Stupid Transistor Question for MOSFET

Discussion in 'General Electronics Chat' started by automagp68, Jan 8, 2013.

  1. automagp68

    Thread Starter Member

    Nov 13, 2011
    81
    0
    Hi all i am having a mental lapse here with some basic transistor stuff

    I get on a regular BJT you apply voltage to the base and you get flow from collector to emitter

    I have no experience with FET stuff and I'm curious why i cant get a simple LED to light up with a MOSFET P channel

    I have 5 volts waiting on S

    I have a 100ohm resistor and an LED connected to the drain and then to ground.

    I put 5 volts to the gate and the led turn on. When i remove the 5V from the gate the LED stays on unless i touch the gate to ground. Is this normal? Im confused? I thought they worked the same way as a BJT did?


    Can someone just give me a quick 3 sentences on what im doing wrong with out getting to technical?

    Thanks
     
  2. panic mode

    Senior Member

    Oct 10, 2011
    1,320
    304
    problem is that you mistake MOSFET with BJTs. do not leave gate floating.
     
  3. automagp68

    Thread Starter Member

    Nov 13, 2011
    81
    0
    OK thats kinda what i was thinking

    If i use a pull down resistor to pull the gate to ground i get the following

    If i give the gate 5v the LED goes off

    If i remove the 5v from the gate and allow the pull down resistor to do its job the LED turns on

    Why does this seem backwards to me?

    Its a P channel so doesn't +5V on the gate turn it on? Not the other way around?
     
  4. Ron H

    AAC Fanatic!

    Apr 14, 2005
    7,050
    657
    That's normal, except you have the gate voltages backward. When you connect the gate to ground, the LED will light. When you remove the ground connection, the LED will stay on until you connect the gate to +5V. This is because the gate has capacitance, but the resistance is essentially infinite. Any charge you put on the gate will leak off very slowly, so if you charge the gate to 0V (-5V relative to the source), it will stay charged until you discharge it (0V relative to the source).
    A resistor from gate to source will keep the MOSFET off when the gate isn't connected to ground.
    BTW, unless you are using a logic level MOSFET, the gate (on a p-channel) needs to be 10V lower than the source to guarantee minimum on resistance, called Rds(on). Logic level devices are specified at Vgs of -5V, -3.3V, -2.5V, etc.
     
  5. automagp68

    Thread Starter Member

    Nov 13, 2011
    81
    0
    Interesting

    Can you explain a bit more on what you mean i have the gate voltages backwards?

    Im not really sure what you mean?

    It seems to work as expected with a Pull down resistor on the gate except the fact that it is backwards from the operation i wanted

    I wanted a +5v on the gate to turn it "on" In this cass it is turning it off So i guess i need an N channel?
     
  6. Ron H

    AAC Fanatic!

    Apr 14, 2005
    7,050
    657
    You said the transistor turns on when you put 5V on the gate, and turns off when you put 0V on the gate.
    This can't happen with a PMOS transistor. When the source is at 5V, 0V on the gate turns it on. 5V turns it off.
     
  7. automagp68

    Thread Starter Member

    Nov 13, 2011
    81
    0

    Did you see this part i posted?
    When i give the gate 5v + it turns off?

    Thats correct right?
     
  8. Ron H

    AAC Fanatic!

    Apr 14, 2005
    7,050
    657
    That's correct, but that wasn't what you said in your first post. I was responding to your first post when i said you had the gate voltages backwards.

    Tell us what you are trying to do.
    Are you just wanting to light an LED?
    Is your power supply going to be 5V?
    Where does the control signal come from?
     
  9. automagp68

    Thread Starter Member

    Nov 13, 2011
    81
    0
    Hey there, thanks for the replies

    Yea the first time i wrote what i was doing was wrong the second reply i made made me think pull down resistor and now it works

    Can you answer this for me?
    So a P Channel does what we were just talking about If i go to an N channel to get the operation mode i want i have to reverse the supple and drain right? And of course the main difference being that the N channel will turn on with +5 instead of turning off like the p i was just talking about. N channels are reversed drain and supply though also? Can you verify that?

    Im just wanting to control something that pulls a few amps using a micro controller.
    All low voltage.

    MC will control the gate of the transistor which will have 5 volt source with a load of say 2 amps
     
  10. Ron H

    AAC Fanatic!

    Apr 14, 2005
    7,050
    657
    If you don't mind switching the ground side of the load, you can use an n-channel. Source to ground, drain to low side of load, +5V to other side of load. Zero volts turns it off, +5V turns it on, IF you use a logic level MOSFET like IRLB8721PBF. Mouser has them for US $1.25. There may be cheaper logic-level devices available. This is just the first one I found with low Rds(on). If you choose another one, Rds(on) should be less than .025Ω, to avoid needing a heat sink.

    If you're driving it with a microcontroller, why are you concerned about the polarity of the gate drive? You should be able to program an output for active high or active low.
     
  11. sheldons

    Active Member

    Oct 26, 2011
    616
    101
    have a look at the output stage used here-it can easily be used for your application with a few changes(led series resistor etc...)
     
  12. kubeek

    AAC Fanatic!

    Sep 20, 2005
    4,670
    804
    This is wrong, try putting voltage on a base of transistor and it will instantly blow up, unless that voltage happens to be less than ~0.65V. BJTs are controled by current through the base-emitter PN junction, not voltage.
     
    cabraham likes this.
  13. automagp68

    Thread Starter Member

    Nov 13, 2011
    81
    0
    What?

    I have no idea what you are talking about?

    Then your trying to tell me this link is wrong? It does say current not voltage.


    http://people.usd.edu/~schieber/psyc770/transistors101.html



    If you want current on the base then clearly you need to complete the circuit from the base to --------- something.

    So how does current exist on the base without completion?



    Please explain?
     
    Last edited: Jan 10, 2013
  14. automagp68

    Thread Starter Member

    Nov 13, 2011
    81
    0
    Ive been doing some more reading on the difference of BJT and mosfet

    If you could maybe give a simplified version of how they are different i would be forever grateful
     
    Last edited: Jan 10, 2013
  15. crutschow

    Expert

    Mar 14, 2008
    13,009
    3,233
    The very simple difference is this:

    A MOSFET is strictly voltage controlled with the gate having a very high input resistance (it looks like a capacitor). The drain-source current is proportional to the gate-source voltage above the threshold voltage. It generally requires from several volts to ten volts to fully turn on as a switch (this is not the threshold voltage), depending upon whether it is a standard MOSFET or a logic-level type MOSFET.

    The BJT has a much lower input impedance with the the base-emitter junction looking like a forward-biased diode (thus typically requiring a series base resistor to limit the current). It requires a small voltage to turn on (approximately equal to a forward biased diode) with the collector emitter current proportional to a fraction the base-emitter current as determined by the transistor gain (Beta or Hfe). For full turn on as a switch (saturation) typically the base current should be at least 1/10 of the maximum collector current.
     
  16. bratsuzao

    New Member

    Jan 11, 2013
    5
    0
    Hey there, thanks for the replies[​IMG]
     
  17. automagp68

    Thread Starter Member

    Nov 13, 2011
    81
    0
    Interesting so let me now ask this

    on a mosfet can one intentionally limit the voltage allowed to pass by not applying "10 volts" on the gate? Say my source is 5 volts but i really want it to be 1.5

    Can i intentionally not apply all the voltage to the gate
     
  18. kubeek

    AAC Fanatic!

    Sep 20, 2005
    4,670
    804
    You can, but.. The transistor would be in linear mode where it acts as a resistor, not a strictly voltage drop. Other thing is that the Vth and thus the voltage needed to get just that resistance varies greatly, even in the same batch, so you will need some feedback mechanism to control the output voltage anyway.
     
  19. crutschow

    Expert

    Mar 14, 2008
    13,009
    3,233
    That is how a linear voltage regulator works. It uses negative feedback to adjust the gate voltage of the series pass transistor so that the output voltage stays at the desired value, independent of the load.
     
  20. cabraham

    Member

    Oct 29, 2011
    82
    30
    The answers given above are good ones. I would just add that if a standard level FET is driven from a g-s voltage of 5V instead of 10V, the resistance of the drain-source channel, "Rdson" is higher in value. At 10V Vgs, the Rdson may be 0.10 ohm, but at 5V Vgs, it may be 0.15 ohm. As stated above, if Vgs drops further, the FET may not stay in the Ohmic mode of operation, acting like a resistor. The voltage drop increases as does the power dissipated.

    Also, regarding the differences between bjt & FET, there are 2 not covered yet. A bjt is a "bi-polar" device. A FET is "uni-polar". The prefix "bi" indicates that it conducts via 2 polarities of semiconductor material, where as "uni" infers 1 polarity type. This leads us to another distinction. Since a unipolar device has current conduction in n-material only (or p), there is little stored charge except for capacitance. A bjt conducts current in n & p type material. As carriers cross the junction storage occurs in a depletion region.

    Before the signal polarity can reverse, this stored charge must be removed. FETs are called majority carrier devices. The bjt is minority carrier in nature. So when used as a switch, a FET can switch polarities very fast. A bjt when allowed to saturate, such as in common emitter configurations, will take a long time to come out of saturation.

    Books will elaborate but I thought this should be mentioned. BR.

    Claude
     
Loading...