Ruhmkorff coil, Ignition coil, Tesla coil, Flyback converter, SEC exciter, Slayer exciter high voltage fun

@author Andre Adrian, DL1ADR
@date 17jan2015


Ruhmkorff coil or Tesla coil experiments are fun. High voltage sparks and Geissler tubes electrical glow discharge create some "Frankenstein laboratory" feeling. As teenager the author experimented with car ignition coils and electromagnetic interrupters. The television set with a vacuum picture tube uses an electronic device instead of a mechanical interrupter to produce high voltage. The active device in the flyback converter in the TV was first a vacuum tube, later a transistor. A very simple flyback converter design is the Slayer exciter. It was (probably) published by Dr. Ronald R. Stiffler, WA7RTQ. This simple circuit allows to build a tiny high voltage generator. The generator does produce alternate current in the radio frequency range of typical some hundreds kilohertz (some 100000 cycles per seconds).
Some people attribute "extraordinary" abilities to the simple exciter circuits of Dr. Stiffler and others. The author does not believe in "perpetual motion" engines, engines that have a efficiency above 100%. The "Panacea-BOCAF On-Line University" is a little cautious, too. At least they write: "SEC Exciters (Do Not) create energy".
Günter Wahl, the author of the german book "Experimente mit Tesla Energie" wrote wisely: "wobei der Fantasie des Lesers die Interpretation überlassen bleibt, ob die Energie-Übertragung mit Teslawellen, Skalarwellen oder einfach nur mit leitungsgebundener Hochfrequenz stattfindet [the interpretation is left to the reader's imagination, whether the energy transfer takes place with Tesla waves, scalar waves or just over wired high-frequency]".
The author wonders sometimes about the progress physics makes. Some ten years ago there was still strong believe in the "aether", a mystical substance that is the transport medium for electromagnetic waves. Maxwell and others did prove that electromagnetic waves can propagate through vacuum and do not need an "aether". Today physicists are talking about "black matter". Again some mystical substance that you can not experiment with, but is needed to make some theories work. Today we know that the "aether" theory was wrong, but not the practical work of the experimenters. The author does not know if the "black matter" theory will be proved wrong, too. But at all times the scientists were 99.99% sure that they know nearly everything and that nothing interesting was left in the still unknown. But they were proven wrong again and again. Electrical super conduction or superconductivity at liquid nitrogen temperature of 77 Kelvin (−195.79 °C) was though impossible by the physicists one generation ago. The same is true for the giant magneto-resistance effect that gives us today magnetic hard-disk storage in the Tera bytes range.
The author likes simple experiments that make a "wow" effect. The pop-pop boat with the pulsed jet engine is a good example. The one vacuum tube radio receiver is another.

Picture: Build your own spark generator. Günter Wahl; Experimente mit Tesla Energie; 2002; Franzis Verlag

Ruhmkorff coil high voltage generator

Let's start with the Ruhmkorff coil from Heinrich Daniel Rühmkorff. The battery G provides direct current (DC) voltage. The primary winding A and the magnetic core C work as an electro magnet. The iron armature E carries an opener contact. The spring D brings the iron armature back in default position if no magnetic field pulls it away. The interrupter E operates the electro magnet in an on-off manner. The spring D closes the contact E and current can flow into the magnet. As the magnetic field builds up, the magnetic core C pulls the iron armature (left part of E) and the electrical circuit is interrupted. The magnetic field collapses and the iron armature moves again to the original position. The cycle starts again. The capacitor F improves the life time of the interrupter. Every time the magnetic field collapses, a (relative) high voltage is induced into the primary winding. This induction voltage causes a little spark at E. The spark does remove material from the interrupter contact. The capacitor F has no electrical change at the moment E opens. The induction voltage does charge the capacitor. After contact E closes again, capacitor F gets discharged. The capacitor F operates as an EMI/RFI (electro magnetic interference/radio frequency interference) suppression capacitor. This capacitor influences the Ruhmkorff coil radio frequency and the peak secondary voltage at H. Less capacitance of F gives a higher secondary voltage but more EMI/RFI, too.
Because the magnetic field changes between zero and some maximal value, a voltage is induced into the secondary winding H. The secondary winding has much more turns than the primary winding. The Ruhmkorff coil works as a step-up transformer.

