Two Mosfets for AC

Discussion in 'The Projects Forum' started by urb-nurd, Jan 23, 2015.

  1. urb-nurd

    Thread Starter Member

    Jul 9, 2014
    269
    3
    Im having trouble getting my head around using two mosfets to pass AC.
    [​IMG]
    I have reduced my schematic to its bare bones to help keep it un-cluttered.

    I am taking 25V AC rms through a bridge rectifier to allow me to maintain a DC level at the gate through the voltage divider with the aid of a cap, the diode ensures the cap discharges onto the gate only.

    I understand the above circuit does not work, however i am a little lost at where to go about correcting it.

    Any assistance is greatly appreciated!
     
  2. wayneh

    Expert

    Sep 9, 2010
    12,085
    3,024
    How do you know it doesn't work? What is the point of this circuit? It seems like a way to waste power.

    Don't forget the body diodes, which will conduct no matter what the state of the gate.

    Once current passes through the MOSFET, you route it back to rectifier "-". That makes no sense and might send the gate voltages up above their rated max.
     
  3. urb-nurd

    Thread Starter Member

    Jul 9, 2014
    269
    3
    I know it doesn't work because when i tested it, the mosfet fried instantly and my Breadboard melted a little! haha.
    Excuse my brain today if you will, i am on a new medication that has quite the psychotropic effect.

    As you say, the return current passes to the bridge - this was evident from the location of melted plastic.
    I knew when i hooked this up that it was incorrect, yet i tried anyway!
    I foolishly hoped the body diode in the lower fet would be a favourable path for the return current than the bridge -ve terminal - this was not true!

    As for the purpose, i omitted my PWM circuitry that shorts the gate voltage via a gate drive opto. This is being use to dim a halogen bulb.

    I had the circuit working with a single mosfet using the bridge to rectify the ac for my bulb, however i intend to use AC and then an LC circuit of sorts to reconstitute the sinusoid from its pulsed representation.

    I am a little confused with a few things.
    One being the position of the bulb, using a single mosfet - it was easy to see the low side configuration and work with it.
    Another being the common source voltage - in the schematic above i had it going to the negative output terminal of the bridge rectifier, hoping that this would allow the necessary gate-source voltage .
    Now i want to introduce a second fet for the negative portion to eliminate the rectifier, it is not so obvious.



    If you could suggest how i could make a change to put this into effect.
    I would appreciate any help
    THanks!
     
    Last edited: Jan 23, 2015
  4. wayneh

    Expert

    Sep 9, 2010
    12,085
    3,024
    Ahh, OK. You need to read about triacs.
     
  5. urb-nurd

    Thread Starter Member

    Jul 9, 2014
    269
    3
    Why do i need to know about triacs?
    I understand triacs and the role SCR's play in controlling an ac load - however it is not simple to PWM a triac - maybe it is and you have the secret!
     
  6. urb-nurd

    Thread Starter Member

    Jul 9, 2014
    269
    3
    this link shows a person suggesting a similar layout.
    Im failing to see how mine differs aside from the fact they use a battery for the gate where i aim deriving the gate drive voltage from my ac source.
     
  7. ScottWang

    Moderator

    Aug 23, 2012
    4,848
    767
    I'm not sure the circuit will work correct or not, but the mosfet is not a logic level and it needs 10V for Vgs, and the Vgs is not enough for the up mosfet.
     
  8. urb-nurd

    Thread Starter Member

    Jul 9, 2014
    269
    3
    The mosfet works just fine at 5v - check the datasheet.
    Anyway, that is besides the point - with a 34v peak the v divider will maintain +6.5V which is more than adequate.
    But that being said, the voltage divider is not the issue with the circuit.
    Thanks for the response
     
  9. ScottWang

    Moderator

    Aug 23, 2012
    4,848
    767
    Did you measure the Id?
    When the Vgs <4V then you only can get less than 1mA Id current, if the Vga<10V then the mosfet can't get into the saturation area.
     
  10. urb-nurd

    Thread Starter Member

    Jul 9, 2014
    269
    3
    I did not measure Id, however - when i used this with a single mosfet and passed pulsed dc via the rectifier through the bulb: all was well.
    So the gate voltage and current isnt the issue here, i can watch the gate capacitance charge on my scope. its fine.
    [​IMG]
    Here is a snippet from the datasheet showing that the voltage at the divider is no the issue.
     
  11. wayneh

    Expert

    Sep 9, 2010
    12,085
    3,024
    I don't see how the MOSFETs are ever NOT conducting (assuming the gate voltage is enough to turn them on). The sources are tied together at DC ground, so the gates are always on?
     
  12. urb-nurd

    Thread Starter Member

    Jul 9, 2014
    269
    3
    Correct!
    I will obviously PWM the gate voltage, however for the test circuit i ensured the gates were held high enough to keep them both fully on.
    The sources are tied to DC ground to ensure the voltage at the gate is above the source voltage.

    This is where i am stuck, is it correct to do as i have?
    Should i use a resistor between the sourced and the DC ground to ensure the load current takes the fet and its partners body diode?
     
  13. ScottWang

    Moderator

    Aug 23, 2012
    4,848
    767
    The block diagram show that the Vds=25, but you want the Vds=0V when the mosfet get into the saturation status.
     
  14. urb-nurd

    Thread Starter Member

    Jul 9, 2014
    269
    3
    im afraid im lost.
    What block diagram, if you mean my schematic above? then yes - i guess technically it is at 25V rms when not conducting.
     
  15. ScottWang

    Moderator

    Aug 23, 2012
    4,848
    767
    Sorry, I meant to the Vgs-Id graphics that you shown on #10.
     
  16. urb-nurd

    Thread Starter Member

    Jul 9, 2014
    269
    3
    Im thinking a resistor between the source and DC ground will ensure the load current takes the body diode.
    Essentially allowing one mosfet to control the switching action per cycle, allowing the second to use its body diode.
     
  17. ScottWang

    Moderator

    Aug 23, 2012
    4,848
    767
    In the circuit that you didn't consider the power of low side, when it will work(Vgs)?
     
  18. crutschow

    Expert

    Mar 14, 2008
    12,973
    3,219
    Why do you want to use PWM rather than a TRIAC phase dimmer?

    Note that D1 is redundant in your circuit since it is series with the bridge diodes.

    The MOSFETs block the current when they are off (gates connected to the common source) but when the MOSFETs are turned on there is a short between M2 and the bridge diodes when the top terminal of the AC source is negative. This allows current to flow from M2's source through the top bridge diode, limited only by the diode and source impedances (poof).
    If you draw the diodes in the bridge you will see the path.

    Using a battery for the gate allows the gate-source voltage to be isolated from the mains, thus there is no short-circuit path.
     
  19. urb-nurd

    Thread Starter Member

    Jul 9, 2014
    269
    3
    Hit the nail on the head.
    Thanks!
    I can see the diodes redundancy now, however i have noticed strange behaviour without it.

    The reason i am not using a triac for phase angle control is that i am using the bulb as a heating element to effect PID control.
    That isn't so easy with a triac.
     
  20. urb-nurd

    Thread Starter Member

    Jul 9, 2014
    269
    3
    i also do not have the battery option, as i am trying to power this from a single wall outlet (all fully isolated!)
     
Loading...