Step 3: Driver Circuit

In normal automotive use the current in the primary is interrupted at the appropriate time to generate a spark at the spark plug to ignite the fuel-air mixture in a car’s engine. In old cars the current is interrupted by a mechanical switch called the points, which are opened and closed by a cam on a shaft that rotates when the engine is running. In newer implementations a sensor sends a signal to an electronic module to cause a semiconductor switch in the module to open and generate the spark at the proper time.

All that is needed to use the coil to generate arcs for experimentation and display is a circuit that turns a switch on and off to interrupt the primary current of the ignition coil. Concerns such as adjustments to the timing to ignite the fuel in cylinder at the proper time do not apply for a display application.

The diagram below shows the basic blocks of a system used to generate acrs using an ignition coil. The general drive circuit consists of a pulse generator that turns a semiconductor switch (a transistor, FET, or in the circuit presented later, an IGBT) on and off at a desired rate. The power supply can be any battery or DC power supply of an appropriate voltage, capable of supplying enough current.

The function of each section of the system is described in detail next. Refer also to the detailed schematic.

+5V Regulated Supply
A LM7805 regulator, VR1, provides the +5V regulated voltage for the pulse generator circuitry. Diode D1 in series with the input to the regulator is used to to prevent damage if the power source is connected backwards. There are electrolytic and ceramic capacitors (C1, C2, C3, and C4) on both the input to and output from the regulator for filtering.

Pulse Generator
As described earlier, energy is wasted if the switch remains on past the time the steady state primary coil current is reached, so the pulse generator circuit must be designed and adjusted with this in mind. Again, the time constant of the primary circuit will determine how long it takes for the primary current to reach its steady state. The inductance and resistance of the primary will vary between different coils, so the maximum on time of the switch will need some adjustment depending on the characteristics of the coil you are using. Some coils are intended to operate with additional resistance in series with the primary coil, and this resistance will need to be added to the other resistance in the primary circuit when calculating the steady state current and time constant.
Two different designs of the pulse generator are described in the next steps. One is based on a PIC microprocessor, and the other based on a 555 timer.

Pulse Generator Based on Microprocessor
A PIC microcontroller based pulse generator circuit is used to generate the signal used to turn the IGBT on and off. Using a microcontroller gives the flexibility to control both the on time and the frequency independently. The software allows the spark repetition frequency to be set between 10Hz and 400Hz, and it allows the on time of the pulse to be set between 0.5 and 2 milliseconds.

The maximum current can be limiting by adjustment of the on time. If the on time is set to less than about four time constants, then the current in the primary will never reach its steady state.

The circuit uses two potentiometers, R3 and R7. One is used to set the on time, and the other to set the repetition frequency. The potentiometers are read by the microprocessors A/D converter. R4 and C5 and R6 and C6 are used as low pass filters on these analog inputs. The software generates the output pulses based on the on time and frequency readings.

One processor input is used to read an enable switch. When the switch is pressed, the input is brought low, and the microprocessor generates pulses on the output with the on time and frequency as set by the potentiometers. When the switch is released R1 pulls the input high and the output is turned off.

The program flow chart diagram, *.asm source code file, and *.hex file for programming are included.

Alternate Pulse Generator Based on 555 Timer
This 555 timer circuit can also be used for the pulse generator, but it can’t generally let you control the on time and frequency independently. Refer to the separate schematic for the 555 timer version of the pulse generator.

The schematic shows a 555 timer circuit which uses diodes to create separate charge and discharge paths. The on time is determined by how long C1 takes to charge through R1, potentiometer R2, and diode D1. The off time is determined by how long C1 takes to discharge through R3, potentiometer R4, and diode D2. By separating the charge and discharge paths, the on and off times can be controlled independently, allowing the user to set the on time so that it is not so long that it results in wasted energy. Note however that changes to either the on time or the off time will result in a change to the pulse repetition frequency, unlike the microprocessor based version of the pulse generator.

