! Note that commercially available LED lamps are now appearing on the market with similar spec.



nutcase circuit 












Randal Munroe worked on robot technology at NASA’s Langley Research Labs. Randal has kindly let me display the circuit to the left.


Unfortunately, Randal went missing while working on the construction of his prototype. It is believed he was trying to connect the wiring at the bottom left of the circuit. It is thought that he ran out of wire.


Randal hopes that the publication of his circuit will inspire other electronic enthusiasts to help develop this new ground breaking technology.























Thanks for reading.