Back EMF pulse

Discussion in 'General Electronics Chat' started by Jaguarjoe, Nov 2, 2010.

  1. Jaguarjoe

    Thread Starter Active Member

    Apr 7, 2010
    770
    90
    The back EMF pulse from an ignition coil primary (-) side switched on/off to ground by a GM HEI module is about 30nS wide and maybe 300 volts tall. Current flow starts out at about 5 amps. The other side of the coil is connected to the (+) power rail of the car through 5 ft of #18 wire to the battery and alternator.
    How does the wire and battery/alternator appear to the back EMF pulse? At that short of a duration is it a low impedance path to ground or (+) 12v?
     
  2. thatoneguy

    AAC Fanatic!

    Feb 19, 2009
    6,357
    718
    The side closer to the ECU (supply) has blocking diodes preventing the pulse from reaching the ECU/Engine Control Unit/Power Supply.

    The other side of the coil is a spark gap that ignites the air/fuel mixture in the compression chamber.
     
  3. Jaguarjoe

    Thread Starter Active Member

    Apr 7, 2010
    770
    90
    On the (+) side of the coil primary there is a 500nF capacitor to ground for noise suppression. My friend is telling me the ignition module needs that cap to provide a "low impedance path to ground for the module to work". I'm saying that the module does not need that cap to work and that the 12v power supply provides a low impedance path for the (+) side of the coil primary.

    Here is the HEI module:

    http://www.jaycar.com.au/images_uploaded/MC3334R0.PDF
     
  4. thatoneguy

    AAC Fanatic!

    Feb 19, 2009
    6,357
    718
    The cap does pass high frequency noise, "ringing" from coil, to ground to prevent broad band electromagnetic interference both back to the ECU, and to nearby electronics, such as the radio, or the radio of the car next to you. The diode prevents a reverse voltage spike which would damage the logic of the ECU Driver.
     
  5. Jaguarjoe

    Thread Starter Active Member

    Apr 7, 2010
    770
    90
    OK, I'll ask another way.
    That 500nF cap is in parallel with a 5 ft piece of #18 wire and a car battery/alternator. Which one will conduct a 30nS pulse better, the cap or the wire and batt/alt?
     
  6. thatoneguy

    AAC Fanatic!

    Feb 19, 2009
    6,357
    718
    It depends how close the cap is to the source of the noise.

    Either way, it becomes an RC lowpass filter, with the R being the resistance of the wire.

    See: http://www.allaboutcircuits.com/vol_2/chpt_4/3.html

    The cap should connect the coil to ground as close as possible, so that the wire doesn't become a "better antenna".

    The circuit should function without it, but the interference could cause strange things to happen to the ECU or other computers onboard the vehicle. The diode isn't optional.
     
  7. Jaguarjoe

    Thread Starter Active Member

    Apr 7, 2010
    770
    90
    Actually, the circuit works fine without the zener diode as long as all of the spark plugs are in service because the plugs limit the secondary voltage to about 20- 25kv which, in turn, limits the back emf spike amplitude. OTOH, with no secondary load voltages approach 40- 50kv which cause a huge back emf pulse which is death for the output transistor.

    The cap is mounted as close to the source as practical, within inches.
     
  8. DonQ

    Active Member

    May 6, 2009
    320
    11
    It depends.

    The 'pulse' can be thought of as being composed of different frequency components. The sharp rise, and any sharp corners, generally are composed of higher frequency components, whereas the flat or gently curving parts of the pulse are composed of lower frequency components. This matters because the highest frequency components will definitely pass through the cap better, and the lowest frequency components will definitely pass through the wire better.

    Somewhere in between is a point where it is half and half, this is known as the 'corner' frequency, fc (called by many as the cut-off frequency, but since nothing is ever 'cut-off' here, I'll stick with 'corner'). If you want, fc can be calculated, but since it means nothing till the frequency components of the 'pulse' are known, it's probably OK to just take it on faith for now that many of the frequency components depend on the capacitor.

    Also, at least some ignition systems use the inductance of the coil and the capacitor to create a resonant LC circuit that will oscillate when provoked, and create a number of sparks, very closely spaced in time, for each opening of the points (or the electronic equivalent). You may not be able to measure this oscillation, and on some CD systems in particular it may not even be used. I have been able to see the effect by looking at and hearing the difference in the spark across a plug when the capacitor is disconnected. Dramatic.
     
  9. Jaguarjoe

    Thread Starter Active Member

    Apr 7, 2010
    770
    90
    Also, at least some ignition systems use the inductance of the coil and the capacitor to create a resonant LC circuit that will oscillate when provoked, and create a number of sparks, very closely spaced in time, for each opening of the points (or the electronic equivalent). You may not be able to measure this oscillation, and on some CD systems in particular it may not even be used. I have been able to see the effect by looking at and hearing the difference in the spark across a plug when the capacitor is disconnected. Dramatic.

    That was one of the functions of the capacitor across the points on a Kettering points ignition system, right? I don't see how a cap across the 12v bus could do much resonating once the points or output xistor open up.
     
  10. shortbus

    AAC Fanatic!

    Sep 30, 2009
    4,019
    1,542
    I was always told that the cap on a points/Kettering ignition was to store the High voltage on the primary side of the coil, from the collapse of the secondary EMF. A points ignition always gives a higher/stronger spark while it's running than it does while cranking because of this.

    As a secondary use of the points cap, it was to slow down pitting of the points. To stop radio interference a second cap was sometimes added , but not all cars needed this extra cap.
     
  11. DonQ

    Active Member

    May 6, 2009
    320
    11
    Several ways available to wire the low-tension system. The way easiest to think about it/describe it online is:

    Code ( (Unknown Language)):
    1.  
    2. Ground --- coil ---long wire -+- points --- battery
    3.                               |
    4.                         capacitor in distributor
    5.                               |
    6.                             Ground
    7.  
    8. or
    9.  
    10.  
    11. Ground --- coil -+-long wire -+- points --- battery
    12.                  |
    13.          capacitor on coil
    14.                  |
    15.               Ground
    16.  
    (Many other options available. You can turn things around, battery-coil-points-ground, but the principle is the same. Other options on another day.)

    Either of these 2 ways, points made, coil 'charges'; points break, you are left with a classic parallel LC circuit, one of L & C sharing a ground, the other ends connected together with a wire, and with the inductor (spark coil) with a full magnetic charge. The high-tension side (the spark), being inductively coupled, goes along for the ride.

    Boinnnggg!


    Another use... CD (capacitor discharge) systems can use a capacitor to force one cycle of the oscillation to return most of the unused charge to the capacitor in preparation for the next cycle. This makes recharging the cap for the next discharge take less time.
     
  12. Jaguarjoe

    Thread Starter Active Member

    Apr 7, 2010
    770
    90
    Its wired like this:

    Battery-- long wire------------------cap-(+)coil(-) -----darlington----gnd

    other side of cap is gnd.
     
    Last edited: Nov 4, 2010
  13. DonQ

    Active Member

    May 6, 2009
    320
    11
    Complicated further by the fact that darlingtons usually have a back emf diode built-in.
     
  14. Jaguarjoe

    Thread Starter Active Member

    Apr 7, 2010
    770
    90
    Which way is the diode pointing?
    If the cathode end went to the collector, it wouldn't do anything because the big back emf spike is of positive polarity. The diode would be reverse biased.
    It would be pointless if the anode went to the collector, it would always conduct, as long as power was applied to the (+) side of the coil.

    MOSFETS have a body diode.
     
  15. debe

    Well-Known Member

    Sep 21, 2010
    947
    184
    Perhaps a circuit of a Ign drive might help you explain it easier? Daryl
     
  16. Jaguarjoe

    Thread Starter Active Member

    Apr 7, 2010
    770
    90
  17. thatoneguy

    AAC Fanatic!

    Feb 19, 2009
    6,357
    718
    It shows a cap from coil +V to ground labled Cfilter, as well as the blocking diode on the output.

    If you've had this all along, I guess I am confused by the question.
     
  18. DonQ

    Active Member

    May 6, 2009
    320
    11
    (edit: this is about the image, not the PDF. Mostly, the PDF is fairly similar.)

    Several things.

    The drawing doesn't show the internal diode on the darlington, but if you look at the data-sheet, the BU941 does have one (with a short search I couldn't find the other transistor, but it probably has one too).

    In this ignition version, the last part of the inductive kick is limited by the zener diodes. While the spark is sparking, the low-tension voltage will be clamped at some reasonable voltage. But once the spark can no longer maintain, the voltage on the low-tension would spike without the zeners. This stops that from killing the transistor. Other strategies would use this energy for multiple hf sparks, or store it for use in the next spark.

    There is no oscillation capacitor shown on this version. For different strategies, it could exist at several places. Some locations would make the resonant LC circuit which would spark, then current would flow the other way for the next half-cycle of the oscillation. This is where the darlington diode would have an effect.

    Another place the capacitor could be would be to make a low impedance point at the positive lead of the coil. This circuit has a resistor in the + lead. Usually, the resistor is bypassed while cranking to get a better spark, especially since the starter will pull down the battery voltage. Once started, the starter no longer loads the battery and the battery voltage is now boosted by the alternator running. The resistor lowers the voltage on the coil, but makes it a high impedance point. A cap to ground at this point would lower the impedance.

    A cap directly across the coil, instead of to ground, makes an obvious resonant circuit. A cap to ground, along with a cap elsewhere in the circuit could also make a resonant circuit.

    So many options, so little time...


    I would also caution everyone on not being limited by what you see in a magazine. Just because it is published, doesn't necessarily mean that it is a good circuit. Often what you see is something that someone was able to get to be functional, without any guarantee that it is an optimized circuit. This circuit seems to be functional. But there is no low impedance after the ballast resistor; No attempt for a multiple spark; It dumps excess energy by using large wattage zeners, etc. Good, maybe, for experimenting, maybe not so good for engine or electrical efficiency.
     
    Last edited: Nov 4, 2010
  19. debe

    Well-Known Member

    Sep 21, 2010
    947
    184
    The circuit i posted was only part of a Programable Igition system for vehicles it comes as a Kit to assemble. It also used inputs from a Knock sensor & Manifold vacumm sensor & the timing could be programed & use different triger systems. I figured if i posted a circuit people might get the gist of what was being described about transistorised ignition. Daryl
     
  20. Jaguarjoe

    Thread Starter Active Member

    Apr 7, 2010
    770
    90
    Read post #3 for the question.

    What blocking diode? There's a 350v zener clipper and a diode across C-E inside the MJ10012 darlington.
     
Loading...