Novice trying to measure Qg of Mosfet

Discussion in 'The Projects Forum' started by deplora, Nov 12, 2013.

  1. deplora

    Thread Starter Member

    Jun 11, 2008
    10
    0
    Hi everyone,

    I have not touched electronics in quite a while so I am pretty rusty. So, I've taken upon myself to just try and measure out some of the characteristics of a common Mosfet.

    So far, I have been able to just get a mosfet plug it into a breadboard of sorts and was able to measure out Rdson, Vgth, and BVDSS, but I am having heck of a time trying to begin to measure Gate Charge.

    I've been reading everywhere..appnotes and datasheets, and it just left me confused and have so much information I just need someone to sort out my thoughts the right order for this puzzle to make sense to me.

    Attached is a test circuit that I found through one of the app notes online. I figiured it looked pretty simple to put together and test a mosfet. I just have a voltage supply on Vdrain and I have a voltage supply on Vgate. I don't understand why the image shows that I apply constant current on the gate? From what I read, the gate charge is dependant upon Vdd and Vgs. I have a o-scope as well, and i was hoping I could catch the Vgs Vs Gate Charge waveform as it shows on every datasheet. This is where I really don't know how. I understand some of the theory behind it, but I would love to have someone show me how to actually measure this or tell me a step process.

    I understand that the Fet charge has three stages. One which it slowly charges the input capacitance (Cgd ..which is Crss on datasheet) and after that it reaches a plateau called "Miller", after an x amount of time the voltage ramps up again. I remember that Q = CV, but in order to find this Q for a particular point in Vgs, I have to know the actual input capacitance. I also know that Q = current x time. If i can find how long it takes for capacitor to charge and considering that Ig is the current on the Gate then I can get Q and from there then I can apply it to Q=CV and find the input capacitance?

    My head is wrapped around so much stuff, I confuse myself. Anyone who can help sort me out on how to measure Coss, Ciss, Crss and total Qg?

    Much thanks.
     
  2. DickCappels

    Moderator

    Aug 21, 2008
    2,654
    632
  3. deplora

    Thread Starter Member

    Jun 11, 2008
    10
    0
    Thanks for the response!

    I took a look and I see a nice circuit and where to probe, I will check this on my bench. I do have some further questions.

    1) The App note uses IRF630 mosfet for all their examples. Here is the datasheet for that part.

    http://www.vishay.com/docs/91031/sihf630p.pdf

    Based upon the App note:
    Qgs = Qg - Qgd

    When I see the datasheet, Qgs = 7nC, but Qg = 43nC and Qgd = 23mC. How does Qgs = 7nC? Should it not be 20nC?

    2) I don't understand why they apply 1.5mA constant current on the gate.
     
    Last edited: Nov 12, 2013
  4. ronv

    AAC Fanatic!

    Nov 12, 2008
    3,290
    1,255
    I'm certainly no expert, but I think the gate charges don't add up because they change as the FET turns on.
    As to the current source it is to turn it on slowly enough and to charge the capacitances in a linear fashion as opposed to a voltage source.
     
  5. deplora

    Thread Starter Member

    Jun 11, 2008
    10
    0
    Thanks,

    I will then just take Qgs as Qgs = (t1 - t0)ig (based upon looking at the chart).

    Thanks for clarifying the current source purpose. That being said, why have a current source for Drain? If applying current just enough to slowly turn on the gate is what matters, why just not a voltage source on drain?
     
  6. ronv

    AAC Fanatic!

    Nov 12, 2008
    3,290
    1,255
    If you just had a resistor in the drain the RC time constant would effect the measurement. Although having said that you can see a close curve by just using a large resistor in the gate and one in the drain that won't dissipate to much power in the FET.
     
  7. deplora

    Thread Starter Member

    Jun 11, 2008
    10
    0
    I just connected it all together like on the image I attached on my first post. I got a 2n7000 Mosfet to play with because it is a relatively small voltage part.

    I set 15V for Vds, placed a probe on gate pin, and started sourcing little bit of current and slowly keep incrementing (went to 5mA) but I did not see anything really happen. :(

    So, maybe I'm forgetting something basic, anyone can guide me to how? Would be super.

    Thanks.
     
  8. deplora

    Thread Starter Member

    Jun 11, 2008
    10
    0
    Ah well, couldn't see anything on my scope because the probe I had was apparently not working right. I figured that in the least I should be seeing current drawn from Drain and things rolling. However, all I've done was to source voltage on the gate. I'll try sourcing current instead and see how that goes when I am able.
     
  9. ronv

    AAC Fanatic!

    Nov 12, 2008
    3,290
    1,255
    The 7002 may not be the best example because the charge is so low. But here is a simulation using one with an NC of ~1 and one with an NC of ~2. You can see the RC time constant instead of the nice straight line with the current source.
     
    • NC.png
      NC.png
      File size:
      132.5 KB
      Views:
      30
  10. deplora

    Thread Starter Member

    Jun 11, 2008
    10
    0
    Thanks Ronv :)

    Yes, I agree I need to find a beefier FET to be able to find how to measure gate charge. Last night it was tough to get it on the scope, but finally was able to, but upon measuring the amount of charge it was much larger than expected although still in the nC scale. I don't know if the capacitance of the probe affects it also.

    On that note, I do not have those resistors on mine. I'm just directly probing at the pins. Adding the resistance is what causes the time constant correct?

    That is a slick simulation program. When I went to school years back it was Pspice, but it was a pain to deal with then. Hope there is a free version or at least a trial version of this software to play around with :)

    I'm going to try again later on with another part and see how it goes. I really appreciate your input, it certainly made me revisit and check really old stuff and come up to speed again, though I've never done measuring this spec, figure it would be a good sort of a challenging start.
     
  11. deplora

    Thread Starter Member

    Jun 11, 2008
    10
    0
    Ok, I took a stab at it using a FET that had much higher Gate Charge stated on the datasheet. I used a FDU3N40. I didn't apply that much high voltage as shown on their plots. Although I am wondering as to why they test at ID= 3A when the absolute max is 2A? why they chose 3A?

    My circuit is pretty much applying direct Vg voltage of ~4V so it goes over the thershold voltage and turn on. I just apply a pulse. For voltage on the drain I just source it with voltage at 20V. The image scopes that I have here is what I got. At the very least I can see, at least what I hope it is, the gate charge waveform that is similar looking to that plotted on datasheets. I placed my cursor on what I think is the total Qg amount. I know Qgd is the amount of charge during the Miller plateau stage, and Qgs is the small rise prior to reaching Miller phase. I am calculating the amount of charge by multiplying the time with the current drawn from Vgs. So, I am measring the current that goes into Vgs when I apply voltage there.

    Based upon this, are my conclusions making sense? Someone who can say "oh, that's what you are looking for" or "no, that looks wrong" :) . I'm just doing this for learning purposes. I know there is a test circuit on the datasheet there, but for me adding another mosfet into the mix complicates my life. My test circuit looks very simple in comparison.

    thanks in advanced.
     
  12. ronv

    AAC Fanatic!

    Nov 12, 2008
    3,290
    1,255
    Try hooking it up like the circuit in post 9. then scope the gate.
     
Loading...