The ON time, OFF time, and frequency of the output are determined as follows:

Ton = 0.693 * (R1 + R2) * C1
Toff = 0.693 * (R3 + R4) * C1
Frequency = 1/ (Ton + Toff) = 1.44 / [ C1 * (R1 + R2 + R3 + R4) ]

Pushbutton switch SW1 connects the reset line of the timer high, enabling the pulse generator output. The output goes low when SW1 is released, turning off the HV output.

The switch used here to control the current in the primary is an International Rectifier IRGB14C40LPBF IGBT. This IGBT is specifically designed for automotive ignition system applications.

When the switch in an ignition system opens to disrupt the current in the primary, the voltage across it rises up to hundreds of volts. This IGBT is rated to handle this voltage. It also includes internal clamping diodes to protect against overvoltage. Refer to the internal diagram of the part shown below to see how the internal protection is implemented.

Proper heatsinking is needed for your switch. This is especially true if you ever plan on overdriving your coil with a higher power supply voltage.  Exact heatsinking needs require some calculation of the actual power dissipated in your switch.  To err on the side of caution mount to a piece of aluminum and also use some forced air, such as from a small fan.  In the system shown here, the heatsinking is provided by a large copper area of the PCB.

Circuit Protection
It is beneficial to provide protection to the driver circuit to avoid damage from transient overvoltage “spikes”. MOV1 is placed across the power supply input to clamp voltage spikes. MOVs (metal oxide varistor) are protection devices which are designed to be off when the voltage across them is below their clamping voltage, but then turn on and conduct once that voltage is exceeded. When it is conducting, the voltage drop across it is relatively constant. The MOV therefore “clamps” the voltage across the power supply bus to a safe level, so that it does not rise high enough to damage the driver circuit.

The MOV must be selected with a clamping voltage that is greater than the power supply voltage; otherwise the MOV will be conducting all the time and it will overheat. For example, the MOV I chose for use with a 14 volt power supply clamps at around 20 volts. Transient spikes greater than 20 volts will then be clamped so that the power supply bus will not see a voltage greater than that.

It is also useful to have a larger of capacitor (C7) on the power supply input. A couple thousand microfarads worth of electrolytic capacitors works well. The capacitors provide a low impedance path to high frequency spikes on the power supply. Make sure that the voltage ratings of the capacitors you use are greater than the supply voltage and greater than the MOV clamping voltage.
<p>Do you know what your pulse timings were for the jacob's ladder application?</p>
How did you protect the Microcontroller from the EMC effect ?
I have built one of these using the 555 timer and I now see that a PIC based design allows a greater degree of fine tuning to get to the resonant frequency of the coil. This seems important as different ingnition coils have slightly different capabilities in their output with the same circuit and voltage parameters. Some have used two coils in parallel/antiparallel to achieve even higher output voltages and longer arcs.<br><br>I would love to try your PC based circuit but I currently have no way of programming a PIC and have only Macintosh laptops at my disposal. I suppose I could invest in a PIC programmer and borrow a PC to just load the files on to the PIC. There are other projects such as the Aurora 9x18 RGB LED art (PIC24F08KA101) that also use a PIC from Microchip. Could I just buy the Microchip's PICkit 3 In-Circuit Debugger/Programmer? Would this work with the files you provide here?? Or could you sell me a few programmed PICs? I could also buy a few PICs and fedex to you for programming, if you would be willing to help me.<br><br>Also, could you send me a BOM for the parts list?<br><br>OK to email direct: kuriloff@nyhni.org
Actually, ignition coils don't depend on resonance to generate the HV output. <br>There is a misconception among some that an ignition coil operates like a Tesla coil. An igntion coil is actually an iron core transformer, with a turns ratio on the order of 75:1. Switching off the primary current causes the primary voltage to rise to several hundred volts, which is in turn stepped up by the turns ratio. <br> <br>What is important is to have an idea of the inductance and the resistance of the primary, as that lets you determine the time constant. From there you can configure your pulse generator to turn the switch on for an appropriate length of time. <br> <br>I'm not familiar with the PICkit 3 In-Circuit Debugger/Programmer. If it is a product made by Microchip for general in circuit programming then it should be able to be used. I have actually never programmed any parts using in circuit programming before, but I know that there is an app note on Microchips web site explaining what you need to do regarding connections to the processor. <br> <br>The files I have attached should work regardless of the programmer you are using. You will have to use the 12F683 with the files supplied, unless you know how to update the source code for a different PIC and then assemble to get a hex file for programming the different PIC. That isn't terribly difficult if you know a bit about the assembly language. <br> <br>I don't have a BOM of the circuit per se, but the schematic has manufacturers PNs for all the more critical parts, like the processor and the IGBT. The other stuff is just common resistors and caps. 1/4 Watt resistors work fine. The caps you select must be rated to handle whatever maximum voltage you plan on applying to them. For example, the ones on the input to the +5V regulator will see the power supply voltage, so select them depending on the max power supply voltage you intend to use.
a question here: <br>if ignition coils do not depend on resonance, then why does one need a capacitor in an auto ignition system ? <br>
<p>It will organize and stabilize your current.</p>
Thanks so much for your comments and the files.<br><br>Dan
Your comment got me thinking about something unrelated to the jaccob's ladder, but directly back to automotive applications. Many car modders are trying to extract every last little bit of power from their cars. One common way is with better ignition. Perhaps if coils such as these operate best at their resonant freq and that they all have a unique freq, then it may be beneficial for car enthusiasts to get their coil &quot;tuned&quot; to their optimal resonant freq. Perhaps with a little mass produced 555 IC based circuit built by you. I dunno, but there could be some dollars in a cheap little black box that goes between you coil(s) and power supply, with a potentiometer or something for tuning to the coils best resonance.Im sure you could elaborate further on that idea. Perhaps something is already available, but need simplification Comments :D
<p>flybacks are better at lower voltages and are cheaper</p>
<p>It depends on your circuit. With an coil like this you won't lose most energy in heat. Most of your energy will turn into heat with a flyback coil. So this coils are better if you don't want hotter sparks on an zvs with flyback..</p>
<p>can you help me? do you give circuit that?</p>
<p>it is asome</p>
This will prolly sound quite dumb, but im going to ask anyway, as this will help me understand what the driving circuit does a little better. <br> <br>If I use no driving circuit and just apply approx 13v DC to the coil and have 1 wire out of where a lead to a distributor would go, and 1 wire to the negative ( for the sparks, are these called anodes and cathodes?), then have the positive isolated with a switch. <br> <br>Is it true that I will get just 1 quick spark just a split second after I turn switch off?. <br> <br>What i am trying to understand is if the driving circuit just switches the supply power on &amp; off very quickly!?!?!? and a spark is discharged when supply is interrupted. If this is true, does allowing an earth have the same effect as switching off current, except current is still on. <br> <br> In other words will the mere act of presenting an earth then trigger a spark while un-driven power is applied? Or is it a case that absolutely nothing will happen if constant uninterrupted power is applied?
Nothing will happen if the primary coil is just driven with a constant current. It is the interuption of the current in the primary that causes the changing magnetic flux through the secondary, inducing a high voltage on the output terminal. The presence of an grounded or earthed conductor will not result in a spark under those circumstances.
i wonder how much the value of the coil is ?
If you mean the inductance of the coil, the one I used had a primary inductance of about 4.5 mH. The primary resistance was about 1.5 ohms. <br> <br>The inductance of the secondary was too high to measure with my meter. The secondary resistance was more than 10k ohms if I recall correctly.
cool idea
How would I make this in a negative voltage<br>
The polarity could be reversed by changin the wiring on the primary side so that the &quot;-&quot; connection of the coil connects to the power supply, and the &quot;+&quot; connection is connected to ground when the switch closes. I have not tried this, however. <br> <br>I've seen projects where some people have driven two coils, with opposite polarity output, so that the potential difference between the two is double. If this is what you have in mind, I would suggest that you use two of the exact same coils. The reason being is that the discharge on the output is fairly brief, and the timing is determined by the inductances and resistances and turns ratios of the coils. If two different coils were used, the output voltage peaks may occur at different times between the two coils, and the output may no be as great as you hope. Again, I haven't tried this type of dual coil arrangement, so I am sort of speculating here.
Well designed circuit. I've seen a few designs that are best describes as shabby.
Back in the 1980, we use to build cdi ignitions for cars with points. They had a pot core harting oscillator transformer to charge the 2 0,47uF capacitors to around 400v and a thyristor to discharge it through the coil. The circuit was designed so that it kept the original coil, points and capacitor so that in the event of cdi ignition failure, you would just switch a couple of connector to return to normal. The original coil on a VW combi ran cooler and the points only needed ajustment due to mechanical wear and the ignition never failed.
I've built something similar for a tractor a few years ago. Made a vary noticeable difference in the idle and starting.
What would happen if I would connect a power audio amplifier (12V max) instead of the pulse generator... would I get a singing arch??
This type of setup won't really work for the &quot;plasma speaker&quot; type projects. Those use a much higher drive frequency than can be used with an igniton coil. The drive frequency is usually tens of kilohertz, and then the audio waveform is used to alter the pulse width to modulate the arc to produce the sound. <br> <br>The primary of an ignition coil has a fairly high inductance, and so the current can't be driven with a high enough &quot;carrier frequency&quot; which could then be modulated to produce any kind of reasonable quality audio. Flyback transformers are usually what is used for plasma speakers. <br> <br>At any rate, the drive to the switch needs to be a digital signal to turn the switch either on or off, and so you wouldn't want to apply an analog signal directly to the gate/base of the switch you are using.
Thank you for your comprehensive answer :)<br>I asked because it is easier for me to get for free a ignition coil than a flyback.<br>One more question. What type of material should be used for the electrodes, so they would not melt after a couple of minutes of arc production?
The electrodes in this project are just 14 ga copper wire. I had no issues of the metal melting, but the insulation near the ends started to burn after a while. Brass brazing rods work well, and that is what I used on my Jacobs ladder that was built using a neon sign transformer. The brass can be somewhat better than copper in that it is stiff and so it will keep its shape better. But the copper works fine, just strip off all the insulation off to be safe. <br>The wires may get a little warm after producing arcs continously for a while. <br> <br>I've only run the setup here for about 3 minutes at a time. If you are planning on making something to run for much longer, its best to check it out and see just how hot it gets. An of course, don't leave it unattended. <br> <br>
Fun idea, I may have to try that.
You have taken a complex subject and you have described all the information needed to produce one of the best Instructibles I've ever read. Great job and thanks for sharing!
I was always of the opinion that the main reason for the capacitor across the contact points was to allow the current in the coil to collapse rapidly, thereby inducing a very high voltage for the spark.
Well done. I have been threatening to use a GM HEI coil, with 555 to do something similar, for quite some time. I have an old function generator I could use for a signal. but I'd rather keep from HV.. The old tool ain't worth much, but it works. I ruin a 555, and associate components I'm not out of much
