This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

CSD19503KCS: MOSFET rise time and fall time check

Part Number: CSD19503KCS
Other Parts Discussed in Thread: LM5116, , UCC27511

dear all,

i have made a circuit of buck with following specs,

Vin: 50V,

Vout: 12V, 30A

MOSFET used: CSD19503 as high side mosfet

Driver IC: LM5116,

Gate resistance: 2Ohm

Issue: when testing the schematic with above parameters, my MOSFET gate rise time and fall time are above 100ns.

Datasheet shows Rise time and fall time are 3ns and 2ns. How do i know if the datasheet shows the right parameters or they are just for "sales purpose".

What do i use to make these parameters possible in my schematic. Please suggest.

Vikas Dabas

India

  • Vikas,

    Switching times in a FET datasheet are very dependent upon the drive strength and test circuit, the rise and fall times are based on a specific circuit and test conditions indicated int he datasheet. As these conditions change so will the rise and fall times. There is a blog here that you may find useful:- e2e.ti.com/.../understanding-mosfet-data-sheets-part-5-switching-parameters

    I will ask our apps team if they can give an estimate of rise and fall time

    Chris
  • Dear Chris,

    After looking at the link provided, i have found two things:

    1. "They are so dependent on board and test conditions that one veteran in the FET industry (and personal mentor) often cites these as “the most useless parameters on the FET data sheet.”

    I think my view is matching to author's view.

    2. "TI includes these parameters as tested at the device’s rated current, while others will test these at only 1A ID, to give the appearance of a faster switching device."

    Please specify the test conditions for "CSD19503KCS" rise time and fall time, i.e. Gate drive current, Gate driver Rise time and fall time, the IC used for Gate drive or other Power source?

    This note doesn't says anything about how to find actual rise and fall time using real life conditions, rather given useful theory of current market.

    Can you please provide me some formula to calculate the actual rise and fall time of MOSFET using real parameters. it is most important as i need to choose between different MOSFET's.

    Thanks for reply.

    Help appreciated.

    Vikas Dabas

    India

  • Vikas,
    As I think is mentioned in the video that accompanies the blog, the better parameters to look at when comparing how relatively fast two devices will turn on and off would be the gate charge you are operating at.

    So for instance if you want to operate at 10Vgs, Qg for this part is is 28nC. So t(on) = Qg / Ig where Ig is the current of your gate driver.

    Of course Ig can vary depending on other system parameters, and is at least a function of the series gate resistance of the FET + driver, but all other things being equal, the smaller the gate charge, the faster the turn on.

    I will inquire to our characterization team as to the specific conditions the CSD19503KCS was tested at. 

  • Regarding the measurement specifics:

    The Tester for Switching times is a clamped Flyback topology. Components use are as follows:

    -Drain Inductor is 1uH/100A Drum Core.
    -Clamping Diode is a 10A/100V Schottky in D-Pak package.
    -Driver IC is UCC27511 4A turn ON/ 8A turn OFF or similar

    Rising and Fall switching times are measured on Vds waveform between 10% to 90% gating levels. Gate Delay and Propagation times are excluded.

    Test conditions are:
    -Vdd=40V supply Bus voltage
    -Ids=60A , PWM pulse set for Id peak
    -external Gate Rg=0 ohm
    -Vgs=10V Gate PWM amplitude
  • Hello,

    Data-sheet parameters for CSD19503KCS Mosfet are measured using an inductive Flyback circuit operated at Vdd=40V , Id=60A Peak and L=1uH/100A inductor.

    Very important in defining the Mosfet switching time constants  is the gate driver. For data-sheet characterization, 4A/8A peak UCC27511 driver was used.

    In your case, you're building a Buck converter using CSD19503KCS for HS device and LM5116 wide Vin Buck controller. Your application is essentially different from the Tester used for Mosfet characterization.

    Using Typical and Maximum worst case  data-sheet parameters for both CSD19503KCS and LM5116, we can compute:

    Vgs Rise /Fall Time Typ(nS) Max(nS)
    t1R 10.2 18
    t2R 12.83 22.57
    t3R 12.98 22.5
    t4R 14.8 26
    Total (nS) 50.81 89.07

    Typical Vgs commutation times would be 50.8nS and Maximum worst case would be 89nS.

    The external gate Rg used in the circuit, the LM5116 driver IC pull-up and pull-down impedances together with FET's internal gate Rg determine the switching commutation times calculated above.

    If you measure more than 100nS time, it means you have more than 2ohm gate Rg present in the circuit.

    Hope this helps answer your question.

    Best regards.

  • Dear Lucian,

    Lucian Hriscu said:


    Hello,

    Data-sheet parameters for CSD19503KCS Mosfet are measured using an inductive Flyback circuit operated at Vdd=40V , Id=60A Peak and L=1uH/100A inductor.

    Very important in defining the Mosfet switching time constants is the gate driver. For data-sheet characterization, 4A/8A peak UCC27511 driver was used.



    Please clear one thing, the rise time of driver is 9ns for 1.8nF, so how rise time of MOSFET is 3ns for 2.1nF Cgs?


    Lucian Hriscu said:


    ...... Using Typical and Maximum worst case data-sheet parameters for both CSD19503KCS and LM5116, we can compute:

    Vgs Rise /Fall Time Typ(nS) Max(nS)
    t1R 10.2 18
    t2R 12.83 22.57
    t3R 12.98 22.5
    t4R 14.8 26
    Total (nS) 50.81 89.07
    ...............


    Please tell me How to calculate these values, and what are these parameters? So i don't have to ask next time, it is critical, as many users face the same problem.

    It will help us choose device based on real time parameters.



    Vikas Dabas

    India
  • Vikas,
    I don't believe this is a trivial task and I think Lucian had to use Math Cad or something to that effect. If you email me directly, I can put you in contact with him.