PWM mosfet to drive induction coil and handle the collapse without suppressing it

Discussion in 'The Projects Forum' started by clone477, Oct 3, 2008.

  1. clone477

    Thread Starter Member

    Oct 3, 2008
    11
    0
    Have been over the site and it has alot of great info on here...
    I am a automotive technician by trade, I do alot of work with electrical problems but have only designed a few circuits before......

    I need to build a PWM circuit, with frequency and pulse width adjustments. From my research that I have done, I will need to use a mosfet to drive my high induction coil, and I will need optical isolation to protect my circuit. It was recommended to me to NOT use a 555 based timer for this, what other option is there for this??

    Also is the mosfet is powered separate from the timer part of the circuit?? What is the voltage and current the mosfet can handle through it??

    NOW FOR MY MAIN CONCERN....
    I do not want to suppress the discharge from the colapse of my inductor, thats why I need the optical isolator. Im doing experiment with colapse discharge. How can I design the circuit to not be destroyed by the collapse without using diodes or capacitors.

    I will need a upper frequency of maybe 80k , thats an estimate not for sure. Voltage and amperage are not determined yet, But Im guessing the larger mosfet I get the better for handling the coil collapse asnd not destroying it. I also assume that I will need a heat sink for the mosfet??

    Thanks for any help guys, if someone could just reccomend some componants, and maybe link me to a wiring diagram, I have tried searching with not much luck.
     
  2. mik3

    Senior Member

    Feb 4, 2008
    4,846
    63
    There are lot of different MOSFETs in the market, each one with its characteristics. You have to tell us the voltage and current requirements of your coil to tell you an appropriate MOSFET.
    Also, its better to use a diode in parallel with a small capacitor to suppress the back EMF created by the collapsing magnetic field of the inductor because in this way you will increase the average current through it and wont destroy the MOSFET by exceeding its maximum voltage ratings. A simple opto-isolator is not enough to protect your circuit.
    Yes, you will need a heatsink on the MOSFET with its size depending on your power requirements.

    Here are some links:

    http://www.discovercircuits.com/H-Corner/PWM.htm

    http://www.eleinmec.com/article.asp?28

    http://www.cpemma.co.uk/pwm.html

    http://www.maxim-ic.com/appnotes.cfm/an_pk/3201

    The best way to produce PWM is to use a microcontroller.
     
  3. SgtWookie

    Expert

    Jul 17, 2007
    22,182
    1,728
    Have a look at the attached; it's just a theoretical model.

    The inductor is charged through Rload, a 1 Ohm resistor, via the MOSFET. In this simulation, the MOSFET is turned on for 0.2mS out of 0.5mS.

    The green trace on the simulated O-scope is the voltage at the bottom of the inductor. The yellow trace is the current through the inductor.

    When the MOSFET is conducting, the current through the inductor increases. When the MOSFET turns off, D1, a 50A Schottky diode provides a current path back through the load. This keeps the current through the load relatively constant. Diodes used in this manner are sometimes called "flywheel" diodes.

    If D1 were not present, when the MOSFET stopped conducting there would not be a current path for the inductor. This would cause the inductor to momentarily generate a very high peak reverse voltage as it's field collapsed, very likely resulting in the destruction of the MOSFET.

    It would help to know what you're actually planning on doing.
     
  4. Bernard

    AAC Fanatic!

    Aug 7, 2008
    4,172
    397
    While we'r waiting for your info on L, I , V , maybe we can answer some questions ; Logic voltage about 12 to 15 volts, coil V [cV]TBD ; an optical isolater would be used in the FET gate drive ckt.but any amplifird/drive ckts would be at risk if it cannot take full cV in event of FET failure. There is a continuing battle between BJT's and FET V ratings but either can be obtained with V of over 450 V. [ IRF734; 1.2Ω. 4.9 A, 450 V , about $.65 ]. For trancient limited supression MOV's and thyrector's are available up to 1000 V. Whats wrong with using 555's? Maybe someone will tell us how to drive cascaded FET's.
     
  5. mik3

    Senior Member

    Feb 4, 2008
    4,846
    63
    I am sorry i missed this part in your post "Im doing experiment with collapse discharge"

    You can choose a high voltage MOSFET and when you will want to turn it off dont turn it off too fast. Turn it off gradually (calculate the appropriate time needed) as not to create a higher back EMF than the MOS maximum voltage rating as not to destroy it.

    v=L*(dI/dt)

    The more time you give the current to reduce the less the back EMF. Control this time by turning off the MOS at a reasonable time. The higher the maximum voltage of the MOS the faster you can turn it off without destroy it.

    Also, you will need an isolation transformer which will supply the power to the coil and an optocoupler which will drive the MOS to protect your control circuit.
     
  6. Bernard

    AAC Fanatic!

    Aug 7, 2008
    4,172
    397
    I presume that you can measure or view the inductive kick back from your coil, so you might start with low voltages and work your way up to V s you are comfortable with [500-1000], then pick FETor transistor to match. With + to one side of coil, pulse other end with a wire to _. Not high tec but might save blowing FET's. The [500-1000V] is kick backV.
     
  7. Bernard

    AAC Fanatic!

    Aug 7, 2008
    4,172
    397
    To test my last suggestion,'used a .32 Henry choke,10Ω,pulsed with a 1.5V battery;viewed 1200v pulse followed by a high frequency oscillation.If the short arc could be prevented the pulse would have been greater?
     
  8. mik3

    Senior Member

    Feb 4, 2008
    4,846
    63
    How do you measure such high pulsating voltages?
    Do you step the voltage down with a voltage divider?
     
  9. Bernard

    AAC Fanatic!

    Aug 7, 2008
    4,172
    397
    Used a divide by 10 scope probe on 20V/div. scale;peak V was in the 6th div.& verry hard to see even in a dark room. This is where a higher pulse rate would be helpfull. Next step might be to use .5V & a high V FET as a switch to hopefully eleminate the arc. To go further we need more info.
     
Loading...