Full-on power MOSFET

Discussion in 'General Electronics Chat' started by smoothVTer, Feb 19, 2013.

  1. smoothVTer

    Thread Starter Member

    Jan 13, 2010
    12
    0
  2. shortbus

    AAC Fanatic!

    Sep 30, 2009
    4,004
    1,522
    Yes. At the other gate voltages it's in the linear(resistive) area of operation. And will heat up more.
     
  3. crutschow

    Expert

    Mar 14, 2008
    12,988
    3,226
    The transistor is a logic-level device and is fully on at all the given gate voltages (it is not in the linear region). It just has slighter higher ON resistance at the lower Vgs values, as shown in the data sheet.
     
  4. smoothVTer

    Thread Starter Member

    Jan 13, 2010
    12
    0
    Hmmm OK. That makes sense, so its my application that is giving me pause:

    [​IMG]

    Q1 is a TIP32CTU PNP
    M1 is a STP75N75F4 power nMOS with Rds=0.0092ohm @ Vgs=10V
    DZ1 is an 18V zener


    DC generator charges a 10F cap bank rated to 20V. I want to limit the charge to 18V. Q1 turns on when Vin = 18V. I tuned R2, a 50k pot, so that the center wiper gets to 11V when Q1 turns on. I did this tuning without the gate of M1 connected.

    As soon as I connect M1, the voltage indeed gets clamped to 18V. However, all the power gets dissipated in M1 as heat rather then my large 1ohm chassis mount resistor. I wondered why ... so I continually crank the DC generator ... Vin never goes above 18V ... however now with M1's gate connected, I can only get 4.1V out of the voltage divider. M1 heats up like crazy, my 1ohm diversion load is cold to the touch. Its because I can't get M1's gate barely past threshold...

    This took me by surprise, because a mosfet gate isn't supposed to load the circuit down and draw current. At this point I have no idea why, when I connect the mosfet gate and crank full blast, I cannot get the voltage I did at R2's center wiper.

    Any ideas?
     
  5. ErnieM

    AAC Fanatic!

    Apr 24, 2011
    7,386
    1,605
    Your post is basically unreadable. Make a smaller picture and try again.

    Sounds like the MOSFET is dead Jim, dead. Gate punch thru to the gate is now connected to source and drain and will load down the drive.

    I would *never ever never* put 18 volts onto a MOSFET gate. It's too close to the max, and exceeding the max will... yep, punch thru the gate.
     
  6. smoothVTer

    Thread Starter Member

    Jan 13, 2010
    12
    0
    At first glance I would agree with you, ErnieM.

    However, Vgs has never exceeded 11V. The gate is being driven by the voltage divider on the collector of Q1, and I've tuned that to be 11V when Vin=18V at Q1 is on. Vgs max for M1 is 20V, so 11V is pretty much in the center of the range and 10V is what Rds is rated at, anyway.

    I've also had the same behavior from 3 of the same MOSFETs parts. Did this circuit configuration make me blow 3 MOSFETs? If so, I don't understand how I exceeded the 20V Vgs max.
     
  7. smoothVTer

    Thread Starter Member

    Jan 13, 2010
    12
    0
    I wanted to test the MOSFET in a stand-alone configuration to determine if the gate had indeed been damaged by punch-thru:

    [​IMG]

    I took the same MOSFET out of the original circuit, all the way at the top of this thread. I connected it instead to my benchtop 12V power supply as show in this new schematic. I start with Vgs=0V. The ammeter reads 0A, until about 3.8V when it starts to read ~3ma. 4V is Vth, so that's expected. I crank Vgs to 10V via the pot. Boom, I get 1.1A: pretty close to the expected current through my 10-ohm, 100W load ( not sure where that extra 0.9V or so went, probably dissipated in the fet )

    Anyhow, my test setup works pretty much as expected for this mosfet: Id is zero below Vth; Near Vth Id ~ 3ma; Id rapidly increases @ Vth (4V): Id seems to be at maximum possible at Vgs=10V.

    The behavior is indicative of a fully functional mosfet, is it not?

    What then, could cause Vgs to drop out in the original circuit at the top of this thread? The mystery is on! :cool:
     
  8. smoothVTer

    Thread Starter Member

    Jan 13, 2010
    12
    0
    [​IMG]

    I figured out what's going on, but I don't know what to do about it.

    To summarize: I wanted to use M1 to switch a diversion shunt to keep the cap bank <= 18V while the DC generator is still operating. This sensing is done by zener Dz1, Vz=18V. I tuned the circuit first without the gate connected so that the wiper on the pot goes to 10V when Vin=18V, turning on M1 fully.

    The problem is, as soon as I connected M1, I couldn't get Vgs above 4V. M1 was barely on, and dissipated all the power instead of my 1-ohm power resistor.

    What's happening is, that as soon as M1 switches on, due to the very low impedance now of 0.0092 ohms (Rds) + 1 ohms (resistor) , I basically have a short between the PNP's emitter and collector. In parallel with this low impedence, the PNP cannot sink the current required to establish Vgs=10V, as in my prior no-gate-connected setup.

    I figured this out after I switched the 1-ohm load to a 10-ohm load. The 10-ohm load allows M1 to stay fully-on longer ... but doesn't dump the excess current from the DC generator fast enough when Vin reaches 18V.

    Kind of stuck between a rock and a hard place.
     
Loading...