1. Not finding help here? Sign up for a free 30min tutor trial with Chegg Tutors
    Dismiss Notice
Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

IRF510 Baising with LT-Spice where is my fault? please reply guys

  1. Jan 10, 2012 #1
    there is something strange in my lt-spice
    when i build my circuit mosfet with voltage divider + NO Rs (@Source Lead)
    i did not get the corresponding Id current WHY?

    my parameter:
    R1=R2=100K
    Rd=400
    Id=10 mA
    where i used IRF510 with Kn= gfs^2 / 4Id = (1.3)^2 / (4 * 3.4 ) = 124 mA/v2
    i want to see at least V0= 1 or more than this

    but unfortunately i see very small voltage on Vds about 200 mV

    please some one help me
     
    Last edited by a moderator: Jan 10, 2012
  2. jcsd
  3. Jan 10, 2012 #2
    Re: IRF510 Baising with LT-Spice where is my fault?plz reply guys

    Please supply the circuit diagram along with netlist and output files and detailed model parameters and specify what exactly is the problem.
     
  4. Jan 10, 2012 #3
    111m1pj.jpg

    2qwlb43.jpg
     
  5. Jan 10, 2012 #4

    vk6kro

    User Avatar
    Science Advisor

    Doesn't that just mean the FET is saturated?

    You could reduce the size of the lower 100 K to 20 K and see if that fixes it.
     
  6. Jan 10, 2012 #5
    god help mee;;;;
    it did not work on lt-spice
    what should i do
    please try to build it on lt-spice to loook at by your self
     
    Last edited by a moderator: Jan 10, 2012
  7. Jan 10, 2012 #6

    vk6kro

    User Avatar
    Science Advisor

    It might be better to fix it yourself.

    Just vary the input voltage and observe the drain voltage.

    At some voltage between 2 volts and 5 volts, the FET should turn on and the drain voltage should drop.

    With LTSpice, there is a feature for varying a power source.
    So, if you put a voltage source across the input and vary it from 0 to 10 volts, you will see where the FET starts to conduct.
     
  8. Jan 10, 2012 #7

    vk6kro

    User Avatar
    Science Advisor

    I tried it in LTSpice using an IRF510 model, and the FET switches on at about 3.9 volts.
     
  9. Jan 10, 2012 #8
    but does not like i calc v0=5
    what about UId does its same result?
     
  10. Jan 10, 2012 #9
    why i observe all this wrongs done when i removed Rs from the source?
    i had read more books on oscillators design all of them work as i calc
    but on mosfet i had failed....nevertheless i success design my oscillator on MOSFET only an just only when i build it with Rs (regenerative resistor) .. i need to writing something useful in my book that describe the problem i have encountered .. thats why i'm interesting in understand the correct calculations to give a good analysis in using mosfet in VCO circuit
     
  11. Jan 10, 2012 #10

    vk6kro

    User Avatar
    Science Advisor

    There are variations between actual samples of FETs and what you get in LTSpice depends on the model that you are using.

    You are not trying the input voltage variation I suggested and yet this is the only way to find the switching voltage. The FET will not do anything useful until you get the bias right.

    The 400 ohm load resistor is too high for this FET and it results in a very critical transistion from off to on.

    Having a source resistor makes the FET less critical to set up.
     
  12. Jan 10, 2012 #11
    it is OK!
    i wish if you supply me with the correct method on how to find the value of Kn because the datasheet not mentioned for this value ; all i know is to calculate this value like this below:
    Kn= (gfs)^2 / (4 * Id)
    are there any other way to extract the accurate value of Kn from datasheet ??
    datasheet not mentioned Id(on) else...
     
    Last edited: Jan 10, 2012
  13. Jan 10, 2012 #12

    vk6kro

    User Avatar
    Science Advisor

    When you give a formula, you need to state what the variable symbols mean.

    Until you do that, anyone would just be guessing what you mean.

    However, formulas giving gain etc of transistors and FETs always assume that the device is acting in a linear mode. Your FET is biased fully on and is saturated, so any formulas that assume linear mode will not be accurate.

    I hope you noticed the comment by Berkeman about using TEXT SPEAK.
    It is a rule here that you use complete words and proper spelling and proper sentences.
    For example using "U" instead of "YOU" is not acceptable.
    We realise that English is not your first language and the sentences may be difficult, but the text speak is something you can control.
     
  14. Jan 10, 2012 #13
    Dear vk6kro;
    think you very much i will try to use the full sentences but can not promise if i forget this restrict so be patient with me i am an Arabian man and you know how much Arabs go to worst in learning ;
    my problem started when i was in my lab then i had build my circuit of colpitts for BJT but when i migrate to use MOSFET not JFET to achieve colpitts i successes to design it but i asked my self why not remove Re of the source leg and recalculate my values again to give the circuit 30mA with 5V output ...then suddenly i got what i do not like to guess...after that i decide to finding another way to get the correct output current of the drain using the formula which is known to us. i think the problem came from that I can not extracting Kn (constant of MOSFET) from datasheet ... why manufacturer not included in DATASHEET? series of problems i have ... so please if you can help me or give do not stop to help.
     
  15. Jan 10, 2012 #14

    vk6kro

    User Avatar
    Science Advisor

    Again, you need to say what Kn is. Maybe I should know, but please define it.

    What does this value affect?
     
  16. Jan 10, 2012 #15
    Kn is the constant of Mosfet that is used in calculating the current of the drain
    as this formula give: Id= Kn (Vgs - VgsThreshold )^2 where id = drain current
    Can you give me an example of circuit using mosfet with voltage divider and without putting Rs (resistor on the source leg) which give me 3 volt and 10 mA ? if you do this you answer all my questions
     
  17. Jan 10, 2012 #16

    vk6kro

    User Avatar
    Science Advisor

    If you have a 9 volt supply, then the top resistor in your voltage divider could be 100 K and the bottom one 50 K.

    So, the voltage will be ( 50/ 150 times 9 volts) or 3 volts.

    If you used 80 K as the bottom resistor in your voltage divider and 100 K as the top resistor, the voltage will be ( 80 / 180 * 9 volts ) or 4 volts.

    This is very close to the right bias.

    You could make the drain resistor 50 ohms, too.
     
Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook