FET switching application

Discussion in 'General Electronics Chat' started by baseball07, Aug 26, 2008.

  1. baseball07

    Thread Starter Active Member

    Apr 24, 2007
    39
    0
    Hello everyone. I have a 2N7000 FET which I am using as a switch to drive an LED from an input pulse train. However, when I input the pulse train, the LED does not switch on and off, it stays on. The datasheet of the 2N7000 says that its threshold voltage is .8 min, 2.1 typical, 3 max. So with an input of 1.5V I should be seeing the LED turn on and off. Why is it not working?

    When I attach a function generator of 5V square peaks to the input, the LED blinks on and off, so I know Vth is being reached and turning the LED on and off. Am I missing something?


    I attched a schematic of my circuit
     
    • FET.jpg
      FET.jpg
      File size:
      45.1 KB
      Views:
      45
  2. jpanhalt

    AAC Fanatic!

    Jan 18, 2008
    5,687
    900
    The threshold voltage is where it just starts to conduct. Look at Figure 2 of the datasheet.

    You need a higher gate voltage to handle the current, which is consistent with what you observed. John
     
  3. baseball07

    Thread Starter Active Member

    Apr 24, 2007
    39
    0
    Can I use a BJT, MPS05 NPN,to amplify the voltage? I know BJTs are a function of current, but can they be configured to amplify voltage?
     
  4. Audioguru

    New Member

    Dec 20, 2007
    9,411
    896
    I can't find an MPS05 but I remember an antique MPSU05 in an odd case.
    In your circuit, a red LED will draw 50mA to 70mA which is too high for most ordinary LEDs.
    If your 1.5V source can supply 4mA then a 2N3904 transistor can be used with a 200 ohm resistor in series with its base.
     
  5. thingmaker3

    Retired Moderator

    May 16, 2005
    5,072
    6
  6. baseball07

    Thread Starter Active Member

    Apr 24, 2007
    39
    0
    I just tried to see if I can amplify the voltage out of my device with the MPS A05 NPN transistor before attaching it to the FET but it doesn't seem to be working. My schematic is attached. When I input the 1.5V square pulses to the resistor in series with the base, and I probe the "out" in the schematic, it just reads 2.5V, the transistor didn't turn on. Then when I probe the input at the resistor which is supposed to be 1.5V pulses, the scope has square pulses that start at 0 and go down to -7V. What is happening?
     
  7. jpanhalt

    AAC Fanatic!

    Jan 18, 2008
    5,687
    900
    Put a resistor (R2) between the supply and collector, like this:

    View attachment 4256

    Try 100 ohms.

    John
     
  8. SgtWookie

    Expert

    Jul 17, 2007
    22,182
    1,728
    You do not have a resistor between the collector and Vcc or [eta] between the emitter and ground.
    As you have things wired now, if the transistor turned on, it would be shorting out the power supply - and likely to have it's magic smoke let out :eek:

    You need to provide a ground reference for your 1.5v square wave.
    Please see the attached. The first shows the currents, the 2nd shows the voltages.

    2.5v will not be enough to do a good job of turning on an average 2N7000. You should get the gate up to a Vgs of at least 4.5v.

    You may find some that will turn on reasonably well at a lower voltage, but you will have to screen a number of them on an individual basis.

    Similar to common BJTs (like regular NPN or PNP transistors), MOSFETS will display individual characteristics when you are operating them outside of "normal" operating conditions. As you will find out through experimentation, no two transistors are exactly alike.
     
    Last edited: Aug 27, 2008
  9. baseball07

    Thread Starter Active Member

    Apr 24, 2007
    39
    0
    ahh I forgot the resistor between collector and Vdd. I put a decade box there to see which resistance would be the best, and above 700 Ohm the output produced something that resembeled a square wave on the scope, but it was real fuzzy and unclear. I see that you used 47 Ohm in yuor simulation, but that seems to be too low for me. It seems theres still something wrong. Sgtwookie, you mentioned having a ground reference for my input pulses. How can I do that? There is just an output lead from my device that supplies the pulses.
     
  10. baseball07

    Thread Starter Active Member

    Apr 24, 2007
    39
    0
    Nevermind, I figured out why the NPN was not working, The grounds on my device and my circuit were not the same, I just had to connect them all on the same ground line. I have Vdd as 4V, so my 1.5V input gets amplified to 4V, I attached the screen shot of that. Now that I got that working I connected the amplifier to my FET and the result is shown in the next screen shot. Its weird because the pulses don't begin until about 2V but the peaks are very small. I attached a full schmatic of my circuit. Are there any resistor values I should change? Thank you
     
  11. SgtWookie

    Expert

    Jul 17, 2007
    22,182
    1,728
    Change the 10k resistor on the collector of NPN to under 200 Ohms.

    Replace the gate resistor of the 2N7000 with a piece of wire.

    I don't know what you have on the drain of the 2N7000, but perhaps it is an LED and a 10 Ohm resistor to limit current.

    If you bring the gate of the 2N7000 MOSFET up to 4.5v, the Rds(on) should be <5 Ohms.

    If you replace the 10 Ohm resistor and the diode/LED on the drain of the 2N7000 with a resistor of perhaps 30 Ohms or more, you should see a decent square wave on the drain.

    Be careful to not exceed the current limitations of your semiconductor devices. Datasheets and calculators are your friends.
     
  12. baseball07

    Thread Starter Active Member

    Apr 24, 2007
    39
    0
    Thanks Sgtwookie I will try that tomorrow. However I cannot remove that LED because that is the reason I am building this circuit, to drive this LED. Why would replacing the LED and 10 Ohm resistor with 30 Ohm be better? How can I mod it to keep the diode?
     
  13. SgtWookie

    Expert

    Jul 17, 2007
    22,182
    1,728
    OK, you didn't explain that.

    What current is the LED rated for, and at what Vf? (forward voltage)
    It's best to go by the LED manufacturers typical Vf @ current specification.
    For simplicity, I refer to the LED's current limiting resistor as Rlimit. You calculate it's value by:
    Rlimit = ( SupplyVoltage - VfLED ) / DesiredLEDCurrent
    So, your supply voltage is 4. Just for an example, I'll take a wild guess and say your VfLED=3 and your desired current is 25mA.
    Rlimit = (4 - 3) / 25mA
    Rlimit = 1 / .025
    Rlimit = 40 Ohms

    But, if you're not getting Vgs (that pesky gate voltage) up to 4.5v, the 2N7000 may be operating in linear (resistance) mode. With Vgs=4.5v, Rds is <5 Ohms, and typical is 1.8 Ohms. If Vgs is <4.5v, your mileage will vary, as you are outside the guaranteed specifications in the datasheet.

    [eta] If your LED is white, it will likely have a Vf from 3.4v to 4v. In that case, you will only see a small signal on the drain of the 2n7000.
     
    Last edited: Aug 28, 2008
  14. Bernard

    AAC Fanatic!

    Aug 7, 2008
    4,172
    397
    You said that the LED turns on , so you might have enough drive.If the drive source is somthing like an open collector there might be enough charge on the gate to keep the fet conducting.; then if so just need a ground return for the gate.
     
Loading...