Norcim rc electronics club page 22……
A 2.4 GIGAHERTZ TRANSMITTER WAVE CHECKER by David Caudrey
With the popularity of the relatively new 2.4 Gigahertz radio frequency band now being used by many modellers, David C began just a few weeks ago, to think about the possibility of checking these transmitters for radiated output before crashing a model.
The result came quickly but not without some possible ‘eyebrow raising’ of the hardened ‘electronic drinkers’ out there that will be reading this.
The usual approach with detecting microwave transmission normally involves a special detection diode specifically designed for the job. These are usually of the Schottky variety (1N5711, 1N6263 etc).
This approach was tried but the results were not inspiring.
With much experience of Germanium diodes and indeed ‘cats whisker’ devices, David could see little reason why these early diodes should be listed as ‘low frequency devices’. A decision was made to experiment and the results were so encouraging that the 2.4 GIG WAVE CHECKER was developed.
The device is is very sensitive and easily shows the differences between different makes of transmitters. Another interesting finding was the conclusion that with 2.4 Gigahertz transmitters, the maximum radiation is emitted from the end of the stick aerial. So for maximum range perhaps we should be pointing the aerial at the model ! (more eybrows raised). However it is early days for conclusions.
Just to give an idea of the sensitivity of the prototype wave checker, domestic microwave ovens use the same radio frequency of our R/C transmitters. Microwave ovens have to be have to be leak proof for obvious reasons. This device can pick up microwave oven leakage several feet away from the oven! With normal microwave oven detectors, you have to be touching the oven with the detector around the door joint when closed. Normally nothing will show up using these mass produced safety detectors.
A further eyebrow raiser for this 2.4 Gigahertz Wave Checker design…. It responds strongly to the 5.08 Gigahertz video feedback transmission of drone television cameras. That is also without the easy adjustment of the tuned receiving loop antenna. (two connector screws and a different diameter loop can be used).
It would seem like the ultra high frequency detection capabilities of early Germanium Point Contact Diodes has not been appreciated. They are still listed as ‘low frequency devices’. Production of the OA91 and similar devices has long since ceased in favour of silicon diode technology. Perhaps something has been missed along the way ?
The Wave Checker circuit comes next with notes >>>>>>>>
The germanium detector diode OA91 is listed as a ‘low frequency diode’ and may be difficult to source. Be careful with substituting alternatives. We think that some Maplin stores UK have stock of the OA91.
A Mk2 version of the Wave Checker was disappointing. But provides food for further experiment.
Attached is a photo. of the latest version of the Wave Checker with a helical aerial fitted in lieu of the simple loop. The object of the exercise was to make the checker response more definitely directional in order to attempt to plot the polar response of the in-built printed aerial of my 2.4GHz. Futaba T6K transmitter The dimensions for the 2.4 GHz. helix were obtained via an RSGB. related website so I approached the modification with a fair amount of confidence. However, to date the experiment has been a failure because the helical aerial has resulted in marked loss in sensitivity at 2.4GHz; making the plotting of a response impossible. The modified Wave Checker responds weakly to my Futaba 6EX 2.4 GHz transmitter which is not to the latest EU restricted output level specification but it hardly responds to the ‘pathetic’ output from the newer Futaba T6K transmitter. Paradoxically sensitivity at 35 MHz is excellent; with full scale indication being obtained ten feet from a Futaba T4YF transmitter and even further from a Fleet XP-FM transmitter; making it ideal for plotting a polar response if one wanted to do so. With the original loop aerial the sensitivity to 2.4 GHz (6EX) and 35 MHz. transmitters was similar but it was unsuited to plotting a polar response and at least the T6K output would register full scale with the loop close to the transmitter case. Personally I suspect that the dimensions of the helical aerial might not be ideal although the response to the leakage from my microwave oven is spectacular! (E.g. strongly detects leakage in a narrow beam 8 feet or more from the oven.) D.C. Oct 2016.
A MICROAMMETER PROJECT BY David Caudrey
The Wave Checker circuit above showed that a simple germanium diode listed in the ‘low frequency’ technical specifications, can actually be used for microwave detection up to 5 Gigahertz (and possibly beyond) . the Wave Detector project triggered the desire to look more closely at the characteristics of various diodes for comparison. Particular interest was what happens during extremely low current conditions (micro-amps and nano-amps).
The following project was developed in order to measure diode characteristics at extremely low R/F
detecting levels. >>>>>>>>
THE MICROAMMETER PROTOTYPE IS SHOWN IN THE NEXT PICTURE >>>
Some initial results of various diodes using the microammeter are shown below. The results show that the OA91 point contact germanium diode can detect very low microwave activity compared with the other diodes which were to hand. The internal construction of these early point contact germanium diodes probably helps with these initial findings as the internal capacitance is almost non-existent allowing the diode to still behave as a rectifier even at microwave frequencies. The following static results show that the OA91 needs very little input voltage before forward current begins to flow and rectification begins to happen. It is at least ten times better for small signal detection than the other diodes tried. Unfortunately germanium diodes such as the OA91 were replaced many years ago with silicon diode production which had superior temperature stability.
However it does seem that there is recent growing manufacturing activity of germanium diodes. The new 1N34A device is now available in quantity for as little as 5p each. Whether the microwave characteristic of this new version is as good as the old OA91…we still have to find out. Note that the 1N5711 is a typical schottky diode normally used in microwave applications. The test results suggest that the old OA91 was around ten times better at detecting very low levels of RF. The stunning findings of the 2.4 Gigahertz Wave Checker project showed that these early technology germanium diodes still work at up to 5 Gigahertz.
A MORE VERSATILE VERSION OF THE MICROAMMETER is shown below. This version is also capable of measuring AC Millivolts. The extra circuitry is shown alongside. Selection of Microammmeter or AC Millivoltmeter measurement is done with the toggle switch at the bottom of the case.
Initial testing using the AC Millivolt selection suggests accurate readings with input frequencies below UK mains frequencies (50cps) up to around 25MHz! Interestingly the millivoltmeter still responds to frequencies well over the 25MHz suggested upper limit but its accuracy is not clear. Perhaps for frequencies over 25MHz, the readings may be useful for comparison only.
AND NOW A CHANGE OF SUBJECT from Paul Luby…..Another very valued Micron customer of yesteryear. Paul however is fortunately still very much into DIY R/C radio but with totally up to date technology (he is a clever guy!). His vintage transmitter restorations employ the latest technology using computer technology encoders (Arduino) and FrSky 2.4 Gigahertz transmit sections. These conversions give excellent range for aircraft use and any other use. The radio-link security is better than all of the previously dedicated model aircraft frequency bands.
I started building radio gear from Terry's PL7 Transmitter kits many years ago.
I had many bits sent to RAF Germany and few chats with Terry by phone in the late 80's.
But I digress.
My electronics and model passion has not died over the years and I now acquire ol transmitters and update them using Ardiuno boards and FrSky D8 2.4GHz RF "Hack" modules.
I've attached a few pictures but if you want more information please contact me.
Also check out http://www.modelflying.co.uk/forums/postings.asp?th=94809&p=1
Many thanks for a great site for us electronic junkies.
Wow, and this is just a starter of all of his conversions !
An improved analogue servomotor for radio control
By Jean-Marie Piednoir
An analogue servo design that equals or exceeds the precision of normal ‘digital’ servos used in R/C systems.
Operation principle and schematic:
Op-amp IC1-1 generates symmetrical voltages of +/- 1,2V for the servo feedback potentiometer.
This ensures elimination of neutral drift if the +/- 2,4V battery depletes asymmetrically.
Referring to the schematic, when the voltage at B reaches +0,62V. Transistor T1 begins to drive T3 into conduction. R4 reinforces the conduction of T1 and T3 saturates rapidly.
T3 will be turned off only after the voltage at B has come down to approximately +0,15V (depending on the value of R4).
The same sequence happens with T2 and T4 with thresholds at -0,62V and -0,15V.
Thus, we have a double Smith trigger with a deadband of +/-0,62V.
Let's assume that the servo is in a balanced state. If the input voltage at « A » is raised slowly by 3mV, the voltage at « B » will decrease by 0,62V and T4 will be driven into sturation.
The feedback resistor R5 will cause a positive-going voltage ramp at « B » as C2 is charged. After approximately 3mS (depends on the values of R5 and C2), the voltage at « B » will have risen up to -0,15V and T4 will turn off.
If at this time the servo hasn't moved, C2 will discharged again towards -0,62V with a slope depending on the voltage at « A » and the cycle repeats itself.
The motor is fed current pulses of a minimum width controlled by the values of C2 and R5 as soon as the edge of the deadband is reached either side. The deadband depends on the ratio R2/R1.
The on/off ratio of the pulses increases as the difference between the actual and desired servo positions up to continuous conduction when the error is large enough (around 40 mV in the actual example).
R5 also sets the damping of the system. This why R5 and C2 must be altered together to maintain sufficiently wide minimum pulses (3mS or more).
The attached schematic is for a supply of + / – 2,4V with an ordinary servo mechanics.
Operation is particularly neat and the signal across the motor terminals is amazingly clean.
This circuit brings the operation of an analogue servo on par with digital servos by:
- Feeding the motor with pulses of a minimum width which can be adjusted by component values, as soon as a position change is called for.
- Operating the output transistors in full saturated mode at all times.
- A very low drain when the servo is not moving, of the order of less than 5 mA.
- Eliminating the « soft » centring of the typical analogue servo amplifier.
Deadband for the example: 3,5 mV for an input range of +/-0,40V, i.e. 3,5/800 = 0,43% or equivalent to 232 steps.
Jean-Marie Piednoir 2015
Some further constructional information follows…….
Jean-Marie Piednoir continues with his superb Retro Galloping Ghost model using
60 year old model radio control technology……
After the early days of R/C using valves and sequential left right ruder control, came the first simple proportional system using a small spring centred electric motor to wobble backwards and forwards (around 10cps) via an on/off pulser built into the transmitter. The mark space of the pulser causing the motor and rudder to proportionately 'shiver' proportionally, to the left or right. A pot on the transmitter with an external knob controlled the pulser and hence the direction of the plane. Proportional control was born! It offered the first really smooth control of the direction of the model plane.
It didn't take long before some bright spark (Don Brown US. in 1954) to realize that the same piano wire crank that was shivering the rudder left and right via the on/off mark space of the Tx....could also be used to 'kick' the elevator upwards if the pulsing was slowed down! Looking at the picture and the linkage, Don's idea was to slow down the pulsing of the transmitter so that the crank would swing more in each direction, in doing so it would kick the elevator upwards. Indeed slowing the pulse rate down to around 2 cycle per second would allow the crank to rest the elevator in an 'UP' position for much of the time. (There were mechanical stops on the electric motor gearing to prevent total rotation of the crank). Don then made the transmitter pulse rate variable from 2cps through to 20cps giving proportional control of the elevator using a joystick. With careful setting up of the model linkage, rudder control was still available and very effective.
This early dual proportional system was nicknamed 'Galloping Ghost' after the strange sound of the electric motor and linkage made when the system was working. All early systems used a ‘Mighty Midget’ motor driving a torque rod (piano wire or balsa) to the tailplane, called a 'bird cage' for some reason, (just like the picture). The idea was to proportionally control both the elevator and rudder of a model plane. After flying many of my GG models I can honestly say it worked extremely well! Most of the problems were simply caused by reliability of the home made mechanical components used both in the transmitter which had to have a home-made joystick and the reliability of the models home-made delicate linkage system.
So varying the mark/space ratio of the transmitter pulser (25/75%) still gave proportional control of rudder and varying the pulse rate of the transmitter (2 to 20cps) gave the elevator control. For some reason arranging the CG of the model further forward (25% from wing leading edge) helped with manoeuvrability and dead stick landings (there was no throttle control in those days). Thanks go to Don Brown for bringing our radio control models into the proportional age.
Jean-Marie’s Retro GG model was designed by one of the early pioneers of the ‘R/C proportional period’, C. R. Ralph. It was one of the most popular designs specifically for Galloping Ghost systems and performed superbly using the early Galloping Ghost dual proportional system. His all balsa construction with doped tissue covering plan is shown below. The balsa construction in those days was relatively easy for the modellers of the day and the model could be built in a week of evenings.
Hello Terry Here is a picture of my C.R. Ralph "Galloper" model. Power is a Webra 1.5cc diesel motor.
It is covered with printer's laminating film which is cheap, easy to apply and fuelproof.
Only the fuselage is covered with silk and the nose painted.
Also pictures of the actuator, which I made by converting a small geared electric motor with 6 to 1 reduction (German Michael Gross "Pusteblume")
Originally for 7 to 12V use, but still quite nippy on 2.4V. The battery is a center-tapped 4.8V, 700 mAh NiCad.
I had to put a 1.8 Ohm resistor in series to tame it and still had to use fairly fast rates of 5 to 20 per second.
The GG actuator is shown in the right hand picture and the motor with single gear looks an excellent replacement for the ‘Mighty Midget’ motor of yesteryear. It appears to be a mini electric flight motor. The piano wire seen from the torque rod fixing nut provides the fixing point for the light centring spring necessary for the actuator. It also helps with the alignment of the torque rod with the output gearwheel. The universal joint came from a model boat that is no longer used. A balsa wood torque rod is used from the actuator down to the control surfaces. This provides more rigidity than a piano wire rod often used. The mechanical stops to prevent rotation of the crank with slow pulse rates are at the control surface end of rod which reduces the mechanical shock to the motor and gears with the balsa absorbing the shock without producing a ‘bounce effect’ that a piano wire rod would introduce.
The picture to the left shows the simplicity of the anchor point of the centring spring of the output wheel. A very light spring is used which is necessary for minimising the actuator current consumption. The low 2.4volts motor drive also helps this.
SOME THOUGHTS NOW ABOUT CIRCUITRY FOR A POSSIBLE RETRO GALLOPING GHOST PULSER CIRCUIT….
The following circuit has never been tried and remains an idea for possible retro development. Generally available R/C joysticks are based on 5K control pots. Often these are mechanically adapted to suit the plastic moulding of the stick…..it follows that a change of pot value to suit an electronic circuit is not practical.
This circuit could produce an 80/20% mark space ratio with around 6 to 1 rate variation required for GG.
Must be the most simple circuit using just six external components. Pots would have to be set close to the POS end of the track and CAPs selected to suit the actuator used. Note that the ‘5K POT’s are the ones on the joystick.
David C checked the circuit using LTspice simulation software to see if this simple circuit could possibly come up with the goods. The main fear being that the circuit may not oscillate with the control pot wipers so close to the positive end of the track, however the simulation appears to dispel this. With the circuit values shown to the left, the GG circuit shows the quite promising results above. The pulsing of the transmitter (tone /carrier) would have suited the popular ‘Rand’ Galloping Ghost actuator of yesteryear with plenty of mark/space for rudder and a 7:1 rate variation for elevator control. Note that R3 and R4 simulate the maximum positions of the joystick wipers.
For the circuit to work, both control pots must move together for elevator control and in opposite direction for rudder control. The sketch above shows how this could be done. Spring centring and in-flight trim would have to be added. Fortunately a standard joystick from a defunct transmitter could be used if mounted in the front of a potential GG transmitter case but rotated 45 degrees clockwise.
A PICTURE NOW OF THE FIRST UK’s GALLOPING GHOST TRANSMITTER available in kit form for modellers in 1964 ! The case was folded aluminium (by a small firm in Sheffield) and available in several car cellulose spray colours. (to order, by hand!) The joystick parts were produced by an out of work engineer in his back garden shed in Chesterfield in order quantities of one dozen per order. Printed circuit boards were hand produced, etched and drilled. The kits also included a receiver kit of parts, (which used a plastic box from a chest of component draws!). Also available were folded aluminium parts and spring bits with a ‘Mighty Midget’ electric motor for assembly of the actuator (servo?). Mail order using the model magazines of that time under the name of ‘Aercon Developments’ was used. A year later 1965 the same GG radio control system was published as a DIY article pull-out booklet in ‘Radio Control Models and Electronics’ magazine, September issue.
Later the same company became ‘Micron Radio Control’ and produced not only the majority of Radio Control Electronic Kits for the UK but also for export to dozens of countries around the world during the 1970s.
(Further Galloping Ghost info available on Page4)
A change of subject…………
Back to the Future! …. Just a taste of one of JR’s new transmitters. It’s Retro! So many memories. I want one !
JR's COLT takes RC nostalgia to new levels of sophistication, reliability, and ease.
Housed in a vintage looking all aluminium case just like what we all flew back in the old days, the COLT might look like your basic vintage transmitter, but of course, its looks ARE deceiving.
Just lift the lower cover plate, and you will find a fully capable, 20 model memory, 2.4GHz DMSS fully equipped 6 channel transmitter that will allow you to pilot your prized vintage models with a level of programmability and reliability that was only dreamed of back in the day.
Hand Built, and in limited production, the COLT's 3 Model Type capability makes it capable of flying your entire fleet of vintage models.
No more searching for an old system and crossing your fingers that it's still reliable. No more looking for someone to take your new system mechanicals and install it in your vintage case. JR's COLT DMSS system is Retro, Reliable, and Ready to Go.
THANKS FOR READING so far… ! More to be added as we go along.