Picture: Ruhmkorff coil (from Wikipedia article)

Ignition coil high voltage generator

The author used an ignition coil and a relay to build a Ruhmkorff coil replacement. An ignition coil transforms 12 volts DC from the car battery into 5000 to 30000 volts AC at the spark plug. Rule of thumb is every 1000 volts gives you 1mm spark length. The relay coil was wired in series to the opener contact of the relay. Like in the Ruhmkorff coil the electrical current creates a magnetic field which moves the iron armature and opens the electrical circuit. The magnetic field collapses, the iron armature moves back and the cycle starts again. The primary winding of the ignition coil is parallel to the relay coil. The relay K1 replaces the interrupter E. The relay winding K1 with connectors A1, A2 is parallel to the ignition coil winding transformer T1 with connectors 1, 2. The relay opener contact 12 is connected to the plus pole of the power supply, the opener contact 11 is connected to the relay coil and the ignition coil. The EMI/RFI suppression capacitor is C1. The author used the CAD program EAGLE to draw the schematics.

Picture: Ruhmkorff coil replacement

Picture: sparking ignition coil. Click on picture for Youtube video.

Tesla coil high voltage generator

The Tesla coil high voltage generator is nothing special. It is an ignition coil design. To improve the transfer of electrical power from primary coil to secondary coil both coils are made resonant to each other. Both coils oscillate at the same frequency. The secondary coil of a Tesla (resonance) transformer has often a "top" capacitor in the shape of a sphere or a torus. The inductance and the capacitor form a resonance circuit. The primary coil has the EMI/RFI suppression capacitor to complete a resonance circuit. Without explicit capacities at primary and secondary coils there are still the parasitic capacitances of the coils themselves. Between two adjacent wires in the winding there is a little voltage difference because of the resistance of the coil wire. A capacitor consists of two isolated conductors that have a voltage difference. Therefore every practical inductance has some parasitic capacitance.
The inductances of a Tesla transformer are given, as is the capacitance of the top capacitor. Therefore the capacitor at the primary coil is used to bring the Tesla transformer to resonance. Because both resonant circuits influence each other, both can not oscillate at the same frequency. The resonant circuits form a band pass. One resonant circuit oscillates at a lower frequency, the other at a higher frequency. A loose coupling gives a small difference between the two frequencies, a tight coupling gives a larger difference.
The traditional method to find the "optimum" setting is to measure the resonance frequency of the secondary resonant circuit. The primary coil is far away and does not influence this measurement. In the second step the primary resoant circuit is "tuned" to the same resonant frequency, again without influence from the other resonant circuit. In the third step the coupling factor between both circuits is optimized. Normally the primary coil is located at the base of the secondary coil. For looser coupling the primary coil is moved further away (further down) from the base. To allow this, the secondary winding does not start at the bottom of the secondary coil carrying structure, like an isolator material tube, but a little further up. The length of the high voltage spark is normally the criterion for optimality. But sometimes the capabilities of the interrupter, like the transistor temperature, or the capabilities of the power supply determine the optimum setting.

Slayer exciter or solid state flyback converter high voltage generator

The solid state flyback converter or "Slayer Exciter" replaces the relay coil and relay opener contact with a solid-state device, a transistor. Transistor is short for transfer resistor, an electronic device that changes the resistance value between connectors emitter and collector depending on the current between base and emitter. In the "Slayer Exciter" flyback voltage inverter circuit the transistor switches fast between the minimum resistance and maximum resistance state. The following circuit was published by "Ludic Science" at Youtube. The green LED can be replaced by two 1N4148 diodes in series connection.
The flyback converter is an oscillator. As any oscillator it needs a positive feedback from output of the amplifier to input of amplifier. The "Slayer Exciter" uses a clever (brilliant) feedback solution. After closing the switch, current can flow through the 22kΩ resistor, the base to emitter path within the transistor back to the battery. The base to emitter current brings the transistor into "minimum resistance" state. A (large) current can flow from the battery through the primary coil with 3 turns, the collector to emitter path within the transistor back to the battery. The difference in current creates a changing magnetic field in the primary coil which creates a changing magnetic field in the secondary coil which again creates a changing voltage in the secondary coil. The top connector of the secondary coil has no connection. This is not true. The electro magnetic field closed the circuit. There is an "invisible" load resistor between the top connector of the secondary coil and the negative pole of the battery. A current can flow through the secondary coil, the load resistor and the LED diode. The forward voltage of a green LED is between 1.5 volts and 2 volts. Because the LED cathode is connected to ground (the negative pole of the battery), the anode voltage of a conducting LED is -1.5 volts. Such a negative voltage does switch the transistor into "maximum resistance" state. The current through the "feed" resistor 22k still tries to switch the transistor "on", but the influence of the negative voltage at the LED cathode is stronger. The combined effect (sum of the currents into the base of the transistor) switches the transistor "off". The current through primary coil decreases, the magnetic field collapses, after a short time there is no more voltage at the secondary coil and no more "inhibit" voltage through the LED at the base. The current through the 22kΩ resistor can switch the transistor "on" and the cycle starts again. The LED is anti-parallel to the base-emitter diode of the transistor. The positive amplitude of the radio frequency high voltage travels through the transistor, the negative amplitude travels through the LED. There is no rectification in the secondary circuit (secondary coil, "invisible" Rload and the anti-parallel diodes). Oscillation can only happen if there is a time delay between cause and effect. The cause is the charging of the secondary coil via the primary coil. The effect is a negative voltage at D1 that switches off the transistor. Both coils can store magnetic energy. Without a large enough inductance (storage capacity) of the secondary coil, the effect negates too fast the cause and the oscillations die away. The transistor needs a short time (nano seconds) to forward the "switch off" state from the input to the output. This transition delay supports the oscillation, too.
Note: A flyback converter is a DC/DC converter. In the case of a radio frequency high voltage generator the rectifier diode at the secondary coil is missing. Because of this one can not differentiate between a flyback or forward converter. The author used the name flyback converter because a flyback converter is typically used if high DC output voltage, low output current is needed.

Picture: Transistor Ruhmkorff coil (Ludic Science Miniature Tesla Coil). Click on picture for Youtube video.

If the solid state flyback converter does not work

The correct wiring of the transformer is important for the operation of the high voltage generator. This problem does not exist with the Ruhmkorff coil or the ignition coil. If the flyback converter does not oscillate, you have to change the connection of the primary coil. Put the first connector at the place the second connector was and vice versa. This changes a negative feedback into a positive feedback. Another explanation is: The transistor as an amplifier has a phase shift of 180° between input and output. The transformer as part of the feedback network has to provide another 180° phase shift. Correct connection of the primary coil gives this phase shift, wrong connection gives a phase shift of 0°.

SPICE simulation of solid state flyback converter

LTSpice is a (very good) circuit simulation program from Linear Technology. Let's see what we can learn about the flyback converter circuit from simulation. The "Tesla" transformer consists of a AM radio ferrite rod with primary coil L1 with 6 turns and secondary coil L2 with 275 turns. The inductance of a coil is AL times the square of the number of windings. The AL value depends on the magnetic core material in the transformer. The simulation uses an AL value of 89nH per turn. This gives 3.2μH inductance for the primary coil and 4.8mH for the secondary. The magnetic coupling between primary and secondary coil was measured as 0.6. The resistance of the secondary winding was measured as 10.7Ω. The "invisible" load resistor is now visible as Rload. The flyback converter does not oscillate if the value of Rload is too high. The reason is simple. A large value of Rload allows only a (very) small current through LED D1 which can not overcome the influence of the "feed" resistor R1.
The simulated circuit has one more capacitor than the original circuit. C1 is a filter capacitor. The battery V1 has an internal resistance of 10Ω. At the time Q1 is switched off (no current flows through the transistor) the filter capacitor charges from the battery up to 9V. At the time Q1 is switched on C1 will discharge through the transistor. Because the internal resistance of C1 is (much) lower than the battery internal resistance the filter capacitor reduces the battery voltage fluctuations for the circuit. The transistor 2N2222 or today PN2222A, MBT2222A, PZT2222A was announced by Motorola at the 1962 IRE Convention. It is a modern oldtimer. Central Semiconductor offers a PN2222A SPICE model.
One has to use some tricks to use SPICE successful. That the battery has an internal resistance of 10Ω is no trick but realistic simulation. The first trick is to use the voltage source V_Iin with a voltage of zero volts to have a convenient component to measure the battery current. Another trick is to give the battery voltage a time dependent value. The option startup in the .tran command tells LTSpice to "ramp up" the supply voltage from 0V to 9V within the first 20μs of the simulation. This supply voltage disturbance starts the oscillator. Remember, "an object does not move until an external force ..." and "an oscillator does not start oscillating until a disturbance ...". 

Picture: LTSpice simulator of Slayer exciter (Transistor Ruhmkorff coil)

The simulation calculates the maximum output voltage for a turns ratio of 1:50. If the secondary winding has 275 turns, the primary winding should have 5.5. The oscillator will not start if the turns ratio gets too large. In this case add another turn to the primary winding.

Picture: Slayer exciter simulation diagrams. X-axis is turns ratio. Voutmax is maximum output voltage, voutrms is RMS output voltage, vcmax is Q1 collector voltage, pout is output power in mW and the last diagram shows the efficiency in percent.

The simulation waveform results look realistic. The output of the flyback converter "oscillator" is a fine sine wave. The transistor is in the "linear state" while it travels from the "minimum resistance" to the "maximum resistance" state. Linear state implies power loss. The "switch is off" state has (theoretically) no power loss because the current through the switch is zero. The "switch is on" state has (theoretically) no power loss because the voltage over the switch is zero. In the linear state neither current nor voltage are zero and we have power loss. The average power loss is 624mW. Only the PZT2222A version of the 2N2222A can handle it. A larger value of R1 reduces the output voltage and the power loss.

Picture top: Slayer exciter transistor collector voltage (magenta), transistor collector current (cyan) and transistor power loss (yellow).
Picture bottom: Slayer exciter output voltage.

Dr. Stiffler Exciter (SEC, Spatial Energy Coherence)

The Dr. Stiffler exciter is the "star" of many Youtube videos of the "free energy" community. A good conjurer does not want you to see the trick, he wants you to believe in a miracle. Normally Dr. Stiffler does not show the schematics of his one transistor circuits in his Youtube videos. One exception is the "Hybrid Exciters - Part#1" video from 2010 on Youtube. The presented circuit is a modified Meissner oscillator. Every oscillator needs three major building blocks. A resonance circuit, an amplifier and a positive feedback. In the Dr. Stiffler circuit there are several resonance circuits. The obvious LC-circuit is L1 and C1. More hidden is the LC-Circuit L2 with the L2 parasitic capacity. The transistor internal Miller capacity between collector and base creates together with L1, L2 the third LC-circuit. The amplifier is the base resistor R1, the npn transistor and the collector coil L2. The feedback is via magnetic coupling between L2 to L1 and/or capacitive coupling through the Miller capacity. Capacitor C2 forwards the AC radio frequency to the output but blocks the battery DC voltage.


Picture left: Dr. Stiffler Exciter. Click on picture for Youtube video.
Picture right: LTSpice Stiffler exciter simulation.

The simulation of the Dr. Stiffler exciter is done again with LTSpice. After some experimentation with component values the Stiffler exciter simulation surprises with 91% efficiency. This is very good for an one transistor radio frequency generator. For a load resistance between 75Ω and 150Ω the efficieny is between 90.5% and 91.5%. The transistor collector has a maximum voltage vcmax of 15.3V. The Dr. Stiffler exciter does produce RF energy, but not high voltage. The maximum output power pout is 500mW at 85% efficiency with R1=12k and Rload=33. Enough output power to drive some toy devices.