A good explanation, but you have omitted to describe the purpose of the capacitor connected in parallel with the &quot;points&quot; in the typical automotive application. <br> <br>The method I have used in the pasts is to drive che primary by discharging a 1 /uFcapacitor charged at 400V with a thyristor. With hi-voltage coils one can reach 60,000-70,000V without any damage to the coil.
You are correct, I did not include a mention of the capacitor (&quot;condenser&quot;) that is used with the mechanical points system. That part allows a pathway for the current to flow once the switch is turned off, and in the mechanical setup it prevents pitting of the switch contacts. My design doesn't use such a cap, as the IGBT is use has internal clamps that protect it from damage. <br> <br>The system you describe, where a cap is charged to 400V and then discharged through a coil via thyristor is called capacitve discharge igniton (CDI). I've never experimented with such a system, but I belive they use a different type of coil. Anyway, the older inductive discharge &quot;Kettering&quot; type system that my design is easier to implement, as CDI needs a circuit that can generate the ~400V needed to first charge the cap.
Yes, you are right on the capacitor, but it serves also an other purpose. The 1 &mu;F capacitor, combined with the inductance of the system, creates a resonant circuit with a decaying AC voltage with a frequency in the region of 50KHz, if I remember correctly; facilitating in this way the transfer of energy to the sparkplugs. As a matter of fact, I would avoid using the over voltage clamping of the IGBT. The reason is that any current that is shunted to ground represents wasted energy. I would use a suitable capacitor instead. Because it is a reactive component, no energy will be wasted. I bet that at the secondary one would get a more &quot;robust&quot; spark. Please note that obviously the voltage at the IGBT will go negative also during the inductive kickback. <br> <br>To achieve the 400V in the capacitive discharge method, I used the common voltage multiplicator circuit (Villard circuit and similar) to go from the typical household AC voltage line to the desired voltage, making the circuit quite simple.. <br>
I have wondered what difference it would make is the primary circuit was allowed to &quot;ring down&quot; via a capacitor across the switch, like the arrangement used with mechanical points. <br> <br>I'd like to try a capacitve discharge type circuit, but I think I would need to use a coil intended for a CDI system. The coils I used in the project here were all intended for inductive disharge systems, and have an inductance of about 3 to 5 mH.
Funny, I posted an other answer and it has misteriously vanished. <br> <br>For a confirmation of the purpose of the capacitor please read the following explanation from: <br> <br>http://en.wikipedia.org/wiki/Ignition_system <br> <br>(quote):.... <br> At the same time, current exits the coil's primary winding and begins to charge up the capacitor (&quot;condenser&quot;) that lies across the now-open breaker points. This capacitor and the coil&rsquo;s primary windings form an oscillating LC circuit. This LC circuit produces a damped, oscillating current which bounces energy between the capacitor&rsquo;s electric field and the ignition coil&rsquo;s magnetic field. The oscillating current in the coil&rsquo;s primary, which produces an oscillating magnetic field in the coil, extends the high voltage pulse at the output of the secondary windings. This high voltage thus continues beyond the time of the initial field collapse pulse. The oscillation continues until the circuit&rsquo;s energy is consumed. <br>(end) <br> <br>I personally designed, build and sold SCR ignition units in the early seventies. <br>One of the various occupations that help me to pay for my MSc EE. <br> <br>The standard coil had no problem holding the voltage, because the spark at the sparkplugs acts as a voltage limiter. On the other hand, if one bench drives a standard coil with a large airgap, it is 100% certain that the standard coil will be damaged. As I said with a hi-voltage coil and a SCR unit, it is easy to achieve 60,000-70,000Volts
For the sake of completeness I have to correct you. The capacitor is a fundamental part of the ignition system and without a car will not start.<br> I am afraid there are no misconceptions. In any case job well done.<br> <br> From: http://en.wikipedia.org/wiki/Ignition_system<br> (Quote)<br> <em>At the same time, current exits the coil's primary winding and begins to charge up the capacitor (&quot;condenser&quot;) that lies across the now-open breaker points. This capacitor and the coil&rsquo;s primary windings form an oscillating LC circuit. This LC circuit produces a damped, oscillating current which bounces energy between the capacitor&rsquo;s electric field and the ignition coil&rsquo;s magnetic field. The oscillating current in the coil&rsquo;s primary, which produces an oscillating magnetic field in the coil, extends the high voltage pulse at the output of the secondary windings. This high voltage thus continues beyond the time of the initial field collapse pulse. The oscillation continues until the circuit&rsquo;s energy is consumed</em>.<br> (End of quote)<br>

About This Instructable


190 favorites


Bio: "But I was going to Toshi station to pick up some power converters!"
More by LargeMouthBass: Audio Delay Module Reverse Engineering an Electric Fence Charger Stop Time with an LED Stroboscope!
Add instructable to: