Coil Driver \ Spark Tester

Discussion in 'The Projects Forum' started by electratech, Jan 5, 2009.

  1. electratech

    Thread Starter Member

    Jan 5, 2009
    Hello to everyone, if you check my profile you will see that I am an auto mechanic, I can help you with your car - no problem but this electronics stuff is a brain killer. I can troubleshoot them but have trouble building them. Some of the stuff you guys talk about here is making me sweat!:confused:

    What I would like to build is a coil driver. I want to test a coil off and away from the vehicle. The coils generally have a 1 ohm primary resitance and about 5-7kΩ secondary. they are capable of producing as much as 100κv but really need averge 10κv and max of about 25κv. different vehicles vary slightly but usually I can test the coil output with a special spark tester with a load of estimated value of 25κv. The problem is that I have to crank over the engine and allow the vehicles electronic system to activate the coil. Some vehicles you have to remove the intake manifold to get to the coils and not a good time to be cranking engine over.:eek:

    I have a 2n3055 transistor to fire the coil, a pulse width driver used to pulse fuel injectors to drive transitor base and a condensor to place at positive pole of primary windings to ground (all ignition systems use a condensor at the coil) don't know why they just do. I think it's to prevent arcing from spikes.

    I'm posting my idea to gather insite and any thoughts I should consider before I build.

    I do have one direct question though. Transistor states
    Vebo = 7v what does that represent?:confused:
  2. SgtWookie


    Jul 17, 2007
    Welcome aboard! Don't worry about being confused around here - most are. :confused: :D

    Interesting that you bring that up. I've been kicking around some numbers lately.
    The generally accepted dielectric strength (insulation strength) for air is 3kv/mm; that is, in order for a spark to jump across a 1mm gap at sea level in dry air, 3,000V are required.

    But, that'll change when you increase the atmospheric pressure, like when an engine is running at idle and you mash down the loud pedal. If your compression ratio is 10:1, suddenly you'll need 30,000V to jump that same 1mm gap (just under 0.040"). WOT at idle is a worst-case scenario.

    If you really want to test your coils under similar conditions as when they're in operation, you would need to multiply the standard spark gap by the engine's compression ratio, and make a spark gap accordingly wide.

    I invite you to verify my suggestion; easy to do if you have a diagnostic O-scope.

    Not unless you like having fast moving objects departing through the oil pan!

    The condenser (an old-fashioned term for "capacitor", still popular in automotive circles) is like a storage tank for electricity. Long wires have inductance, which resist changes in current. Having power stored in a capacitor/condenser right next to the coil gives a ready source of power when the ignition sinks current on the ground side.

    You've gotten one or two already - more to come, I'm sure ;)

    That is the maximum Emitter to Base voltage; if you exceed that, you will damage the transistor. Basically (sic) you don't want the voltage on the base to go lower than or equal to the emitter voltage minus 7 volts.

    Note also that the maximum base current is 7A. I don't know what your injector tester's output voltage is, but it can probably blast out a fair amount of current. Seems to me that injectors run on somewhere around 3.3v, if I remember correctly, but I don't know offhand what their resistance is. But, if you used a 1 Ohm 5 Watt non-inductive resistor on the base, that would limit current to 3.3A, and put the 2N3055 into saturation. The non-inductive part is important; many wire-wound resistors ARE inductive. If you use an inductive resistor, when the current from the injector driver cuts off you'll wind up with an inductive "spike" that will "zap" the base-emitter junction - this relates back directly to Vebo.

    Something else is going to happen when the transistor gets turned off is that the ignition coil will try to keep the current flow going. This will result in a very large reverse voltage "spike" across the coil's primary that could reach hundreds of volts quicker than you can say "dang!" Since you have a condenser/capacitor on the + side of the coil, it'll absorb the spike on that side, but the transistor's collector will get a real wallop.

    It's really late here and I'm getting tired - will have to get back to this later.
    Last edited: Jan 6, 2009
  3. SgtWookie


    Jul 17, 2007
    OK then, continuing on a bit.

    In the last post, I ended where the coil primary will be generating a large reverse EMF pulse (aka "voltage spike" or simply "spike") so something must be done with that, or it'll "eat" the 2N3055 transistor. If you look at the datasheet for the transistor, you'll see that the maximum allowed Vce is 100V - that is the maximum allowed voltage when measured from the collector to the emitter.

    One easy way to take care of this reverse EMF is to use what's often called a "flywheel diode" across the coil's primary + and - connections. The diode is connected with the cathode (the end with the band; in a schematic the cathode end is pointed to by the arrow) towards the + terminal, and the anode to the - terminal.

    Have a look at the attached schematic.
    Capacitor/condensor C1 represents the one you usually see in the system. You can use a known good one, new or used. I chose to use 13.6v for the supply voltage, which would be typical when the engine is running.

    T1 represents the ignition coil; primary side on the left.
    Note D1, this is the "flywheel diode". I've used a common power rectifier diode in this case.

    I've reduced the base resistor R1 to 0.33 Ohms; this works in simulation. I'd forgotten about the (approximate) 0.7v Vbe. However, you can start out with a higher resistance and see how well it works - or not.

    If you're stuck using an inductive resistor, you could connect another flywheel diode from the base to the emitter, cathode connected to the base. Radio Shack carries a 0.47 Ohm 10w resistor that should do it.

    V2 represents your injector tester, set to output a 3.3v 0.2mS pulse every 10mS (100Hz).

    R2 makes a poor substitute for a spark gap, but the simulation program doesn't include such esoteric devices.

    All of this being said, it's still not a really accurate simulation of what a real ignition system does. It's my understanding that GM ignitions actually use a brief series of pulses to generate a very rapid series of sparks for each combustion cycle. This helps to ensure more reliable and complete combustion.

    The circuit thus far is purely experimental. Your mileage may vary.
  4. electratech

    Thread Starter Member

    Jan 5, 2009
    Thank you for the welcome and the reply. You seem to know electronics like I know cars. your knowledge is impressive.

    You hit on so many good points I don't know where to begin my reply. First let me say thanks for such intense effort and time you put into this project and sharing of your knowledge.
    several points you made has given me some new in site to the project. First thanks for explaining the purpose of the condenser. I do know that without it, the spark is terrible and the transistor or points will burn out very quickly. Now I know why and it makes sense. That is why there are reverse bias diodes in parallel with many solenoids, relays, ac compressor clutch coils, etc.
    second point.
    depending on the application, an injector generally has a resistance of 12.6 ohms room temp but some may have a value of only 1.5Ω. all working on battery voltage! killer amps here! However, if circuit is pulsed it will decrease the load ( 50% duty cycle = 1/2 the voltage) the average pulse width on injectors is around 3.5ms at idle. If direct constant voltage is applied, the injector would have a short life span and be at full throttle continuesly. Anyway, getting back on track, I think I will build a pulse unit from a 555 chip. Ive found a few circuits here and there that might work better. I was thinking keeping it simple since I had a unit that already produced a pulse. I didn't consider anything else but the on/off pulse. The reason it is such a struggle building electrical circuits is because they don't always tell you everything to do the math. Some data sheets do and others don't. I completely missed the base current amps and that's a little important, don't you think?

    Yes I have a scope but it is more of a digital graphing meter than a true scope. Actually its a Snap On Vantage. Nice unit for automotive use. We used to use a scope to test ignition systems in vehicles that had standard single coil driven via distributors, cap & rotor, ignition cables but now days many systems are utilizing COP (coil on plugs) with each cylinder having a designated coil. Thus making secondary side of coil impossible to measure. Only the primary is accessible if you can even get to that!
    In theory, it all seems soooo easy but berried and entangled with all the engine, trans, chassis, etc it is not so easy to test components. Don't worry, it's only getting worse! Even the pins are getting so tiny you can't even backprobe. They got so many bells and whistles on cars that they have to invert a 42volt system just so they can reduce that amperage. And Hybrids, don't get me started! All I know is that anyone that doesn't have a good understanding in electronics, will not make it in this field. There is a difference between TECHNICIANS & MECHANICS. Tech's keep mechanics busy! Cars that park themselves, navigation control panels, steer by wire, brakes by wire, stability control, oh yeah! guess who gets to fix them when they break? The brite side, maybe this year I'll break 35k.

    How do I upload a picture or schematic without putting it on a web site first? I was going to show you a brief schematic but couldn't upload it from computer.