Picture: Stiffler exciter simulation diagrams. X-axis is Rload in Ω. Icmax is Q2 collector current, vcmax is Q2 collector voltage, pout is output power in mW and the last diagram shows the efficiency in percent.

pnp-npn exciter (complementary astable multivibrator)

The Slayer exciter and the Dr. Stiffler exciter need a transformer. This two transistor circuit needs only one inductance. An astable multivibrator switches a transistor off and on. The complementary astable multivibrator needs only one energy storage component for its operation, the normal astable multivibrator needs two.
Note: The complementary astable multivibrator is a relaxation oscillator. Because the circuit includes an inductance, The circuit can be called flyback converter, too.

Picture: High efficiency 6 to 24 volts DC input pnp-npn exciter

The complementary astable multivibrator uses a high amplification amplifier. The output and input of the amplifier is connected via an energy storage component. Oscillation can only happen if there is a time delay between cause and effect. The state (low or high voltage) of the output is feed back time delayed to the input. Because of the high amplification the transistors are switched fast between on and off. The operation of the circuit depends critically on the value of the base voltage of Q1. This voltage is stabilized by diode D1. R1 closes the current circuit for D1. R2 gives the D1 voltage a high impedance. This allows C2 to change the base voltage of Q1. The energy storage component is L1. The capacitor C1 influences the oscillating frequency, too. R3 limits the base current of Q2. Too much base current slows down the switching time of the transistor and reduces the efficiency. The pnp-npn exciter operates from 6 to 24 volts DC. The wide operation voltage is possible due to D1.
The efficiency is between 91.4% and 93.2% for Rload between 27Ω and 56Ω.

Picture: pnp-npn exciter simulation diagrams. X-axis is Rload in Ω. Icmax is Q2 collector current, vcmax is Q2 collector voltage, pout is output power in mW and the last diagram shows the efficiency in percent.

The KOSMOS Telekosmos-Praktikum Teil 1 from 1966 presented in "Abbildung 126" a complementary astable multivibrator of the npn-pnp type. In 1966 circuit efficiency was no topic, at least not for milli Watt electronics.

Picture: npn-pnp exciter from 1966. The name then was complementary astable multivibrator

Evolution of the Meissner oscillator circuit

The first patent application for an oscillator was for the Meissner oscillator. One can imagine that the inventor wanted to build an amplifier. In the 1910s there was only transformer coupling between stages. One transformer at the amplifier input, another transformer at the output. But unfortunately for the amplifier builder these transformers did couple with each other and did oscillate. After the unnecessary components were removed the Meissner oscillator was invented. Another invention by chance?

A useful Meissner oscillator needs some more components. The amplifier should have some large signal handling capabilities, because for an one transistor amplifier a one volt signal is a large signal. The frequency response should have equal amplification for the oscillator frequency range. The Miller capacitance of the amplifier should not have a negative influence on the oscillator. The LC-circuit should be fed from a high impedance source, but the amplifier output impedance should be low impedance. The amplifier in the picture below fulfills these requirements. The frequency range is 1MHz to 2MHz. Just enough for the oscillator in an AM radio superhet receiver with 470kHz intermediate frequency (IF). The amp is capacitor coupled with C2 at the input and C3, C5 at the output. Resistor R3 controls the DC-operation point and R4 is the load resistor. The RC-circuits R1, C2 and R2, C3 couple the LC-circuit to the amplifier. R1 reduces influence of the Miller capacitance from the amplifier output to the amplifier input. R2 increases the amplifier output impedance. The secondary coil L2 performs a necessary phase shift. The transistor creates 180° phase shift, the RF-transformer L1, L2 creates another 180° phase shift and the positive feedback requirement is fulfilled.
The amplifier should have an amplification of exactly 1. After power on, the amplifier needs an amplification above 1 to build up the amplitude. The presented Meissner oscillator uses two anti-parallel diodes D1, D2 as amplitude limiter. The amplification is always above 1. If the amplitude gets above the threshold voltage of the diodes, the signal gets limited. The building blocks are in the order amplifier output, limiter, LC-circuit and amplifier input to have minimum distortion from the limiter at the oscillator output. Furthermore the amplification is just above 1, the amplifier runs "softly" into the limitation. The values of R1, R2 set the low frequency amplification (less resistance is more amplification), the values of C2, C3 the high frequency amplification (more capacitance is more amplification).
Oscillator simulation with SPICE is always a little challenge. Please remember, SPICE tries to find a numerical approximation for the solution of a linear equation system. An oscillator device fits poorly in this framework. The author uses the LTSpice options numdgt=7 method=Gear plotwinsize=0. These options reduce the integration error and allow more precise frequency spectrum (FFT) calculation. Sometimes LTSpice needs even more guidance to succeed. In this circuit LTSpice needs a small maximum timestep value of 10 nano seconds to start oscillation.

Picture top: Meissner oscillator. Not the best oscillator ever build, but clearly above the average oscillator circuit!
Picture bottom: Meissner oscillator spectrum for five different frequencies. A clean output for a one transistor oscillator.

The Meissner oscillator can be build with a JFET (junction field effect transistor). A JFET does perform like a pentode vacuum tube. An amplitude limiter is build in the pentode and JFET. There is an internal rectifier between gate (grid) and source (cathode). The external diode D1 completes the amplitude limiter. The RC-circuit R3, C2 is the grid leak. The other components perform the same tasks as in the transistor Meissner oscillator above. The BF245A in TO92 housing is no longer in production, but the successor BF545A in SOT23 housing is.

Picture top: Meissner FET oscillator.
Picture bottom: Meissner FET oscillator spectrum for five different frequencies.

Evolution of the Hartley oscillator circuit

We can imagine an evolution of the Hartley oscillator circuit from a simple LC-circuit into the full schematics. The author hopes that his "Hartley oscillator evolution" can show the relations between LC-circuit, Pi-filter (Collins filter) and oscillator.

Picture: Hartley oscillator evolution

The two passive components C2 and R1 are a grid leak. C2 allows the AC signal from the split inductance LC-circuit pass to the input of amplifier Q1. R1 sets the DC operation point of Q1. The Hartley oscillator evolution helps to find good values for L1 and L2. It is obvious that in circuit a) the voltage at C1 and L1 are equal. If in circuit b) L1 and L2 have the same value it should be obvious that the voltage at L1 and L2 is half of the voltage at C1. This makes the Hartley oscillator a poor power oscillator. The AC voltage at L1 is normally the output voltage of the Hartley oscillator. To maximize the output voltage, the inductance of L2 should be small, the inductance of L1 should be large. The voltage at the large inductance drives the output. The correct split of the inductance into L1 and L2 depends on the current amplification value of the amplifier.
The Hartley oscillator is a poor choice for a variable frequency oscillator with variable capacitance tuning. Both connectors of C1 are "hot".

Long tail oscillator

A sine wave oscillator can be build without transformer. Two transistors form a 0° phase shift amplifier. This amplifier is called long tail or emitter coupled amplifier. The amplitude limiter uses two 1N4148 diodes. The oscillator produces a smaller output voltage and less distortions if two BAT54 Schottky diodes are used. But the larger parasitic capacitance of the Schottky diodes reduces the frequency range of the oscillator.

Picture top: Long tail oscillator. It needs only a simple LC-circuit.
Picture bottom: Long tail oscillator spectrum for five different frequencies. It is noisier then the Meissner oscillator.

Wien bridge oscillator

The Wien bridge oscillator uses two RC-circuits for the frequency dependent network. The frequency range of the famous HP200CD sine wave generator is 5Hz to 600kHz. The HP200CD amplitude limiter uses a light bulb. A amplitude limiter with two anti-parallel diodes works, too. The Wien bridge uses the parallel RC-circuit R1, C1 and the series RC-circuit R2, C2. R1 and R2 are a tandem (stereo) potentiometer. Different values of C1, C2 are switched into the circuit. A ten to one variation of R1, R2 gives a ten to one variation of the output frequency. The Wien bridge network has an attenuation factor of 3. The amplifier network R3 to R6 has an amplification factor a little above 3. The anti-parallel diodes D1, D2 are the amplitude limiter. The LM833 op-amp is the European answer to the famous U.S.A. NE5532. Both are low noise bipolar op-amps that are in production for more then 30 years.

Picture top: Wien bridge oscillator. It needs a tandem potentiometer.
Picture bottom: Wien bridge oscillator spectrum for seven different frequencies.

Superhet (superheterodyne) receiver

An oscillator is the first building block of a superhet (superheterodyne) receiver. The second block is the mixer. The mixer is a multiplier. The frequency of the radio station we want to hear (RF) is multiplied with the frequency of a local oscillator (LO). The product is the intermediate frequency (IF) and another frequency. The equation is IF1=LO-RF and IF2=LO+RF. The IF amplifier does only amplify a small frequency range around the IF frequency. One needs a rectifier and an audio frequency amplifier to complete a superhet.

Symmetric double-balanced mixer

The MC1496 or LM1496 is an ancient integrated circuit for a double-balanced mixer. ON semiconductor still produces it in a SOIC-14 package, and Mouser and others distribute it. Unfortunately there is no schematics that uses symmetric ports for radio frequency in (RF), local oscillator in (LO) and intermediate frequency out (IF) in the application note AN531/D "MC1496 Balanced Modulator". But only with this (expensive) circuit a Gilbert cell mixer shows the full potential. The three RF transformers for the mixer are build with ferrite rings FT50-77 for AM radio and FT50-43 for shortwave radio. The multiple voltage divider R1, R2, R3 provides +1.1V at pin 5 for bias, +4V at pins 1, 4 for base voltage of the lower two transistors of the Gilbert cell and +8V at pins 6,12 for base voltage of the upper four transistors. These voltages have filter capacitors C1 and C2. For spurious free IF output the input voltages at RF and LO should be below 10mV RMS. This is a difference between diode ring mixers and Gilbert cell mixers. The diode ring mixer needs a (very) strong LO signal, the Gilbert cell mixer needs a (very) weak LO signal.
The author of the LM1496 LTSpice model is "Ron H".
Some years ago the SA602 (SA612, NE602, NE612) was used as mixer/LO by radio amateurs. But this device is no longer in production.

Picture: LM1496 symmetric double-balanced mixer

Download LTSpice files

The LTSpice from the author and for convenience the third party SPICE model files are included in this Zip file

About the author and other experiments with radio frequency power source

The author made his ignition coil high voltage generator experiments in the late 1970s. At this time he has not heard about micro computers and was (still) fascinated by "brute force" electrical experiments like short circuit the plus and minus pole of a car battery with a wire. Today the author is more refined. But surely he will build a miniature Tesla coil high voltage generator. The days of "Frankenstein laboratory" experiments are not over. I must repeat the Youtube (pseudo) scientific experiments with a radio frequency generator and a 1N4148 diode in a test tube that will produce hydrogen and oxygen bubbles. I am not in search of free energy or black matter, but I am in search of fun and a little oxyhydrogen explosion.

Picture: Dr. Stiffler SEC Exciter Hydrogen production via diode. Click on picture for Youtube video. The government does not want you to know!

Hear the master himself again in the next video. Dr. Stiffler shows how a one wire feed line together with a capacitive ground connection can drive a little DC motor. The RF power in the feed line is locally transformed into DC voltage by a pair of 1N4148 diodes as explained by Jonnydavro. It is still a nice experiment, even if you put away the mysticism. But maybe today you have to add some secret touch to interest people in electronics. Well done, WA7RTQ.

Picture: Dr. Stiffler SEC drives little DC motor. Click on picture for Youtube video.

Another nice example of "ancient" technology in new clothes is the following Youtube video about a SEC exciter. There is nothing wrong with the oscillator circuit, it is of the split inductance or Hartley oscillator type. That the LEDs are connected with only one wire is not strange, either. Tens of years ago a lot of "plain old telephones" where connected this way. The ground connection did close the circuit. That the supply current drops after more load is added should be no miracle by now: The change of Rload moves the operating point of the transistor to a more efficient point. This shift of the operating point does more than compensate the additional current that is needed for the increased load. There is still no free energy here.

Picture: Jonnydavro's simple SEC Exciter drives LEDs. Click on picture for Youtube video.

You can contact the author at this email address: