9 Ways to drive a MOSFET

Поділитися
Вставка
  • Опубліковано 7 лис 2024

КОМЕНТАРІ • 57

  • @JackBarakitis
    @JackBarakitis 3 місяці тому +3

    Excellent video, thank you!

  • @PhG1961
    @PhG1961 5 місяців тому

    Waw, what a great video and at the same time your channel/site!

  • @khimroy3958
    @khimroy3958 3 місяці тому +6

    If two switches press it will short circuit the power supply ! Be careful !

    • @Dazzwidd
      @Dazzwidd 3 місяці тому +1

      Boom 💥

    • @LeeOwen-g9h
      @LeeOwen-g9h 15 годин тому

      Letting the MOSFET Gate float will also go BOOM 💥

  • @GermanMythbuster
    @GermanMythbuster Місяць тому

    You missed direct drive with diode across the resistor to get faster off switching :)

  • @toma.3d
    @toma.3d 3 місяці тому +1

    wrong schematics for floating gate driver and opto-coupler , sources connected to ground shorting the mosfet when on.

  • @fredheilema8683
    @fredheilema8683 5 місяців тому +2

    In the inverted level shifter circuit schema are NPN transistors drawn. They should be pnp transistors. In the comment you speak of pnp transistors which is correct.

    • @ErikBongers
      @ErikBongers 3 місяці тому

      Thank you! The drawing didn't make sense.

    • @RyCorp77
      @RyCorp77 3 місяці тому

      if you pressed both buttons at the same time on the inverted wouldn't the 10v run through both transistors directly to ground?

    • @foogod4237
      @foogod4237 3 місяці тому

      ​@@RyCorp77 Yes! Well done for noticing that! This is actually a very big caveat with these sorts of "push-pull" drive designs. And this can actually still be a problem even if both signals aren't technically on at the same time. Transistors don't switch instantly, so if a circuit like this goes from "on signal" to "off signal" (or vice-versa) very fast (for example, if the signal lines are driven by some other switching circuit which is always either "on" or "off"), then there will be a period of time where the "on" drive transistor has not fully turned off yet, but the "off" drive transistor is already (at least partially) on, so you will get current flowing directly from VCC to ground straight through the two drive transistors, which can actually produce a substantial amount of losses and heat, and can even damage them if they're driven too fast, etc.
      This is why lots of modern dual-output gate driver ICs are built with sophisticated internals which are designed to make sure the two transistors are never driven at the same time, and give one transistor time to fully turn off before turning on the other one, etc. (Yet another of the many many reasons why gate-driver ICs were invented and are a good idea to use.)

  • @frankdearr2772
    @frankdearr2772 4 місяці тому

    great topic, thanks

  • @matanlj
    @matanlj 5 місяців тому

    Very informative!

  • @kabandajamir9844
    @kabandajamir9844 Рік тому +2

    The world's best teacher thanks

  • @geoffgeoff143
    @geoffgeoff143 11 місяців тому +3

    Is a Mosfet like a Masserati?

  • @isacdaimary9908
    @isacdaimary9908 Рік тому +1

    I wish I got your videos a year earlier. Thank you so much Sir

  • @PeetHobby
    @PeetHobby 9 місяців тому

    Great video! 👍

  • @nissaarwz4645
    @nissaarwz4645 3 місяці тому

    If i have a build a "driver circuit" to turn on or off an automotive coil (use 5v pwm signal to turn on) , which method would you rather use?

    • @krakkus
      @krakkus  3 місяці тому +1

      Hi, I would either go direct drive on the low side if acceptable, or dedicated high side driver IC. In both cases I would want a low component count for a project that doesn't live on my workbench.

    • @nissaarwz4645
      @nissaarwz4645 2 місяці тому

      @@krakkus Thanks Sir ! The coil is activated with + 5v , it means i need a low side mofset so that i can a positive signal , right?

  • @ccv3237
    @ccv3237 Рік тому

    Great information, and diagrams. Thankyou much.

  • @vidtechnology7342
    @vidtechnology7342 5 місяців тому

    does bootstrapping limits switching frequency ?

  • @andrew2004sydney
    @andrew2004sydney 3 місяці тому

    Number 6 and Number 7 will not work. Both sides of the lamp are connected to ground. There is no possibility of a voltage across the lamp.

  • @MANINDERSINGH-bw4ms
    @MANINDERSINGH-bw4ms Рік тому

    Thanks fiŕst.sir is it possible that we can isolate drive circuit from mosfet,because when mosfet damaged or shorted(usually we face in repairing circuits) it damages drive ic .can it be seprated through capacitor or by other means?.thanks.secòndly what is best trap for overvoltage zener(sometimes it get open and circuit damages) or crowbar or any suitable provision.thanks

    • @foogod4237
      @foogod4237 3 місяці тому

      A lot of gate driver ICs have protection built in to detect this sort of situation and automatically shutdown to prevent damage to the driver IC or the rest of the circuit (which is another good reason to use a driver IC when you can). Look for driver ICs which list "fault detection" or "overcurrent detection" as a feature.

  • @foogod4237
    @foogod4237 3 місяці тому +2

    There are high-side driver ICs which can be "always on" (and manual control) just fine. That's just a limitation of that particular chip, not of the category in general. (and actually that chip can be controlled by manual signals too, if you wire it up correctly. It has nothing to do with being a "safety feature")
    Also, the optocoupler design can do PWM just fine, if you choose optocouplers with appropriate switching characteristics. A better approach would just be to use one of the other driver circuits and isolate the input using a single optocoupler instead, though.

  • @pamukpasa
    @pamukpasa 2 місяці тому

    Well done putting in efforts to make such a beautiful video however some of your circuit plans won't work according to the scheme you have shown.You can check them.Good luck

  • @ArcanePath360
    @ArcanePath360 8 місяців тому +1

    I'm so confused. I bought a logic level N channel FET and it drives from the Arduino when connected to 5v VCC. For some reason my pin outputs, no matter what digital pin I use, don't output enough voltage (or current) to keep the thing on. I checked my code is set to PINMODE output so I don't know what else to do.

    • @yeahaddigirl
      @yeahaddigirl 6 місяців тому

      Possibly pulls too much amperage from Arduino pin, or possibly not enough. A resistor between the two could help. It could limit the amperage from the Arduino to protect it, and could pull a little more in the case where FET needs a certain current to activate instead of just voltage level.

    • @ArcanePath360
      @ArcanePath360 6 місяців тому

      @@yeahaddigirl Thanks. I changed the FET to an older one I had which I thought was knackered, and it works much better. It's weird because they both have the same info on the data sheet about the logic level and vcc that drives it. It's not perfect, and adding a resistor is a must now, where before it didn't seem to do anything. I now have a new problem though.... when the motor's torque is under strain (like if you force against it) the Arduino resets. I've added a diode and a 10mf cap to the motor which seemed to help give it more oomf, but every time the motor stalls it resets the Arduino.

    • @RexxSchneider
      @RexxSchneider 5 місяців тому +1

      @@ArcanePath360 You always need a current limiting resistor when driving the gate of a power mosfet. Usually something around 100R will do the job. You can test your first mosfet by simply connecting the 100R to the 5V rail and then to ground. It should turn the mosfet on and off. If it doesn't, then you have a problem with your logic-level mosfet. If it does, then you need to check the output voltage of your Arduino.
      When a motor stalls it draws a large current which will drop the supply rail and cause resets. You should really use separate power lines for ics and motors.

    • @ArcanePath360
      @ArcanePath360 5 місяців тому

      @@RexxSchneider I think the problem is the cheapo ebay Arduino doesn't output enough current from the pins. All tests would indicate this, since I even tried 2 motor shields that were rated for the current but were very lacking in power IRL, no matter how I programmed them. This is why I switched to a FET, as I thought the problem was the motor shields, but I get the same underwhelming power. At least I did until I switched to a different FET. Problem I now get is that every now and then, when it's under strain the Arduino resets itself. I've put a 100mf cap on the motor and a diode which helped, but maybe I need another cap on the power input of the Arduino. The motor runs off the same power brick but with a buck/boost in front of it.

    • @foogod4237
      @foogod4237 3 місяці тому +1

      ​@@ArcanePath360 voltage ratings aren't the only things that can matter in this sort of situation. FETs are basically sorta like capacitors internally, which means that when you first turn them on, they will "suck up" a lot of current to charge up to their "on" state (and then they won't need much to maintain that state). Different FETs can have very different internal capacitances, which means they can have very different switching currents. These switching currents are very brief, but they can actually be pretty large if there's no resistors or anything in the way to limit them, which is why you should never drive a FET directly from something like an arduino without a current-limiting resistor. I suspect the first FET had a somewhat larger capacitance than the second one did, and was pulling more current when switching. However, even the second one is probably borderline and could potentially stop working (or even damage the arduino) if run directly (without a resistor) for a long time (this could also be why the arduino is resetting sometimes).
      It is also quite possible that switching the motor is sending noise, spikes or voltage dips back through the power system and causing the arduino to glitch or brown out. I'd suggest trying running the motor off of a completely separate power source than the arduino, and see if that makes things more stable. If so, you could try putting them back on the same source, but adding some much larger capacitors and maybe some inductors (chokes) to filter out noise/spikes between the two. If you have one, monitoring the arduino's supply voltage line using an oscilloscope would be a very good idea, as it can give you a much better idea what type of problem you may be dealing with.
      As a matter of principle, I always recommend adding at least a 10uF cap to the power input of any digital device like an arduino, regardless of what you're doing with it. With something big like a large motor in the mix, though, you might even want to go 100uF or bigger, though.
      Also, if the motor shields were not producing the amount of drive power to the motor you expected, that is *not* a problem with the arduino. That is either a problem with the motor shields, or (more likely) a problem with your power supply not being able to keep up with the demands (and the voltage to the motor dropping under load).

  • @yousefalhelabi8601
    @yousefalhelabi8601 9 місяців тому

    The circuit at 1:50 works well with me using IRZ44N MOSFET and supplying a PWM from the Arduino.

  • @wahyu_zx
    @wahyu_zx 11 місяців тому

    what transistor can i use for pushpull driver for esp32 with 3.3v logic level

  • @xfxox
    @xfxox 10 місяців тому +1

    7. isnt it shortcircuited ?

    • @foogod4237
      @foogod4237 3 місяці тому

      Yes, the diagram for #7 is wrong. The "GND" the switches are connected to must be a *different* GND than the one the load is connected to, which means you need two separate, isolated supplies (the switch supply cannot use the same ground as the load supply does). This is part of the reason almost nobody ever does things this way.

  • @Dc_tech386
    @Dc_tech386 Рік тому

    Question the high side switching is very difficult to drive with high voltage of the voltage is 160vdc to the higher side mosfets the gate can handle voltage higher the 20volt so how can the gate get change higher then the supply voltage of 160v to work explain this I use driver transformer to drive the low side switching but the mosfet get hot at 20v to get still this how can I drive 160vdc in a high switching or it’s not possible

    • @foogod4237
      @foogod4237 3 місяці тому

      This is why high-side switching is less commonly used than low-side switching. Depending on your switching requirements, there are high-side gate driver ICs that can do this for you, or you can use something like a charge pump circuit to generate a voltage higher than your high-side voltage for switching the MOSFET.
      Alternately, if you use a P-channel MOSFET on the high-side, then it will turn on when the gate voltage goes _below_ the source, so the operation will be inverted, but you will not need to have voltages higher than your supply voltage to drive it.
      Or you can just stick with low-side switching, which is generally easier anyway and works fine for lots of things, unless there's some specific reason that won't work for you (how hot the MOSFET gets will be exactly the same, regardless of whether it's on the high-side or low-side).

    • @Dc_tech386
      @Dc_tech386 3 місяці тому

      @foogod4237 thank for your feedback but I have bing to alot if classes on integrate circuit and I have fully master the ability to driver high side mosfet and igbt without using gate driver ic I wanted to use transformer instead and I got it to work perfectly 👌 without generating any heat what I learn from the graft machine that show the fully functional gate charge at every level of current to hand large load .the machine cost me alot of money to buy but this machine don't sell on the market ,i was lockly to get it from a company that was closing down what I found out is all mosfet use a current of 900 milliamp at full load some take up to 1 Amp of continuous current so when I hear alot of so call engineer on UA-cam said mosfet is a voltage Driven transistor I just laugh now because they don't have such a machine to measure the mosfet performance and show the data live I test a mosfet of 500 watt with a gate driver ic ir2110 at fully load the mosfet get so hot really fast I have to stop the testing and when I use a gate driver transformer that I design that provided a peak current of 12 Amp and 2 Amp continuous current with the same load the mosfet just get really warm but not hot and the current at the gate charges was taking continuous was 1 Amp so you see all the lies about mosfet is a voltage driven transistor is all garbage

  • @shafikgeorgeskhouzam5810
    @shafikgeorgeskhouzam5810 6 місяців тому

    Error in the floating gate driver at 5:52, the lamp is always off

    • @foogod4237
      @foogod4237 3 місяці тому

      The diagram is drawn wrong. He mentions in the audio (but not in the schematic) that it requires an independent power supply for the switching, which means that the "GND" that the switches are connected to is _not the same GND_ as the one the load is connected to. For this to work, there should be a "switching VCC" (+5V) and "switching GND", and a completely separate (unconnected) "load VCC" (+10V) and "load GND". (the "switching GND" is then connected to the MOSFET source pin, so "switching GND" and "load GND" will always be different.)
      But this means you need two completely independent power supplies, which is usually a pain, so this technique is actually not that commonly used...

  • @kennmossman8701
    @kennmossman8701 2 місяці тому

    i can ride a bike, fly an airplane, sail a boat but can not drive a mosfet

  • @swedensy
    @swedensy 4 місяці тому

    Its Ground and not Grounds

  • @quevineuxcrougniard2985
    @quevineuxcrougniard2985 10 місяців тому

    Sa démonstration est très intéressante et très bien faite mais il nous agace avec son accent auvergnat, le gugusse ! On ne dit pas : "moche Fête" mais on dit : "MOS FET". Même chose pour le : "pouche poul".

  • @manfreds.5242
    @manfreds.5242 Рік тому

    schorsch, proosch. Nice spelling.

  • @Lightrunner.
    @Lightrunner. Рік тому +6

    02:47 your schematic is wrong.

    • @RexxSchneider
      @RexxSchneider 5 місяців тому +1

      I was about to say the same thing. Neither BJT will turn on whatever you do with the switches.