    Anyway, thank you very much for your help. I'll keep you posted on my progress and be sure to study and apply everything I've learned. I like this site, excellent videos. A++++​
  5. Audioguru


    Dec 20, 2007
    The max allowed collector to emitter voltage of a 2N3055 transistor is 60V. If you use a reversed diode to limit the voltage to only 14.4V when it turns off then the high voltage from the ignition coil will be much too low. The ignition coil is designed to stepup the 100V at the transistor when it turns off.
  6. electratech

    Thread Starter Member

    Jan 5, 2009
    Can you put that in dumb terms, I'm a little lost on that one.
  7. SgtWookie


    Jul 17, 2007
    I just spent the better part of an hour typing up a reply, and my browser froze up solid.

    Sorry, but it's gone. I'll try again tomorrow.
  8. electratech

    Thread Starter Member

    Jan 5, 2009
    Well, I did it! I made up a 555 astable circuit and powered up a 203055 transistor along with the use of a condenser to keep destruction of voltage induced backward EMF from destroying circuit. So far it is working at 1" spark gap on a GM DIS coil. I would like it to go from terminal to terminal but need to find out what is actually going on with the transistor overheating when putting it under full stress level of the coil. This circuit has no problem cracking out a strong spark on a standard coil 20kv or so but GM DIS coils can produce 100kv! Somehow I need to figure out how to step up the current or cool down the transistor. I have it on a heat sink but should be much larger and perhaps a fan to keep it cool. I should also put a current limiter or even a fuse on coil primary circuit to prevent transistor failure.
    This unit works very well for driving fuel injectors. You can adjust the Hz by T=R1+R2*C1, love to upload a schematic but ??????
    Anyway, increasing Hz makes spark thin and long. Decrease Hz makes spark fat and short.
    I'll keep you all posted as I work things out. Feel free to share your thoughts.
  9. SgtWookie


    Jul 17, 2007
    OK, I'll suggest that you don't want to force a coil under test to to jump a spark gap of more than 20mm (just over 3/4") because that's 60,000 volts. That's enough to fire an engine that has 12:1 compression ratio with a 0.065" plug gap (1.66mm) running WOT from slow idle. You'll risk carbon tracking/surface arcing/insulation breakdown of the coil - then it's time to dump it.

    Would you explain what you mean by this?

    The primary of the coil has low resistance - about 1 Ohm you said. If the transistor is conducting for long enough in a duty cycle, and you have enough current (3.3A) going through the base, the transistor will have about 10 Amperes of current flowing through it, with as much as a 3v collector-emitter voltage drop. This is right from the datasheet; look up Vce(sat). So, we have 10A across a 3v drop.

    Referring to Ohm's Law:
    P=EI, or Power(Watts) = Voltage * Current(Amperes)
    P=3V * 10A = 30 Watts
    You have a nice little room heater. ;) You'll need a good-sized heatsink.
    This is why MOSFETs are so popular nowadays; instead of a large Vce(sat), they have a low Rds(on) (Resistance from Drain to Source) when they are conducting.

    If you change to a capable power MOSFET, you will get both more current through the coil (because of the low voltage drop across the MOSFET) and less heat dissipation in the MOSFET than the transistor. One caveat: they are sensitive to static electricity; one "zap" and they're done. Transistors are much less sensitive to static electricity.

    You'd need a pretty large fuse.

    Tell us the values that you are using for R1, R2 and C1, and put a link to where you got the schematic you're using.

    If you are leaving the transistor ON too long, you are saturating the coil, and you may be causing damage to it; and you'll certainly heat up the transistor a lot faster.

    What do I mean by saturating the coil, anyway? :confused:
    When current is first applied across a coil, it doesn't start flowing immediately; it takes time to build up the magnetic field around the coil. The current gradually increases at a steady rate, until at some point it saturates; up until this point the rate of current increase was like coasting down a hill in a car, at saturation it's like driving off a cliff. :eek: The current flow increases very rapidly after saturation.

    Here's a very helpful page:
    Scroll about halfway down. To the right of Figure 6, you'll see a 'scope image of current flow through an inductor when it's saturating.

    That whole page is well worth the time reading through, even if you don't understand it all on the first few go-rounds.

    Do you have a condenser(capacitor) on both the high (+) and low (-) sides of the coil, and are they well grounded?

    Which resistor are you varying to change the frequency?

    I don't happen to have any spare ignition coils sitting around here to play with, and really don't feel like pulling one out of our cars while they're working just fine... ;)
  10. electratech

    Thread Starter Member

    Jan 5, 2009
    Ther is alot that goes into making this thing. I am amazed of everything I've learned such as the mc3334 chip and the mj10012 transistor designed for GM HEI ignition systems. The pulse is one thing but dwell is how long the coil is energized. The project is to advanced for me. I understand how to test them what what the outcome should be but to actually design it, thats alot of work and I lack alot of the knowledge I need. Ho do you keep up with everything? So many thousands of components with tons of spec's to consider. It will drive you to the looney ben for sure!
    Maybe, I'll play around with it just for kicks but not for ay testing purposes. I also wonder about the coil designs of other vehicles. They are all different per manufacture and that might hinder test results or even destroy coils thatare being tested. There is alot to think about here. However, I had fun learning and exploring. If anything, I did make a couple cool gadjets from the 555 chip. Kinda cool! Thanks for your help.
  11. SgtWookie


    Jul 17, 2007
    It does require some exercises of the noodle ;)
    MC3334? I can't seem to find a datasheet for that. I've found the MC33340, MC33341, MC33342 - but those are basically charge controller IC's.

    Useful NPN Darlington power transistors. MJE13009's are even more capable.
    In the "good old days" of points, the dwell was for how many degrees of rotation the points stayed closed. The problem with this scheme was that as distributor RPMs increased, the time that current flowed through the coil's primary before being cut off would vary widely. Some manufacturers like Mallory came out with dual point distributors that helped increase the dwell time for high RPM operation. But once electronic ignition came on the scene, the design of the coil could be optimized for a single pulsewidth (or series of pulses) to give a much better controlled spark over a wide range of operating conditions.

    Well, all you really need is to duplicate the signal that the ignition system is using to ground the - side of the coil. It may be just one pulse, or in the case of GM HEI ignitions, it may be a series of pulses.

    If you can take a picture of your O-scope monitoring the coil's negative terminal, and what your scope settings are, we could figure out how to mimic it using a timer or pairs of timers, or perhaps a microcontroller.

    I don't. :)
    Variety is the spice of life. ;) One can't keep up with everything, but parametric search engines help quite a bit in finding a "best fit" for a part to use in a project.
    That's the thing; you'd want to simulate the actual input from the ignition system for each given coil. It can differ significantly.
    Glad you're having fun with it :) You can do lots of things with the 555 timer IC; it's one of the most versatile ICs ever invented, and certainly the most popular. Even though it'll celebrate it's 40th birthday next year, it's more popular than ever. As of 2003, more than1 billion of these were being made each year.
  12. electratech

    Thread Starter Member

    Jan 5, 2009
    Here ya go. here is the data sheet. Problem now is can't seem to find the chip BUT I can use an ignition module that has the chip and transistor already inside. I just need to generate an AC signal like the one you see in the data sheet. I could use a little motor I suppose but who wants to sit there and spin it? I'd rather just push a button:cool:

    Anyway, as far as I understand, the series of pulses you are talking about is actually the capacitance inside the sparkplug. The plug itself is generating the series of pulses not the coil or module.

    And... the two terminals on the coil we talked about is actually the two towers for better words. DIS Distributorless ignition systems use one coil for each pair of cylinders known as waste spark system. It fires one plug in correct polarity and the other in reverse. on fires on the compression stroke and the other on the adjacent cylinder exhaust stroke as wasted.
    It take about 5kv more to fire a plug in reverse. It's actually a "loop" firing system. The wasted kv are minimal about 2-3kv at most.

    If you remove the wires from the coil, it will jump from tower to tower about 1". We test them this way to see if spark has consistency power and not shorting to ground. Of course only for a few seconds to keep from overheating. But all things considered the module is mounted on a huge heat sink!

    I had it sparking really good until the heat killed it. One thing I did notice in the data sheet that makes me want to try it a few more times is the use of a 450v zener diode on the transistor. The capacitor is not fast enough to absorb that type of surge. It must be channeled to ground. That is alot of backward emf.
    If weren't trying to jump such a large gap, it may have lasted longer of course.
    Like I said, I'm gonna just play with it and see how it goes. The data sheet helps alot and also this site I recently found.
    I am going to also give you the scheme I was working with. I made a few mods but started with this as a starting point.
    when I redo, I'll post the schematic here.
    Here is my summary of research.

    data sheet