High-efficency MOSFET relay

Discussion in 'General Electronics Chat' started by cmartinez, Oct 6, 2014.

  1. cmartinez

    Thread Starter AAC Fanatic!

    Jan 17, 2007
    3,555
    2,512
    Hello, everyone.

    I'm trying to build a high-efficiency AC MOSFET relay, configured as the classic back-to-back n-channel pair, such as the one shown here.
    JDT actually drew an excellent driver for this configuration a few years ago, which can be found at the end of this thread.

    The basic circuit works because most (if not all) MOSFETS have a inherent reverse diode built into them. So when current flows in one of the MOSFETS in the allowed on-direction (from positive Drain to negative source), the other MOSFET will let it through its body diode (from positive source to negative drain) independently of the voltage present at its gate. I hope I'm making sense here... This behavior is reversed as the polarity changes.
    The issue here is that even though a MOSFET's on-resistance is usually quite low (about 0.55 ohms in the IRF740, for instance) most MOSFET's body diodes are not very efficient. The body diode present in the IRF740 has a voltage drop of about 2V, for instance. That means that the MOSFETs will have a tendency to heat up when used in this configuration, even if we're running them well below their rated current.

    Consider this example, if we were to run 8 amps through the IRF740's body diode, the 2V voltage drop would force it to dissipate about 16 watts of heat, which in my book is quite a lot... (feel free to correct me if I'm wrong or just saying something stupid)
    If I were to add an external normal or schottky diode in parallel with each of the MOSFET's body diode then maybe things would improve quite a bit, since they would be relieving the MOSFET's body diodes from some current.

    Considering that a MOSFET can behave like a diode when properly configured (like shown in this video) and that their equivalent resistance is very low, my question here is this:
    Is there a way to add a couple (or more) additional MOSFETs to the basic back-to-back AC relay circuit, so that they could behave as the external diodes I just described?
    Maybe this way the circuit would become far more efficient and would generate less heat.

    Thank you all for your patience and your time.
     
  2. wayneh

    Expert

    Sep 9, 2010
    12,090
    3,027
    Not sure if this helps, but in an active rectifier, a comparator detects the forward voltage drop across the body diode and turns on the gate to increase the conduction and eliminate the diode voltage drop.
     
  3. cmartinez

    Thread Starter AAC Fanatic!

    Jan 17, 2007
    3,555
    2,512
    Thanks for the advice... but could you please elaborate?
     
  4. shortbus

    AAC Fanatic!

    Sep 30, 2009
    4,003
    1,521
    This is a total guess, but the IRF740 and others in that type are fairly old technology (for components). The high Rds is the give away on this. Maybe using a "newer" mosfet, with better characteristics would be a place to start? Pretty sure that the 'newer' mosfets with lower Rds rating will also have better intrinsic diode ratings too. Worth looking at a few data sheets to see?
     
  5. wayneh

    Expert

    Sep 9, 2010
    12,090
    3,027
    Take a look at this thread. That's pretty much the extent of anything I know about it.
     
  6. Sensacell

    Well-Known Member

    Jun 19, 2012
    1,127
    266
    If BOTH MOSFETs have the gates driven positive with respect to the source at a sufficient level, the on resistance of the circuit will be very low (Rds ON x 2) the MOSFET body diode will not carry the current, the FET will. If the Rds ON is low enough, the body diode will not be in play.

    No need for extra parts- just find a part with low Rds ON.
     
  7. cmartinez

    Thread Starter AAC Fanatic!

    Jan 17, 2007
    3,555
    2,512
    See, that's what I fail to understand, and maybe that's what wayneh was trying to tell me. Do you mean that (in this back-to-back configuration) each of the MOSFETs is capable of conducting current in BOTH directions? I thought that a MOSFET could only conduct in one direction...
     
  8. cmartinez

    Thread Starter AAC Fanatic!

    Jan 17, 2007
    3,555
    2,512
    You're right... the IRF740 is indeed an old model... I'll do some more browsing around and see what I can find... thanks.
     
  9. wayneh

    Expert

    Sep 9, 2010
    12,090
    3,027
    MOSFETs definitely conduct in both directions.
     
  10. cmartinez

    Thread Starter AAC Fanatic!

    Jan 17, 2007
    3,555
    2,512
    Oh my... you're absolutely right... Check this video, in min 4:40 to 6:30 the narrator explains why the MOSFET heats up when current is driven through its diode, and what happens when voltage is applied to its gate in this mode.
    Thanks Shortbus for your suggestion, I'll do some experimenting with the TK10P60W, it seems to have far better characteristics than the IRF740
    Thank you all of you for your time. And thanks Wayneh for directing me to this thread, there's tons of things to learn from it.
     
    shortbus likes this.
Loading...