Ask Hackaday: Dude, Where’s My MOSFET?

(Bipolar Junction) Transistors versus MOSFETs: both have their obvious niches. FETs are great for relatively high power applications because they have such a low on-resistance, but transistors are often easier to drive from low voltage microcontrollers because all they require is a current. It’s uncanny, though, how often we find ourselves in the middle between these extremes. What we’d really love is a part that has the virtues of both.

The ask in today’s Ask Hackaday is for your favorite part that fills a particular gap: a MOSFET device that’s able to move a handful of amps of low-voltage current without losing too much to heat, that is still drivable from a 3.3 V microcontroller, with bonus points for PWM ability at a frequency above human hearing. Imagine driving a moderately robust small DC robot motor forwards with a microcontroller, all running on a LiPo — a simple application that doesn’t need a full motor driver IC, but requires a high-efficiency, moderate current, and low-voltage-logic compatible transistor. If you’ve been here and done that, what did you use?

Bipolars

Years ago, the obvious answer to this dilemma would be TIP120 or similar bipolar junction transistor (BJT) — and a lot more batteries. The beauty of old-school Darlington transistors in low-voltage circuits is that the microcontroller only needs to produce a small current to push relatively large currents on the business end. With BJTs, as long as you can get over the base-emitter junction voltage (typically under one or two volts) you just pick the right base resistor and you’re set. This is in contrast to FETs of the day which require a given voltage to pass a current through them. Gate voltages for the big FETs are optimized for the 4-5 V range which is lousy if you all you have is a LiPo battery.

TIP122/127 H-Bridge: Easy to Build, but a Battery Hog
TIP122/127 H-Bridge: Easy to Build, Battery Hog

While the power Darlington is easy to drive, it has a few drawbacks. First is the voltage drop through the device when it’s conducting. Drop one or two volts on the transistor and you’ve pretty quickly got a few watts of power going to waste and a hot chip. And that’s assuming that you’ve got the voltage drop to spare — a volt or two off of the 3.6 V on a LiPo battery pack is a serious loss.

With apologies to [Adam Fabio], the BJT is off the list here. It’s easy to drive at low voltages, so it would work, but it won’t work well because of stupid quantum mechanics.

FETs

MOSFETs should be great for driving small motors, on paper. They have incredibly low on-resistances, easily in the milliohms, and they can turn on and off fast enough that the PWM will be efficient and noiseless. The flaw is that garden-variety power MOSFETS, for driving big loads, tend to have similarly large gate threshold voltages, which is a showstopper for low-voltage circuits. What can we do?

If the motor were being driven by a higher-voltage source, and you were switching the MOSFET on the low side, then you can use the motor’s power supply to drive the MOSFET, switching it on and off with whatever is handy — a small-signal BJT is just about perfect here. That’s the classic solution, illustrated here. As long as the motor voltage is high enough to fully open the MOSFET, you can just use that for the switching voltage.

In the actual application that spurred this column, I wanted to use a LiPo cell for the motor and the logic, but I ended up doing something ridiculous. I started off with a go-to MOSFET from my 5 V logic days, the IRF530, but it barely turns on at 3.3 V. So I cobbled on a 9 V battery to provide the switching voltage — purely to drive the MOSFET into full conduction. This 9 V “high” voltage is switched by a 2N2222 small-signal BJT and seems to do the job just fine. It works, but it’s a horrible hack; I wanted to drive everything off the LiPo, and failed.

Other Options?

Big power MOSFETs, in addition to having a higher gate voltage, also have some capacitance that needs to be overcome to turn them on and off. Between the fully-on and fully-off states, they get hot, so it’s important to push enough current into the gate fast enough that they transition quickly. Thus, big power MOSFET circuits use a gate driver circuit to drive them. A low-voltage gate driver, paired with my IRF530, would certainly be an option here. But all this just for a medium-sized DC motor? Seems like overkill.

7307Once we embrace complexity, there are small H-bridge and push-pull driver ICs that might fit the bill, and they’ve naturally got MOSFETs inside. Now that I think about it, I’ve built small-motor H-bridges from N/P complementary pair MOSFET chips in the past, and they work for low voltages. Somewhere in my pile I have some IRF7307s that will just barely do the job. I’d be ignoring one of the two paired FETs, but who cares?

Taking the next step in IC complexity, the various stepper-motor driver ICs can usually push and pull an amp or two, and operate on low voltages. You could conceivably drive a DC motor off of one phase of a stepper controller, but that just seems wasteful. But something like (half of) a TB6612 would work.

On the other hand, the fact that these various gate-driver, H-bridge, and stepper controller ICs can handle the currents I want with low logic voltage thresholds suggests that there should be at least a few monolithic, and cheaper, MOSFETs that can switch a few amps around on low voltages. Where are they hiding?

The Ask

So what would you do when you need to push up to two amps DC in one direction at LiPo battery voltages, with low loss, driven (potentially by PWM) from a 3.3 V microcontroller? Feel free to take this as a guideline, and deviate wherever you’d like from the spec if it brings up an interesting solution.

Whatever you do, don’t give me current figures out of a datasheet headline that are based on microsecond pulses, only to find out that it’s outside of the part’s DC safe operating area. I’ve been down that road before! It never ceases to amaze me how they design parts that are rated for 100 A at 10 microseconds that can only handle 300 mA steady state.

This has to be a common hacker use case. Does anyone have the MOSFET I’m looking for? Or do you all just use motor driver ICs or tack random 9 V batteries into your projects? (Ugh!)

[MOSFET tattoo image from Google image search; Make Yr Mom Sad on Tumblr (dead link)]

126 thoughts on “Ask Hackaday: Dude, Where’s My MOSFET?

          1. That’s a good ‘un. There are a couple of chips in SOT packages listed below that will work marginally at 1-2 A but might get hot. This is a big fat TO-220 that will provide a bit more headroom in terms of head-shedding, and it still turns on well enough at 3.0 V.

            This is a great choice. Thanks.

        1. The variant you want is an NTD6416ANL, sorry – left off too much of the alphabet soup part number. Looks like the version without the “L” has a higher threshold voltage. I’m used to working with 5 volt automotive processors, but the “L” version will work on 3.3 volts.

          1. Ah! Right on. That one is great. I think part of my problem in finding these parts is that I don’t know all of the various manufacturers’ alphabet soups well enough.

            Looking at the max safe operating area, this will push 10 A at 3 V DC without complaints. That’s more headroom than I need. Can’t complain about that!

    1. IRLZ44N is pretty common to find around here and is cheap.
      I also have some IRLB8743 I bought from China that are good for higher currents.

      Never tried 3.3V, but I use them with MCUs running at 5V with no problems.

    1. That is a boost circuit. And being from LT it is really expensive. 4,50 in 1pc and 3,- in 25pcs (digikey). The NXP PMV16XN Michael O’Brien mentioned is just 15ct in single pieces.

    1. threshold voltage is almost irrelevant it is spec’ed at 1mA. Rdson is only spec’ed for 10V and 4.5V so with a 3V uC you can only hope it works. and it obviously can’t handle 10A*30V=300W dissipation

    2. It’s the graphs (fig 5, Output characteristics) that really need to be looked at to figure out how much current you’ll be able to switch at what gate voltage. At 3V you can only switch 5A or so before the voltage drop starts getting too much.

    3. Nope! 20kHz is NOT above hearing range. I’m sick to the back teeth of switch mode power supplies that make a really uncomfortable whine because some designer has made that same stupid assumption. Most people under 25 can hear 20khz quite well. If you want your system to be inaudible operate at 30kHz or higher, please. My ears thank you

      1. It’s not a question of age or hearing range, most people above 25 can hear high frequencies just fine, there is no magical numeric “cutoff” like that.
        But what you usually hear are intermodulations and interferences or resoance frequencies.

  1. Hmm, how about you make a little inductive boost circuit to push an injection of charge into that gate? No need for complete SMPS, it should work discontinuously. You can control it with the micro (you’d need another spare port, though) with addition of another BJT. Once its gate is charged, the MOSFET can stay on for seconds or longer, so for your purpose it should suffice.

        1. Well, you could add a larger capacitor after your charge pump to store the charge and drive the gate… Depending on size it could keep it going for quite a while, since MOSFETs don’t need that much current. Mind you, I have no idea if this works.

          1. I test this at work, and it doesn’t work for very long. A bigger cap takes just as much longer to charge up as it buys you on the discharge side, so your minimum on-time requirement isn’t changed.

    1. No need for an inductor, use an oscilator and voltage doubler.
      Seen on the net somewhere a diagram for making a voltage doubler out of a 555 timer, capacitors and some diodes.
      doesn’t put out much current though.
      Oh and an amplifier in oscillation for more current….
      Someone in the comments around this site posted a link that lead me to those circuits.

        1. I was thinking of a DC-DC charge pump to raise the gate voltage of the MOSFET. Then to discharge said voltage would be to short the gate voltage to GND (assuming low-side switching).

          3v3 boosted to 6v6 will easily turn on a low-side MOSFET. (can’t remember if N-type or P-type, done some learning experiments to achieve on, off and a bonus heater mode LOL )

          The load is then wired via the drain-source channel for more power.
          Alternatively if the load needed more voltage to work properly then a boost SMPS could be used as the load’s VCC rail. With the down side is more current is drawn at the lower voltage to give enough current at the higher voltage for the load (wattage stays similar).

          With all the ideas so far the following rails could be produced:

          Vbat -> G-VCC -> MOSFET gates (G- for gates)
          Vbat -> L-VCC -> motors, WiFi, spinney flashy thing (L-oad)
          Vbat -> 3VCC -> unregulated or 3v3 regulated rail

          Also if the load voltage is around or below the gate voltage then the same setup can be used as high side switching.

          1. http://i.imgur.com/UGKu11m.png

            Next to the characteristic curves, this is the most important graph in a data sheet: the safe operating region.

            You really want to look at the DC line, but the 10s line is probably good enough for me. And that crosses the 3V DS at around 450 mA. Really not too shabby for the small package size, but not something you want moving 1A for a long while.

          2. That said, put 2 or 3 of these in parallel, and I’d be totally happy to run 1A through them all day. And they’re small enough that you could do so easily.

            (And if you’re curious: the graph that proves you can parallel them is here: if the resistance increases with temperature, more current will flow through the colder devices until they’re all equalized. As long as this curve slopes up, you’re good.)

            http://i.imgur.com/hRvIgvB.png

          3. Sorry to necrobump, but Elliot, you forget that in the on state Vds is low, not 3.3V. You can safely look at the peak of the DC line in that graph and the corresponding Vds willl give you voltage drop on the fet at that (high) current.

  2. I teach robotics and interfacing to students ranging from 8th grade to college. I find regularly that the static-sensitive MOSFETs die in handling. For that reason I tend to offer them TIP120s since they won’t be switching large loads AND we go through fewer parts in a class period. The MOSFETs just aren’t very robust. A science website I work with also promotes designs with MOSFETs and the teachers who buy those kits say the same thing.

    MOSFETs more efficient? Absolutely. Suitable for teaching labs? Not so much.

    1. You could make them work by soldering protective zeners to their gate and source terminals. I did that on a test batch; no casualties yet. Before that, I got away with sticking the MOSFETs and zeners in breadboards before class and telling my students not to move them around.

      1. While they’re hard to find, I HAVE seen MOSFETS offered with a gate protection zener diode inside the package. I think they’ve all been surface mount parts so a breakout PCB would be needed on a breadboard anyway. (and at that point just put a FET and zener on the pcb)

    2. We see damaged mosfet gates all the time. I’ve thought about making tiny clip-on shorting bars for to220’s that clamp the gate to drain/source, so you fully assemble the circuit and only then remove the shorting bar, for teaching kids electronics.

    3. If they do not switch large loads, I would use a high gain transistor like BC337-40 I_max 800mA and a current gain of around 400. But NO darlington, as the additional voltage drop is a heavy waste.

    1. This is another one that’s not really “on” until 4.5 V or so. It’ll work great for small currents, but it’ll get really hot and kill your battery life at 3.3 V.

      (I’m totally sure that it works great in your circuit, btw. We have different uses.)

    1. Only if necessary. For HS I normally use intgrated HS-switch, for LS only a FET, sometimes with external protection, if required. And of course I use LS if possible. Just to be cheap :-)

      1. I’m guessing deep water, to keep things light it’s common to design the electronics to cope with the water pressure then put it in a lightweight oil-filled container, rather than a pressure housing to keep the electronics at 1 atmosphere, which gets big and heavy at depth.
        It works well, but you have to be careful with component selection to avoid packages with air bubbles or other reasons why components might fail at pressure.

    1. sure, why not. What should be better? The semiconductors in plastic packages are your least concern in this case. Anything what has a void in it (metal package) will crush and any wet capacitor would get oil into it’s electrolyte. At least this was the idea at 1/3 of this pressure, but anything what stops working here will not magically start again at higher pressures :-)

  3. The intro paragraph doesn’t make any sense. I assume from the rest of the article it’s discussing BJTs versus MOSFETs. Saying “transistors versus MOSFETs” is like saying “vehicles versus cars”. One is a subset of the other.

    1. +1 This.

      It bugs me to no end when people make this ‘distinction’.

      What does the T stand for in MOSFET? And the T in BJT? What then is a transistor which can be compared against a MOSFET!?

      1. What’s my PIN number? I hate it when I loose my PIN number! (We all have our pet peeves.)

        I wrote BJT like 70,000 times, and snuck one “transistor” in maybe once just to ease the monotony. You try coming up with interesting synonyms for “bipolar junction transistor” to keep your prose peppy.

        1. This is not a pet peeve, this simply wanting a factual representation of a technical matter.

          “You try coming up with interesting synonyms for “bipolar junction transistor” to keep your prose peppy.”
          When writing about technical things, peppy prose is profoundly prohibited per peppy prose’s powerlessness in portraying pertinent principles. In other words, pep is fine when not misrepresenting facts. Using the wrong word is not a stylistic choice.

    2. +1

      This is the second time I’ve seen ‘transistor’ used to mean ‘BJTs but not FETs’. The first time was at an interview, and I was super confused after laying out a circuit full of MOSFETS, when the interviewer asked “great, but I asked you to do it with *transistors*, can you do that?”.

      I don’t understand how someone can expect the word transistor to not include Field Effect *Transistors*, the ones included by the millions in almost every consumer device. Strange usage.

      1. Probably historical, since BJTs were used more than *FETs initially, as reflected in terms such as RTL (resistor transistor logic) and TTL (transistor-transitor logic) logic, which all refer to BJTs.

    1. This is common just like people expanding MOS to Metal Oxide Silicon which is incorrect because it’s actually Metal Oxide Substrate. MOS can be made with Gallium Arsenide (GaAs) as well.

  4. I’m a fan of Infineon’s Intelligent Power Switch (IPS) line. They’re basically a MOSFET with integrated drive and protection circuitry that defends them against over voltage, over temp, over current, and under voltage. They come in both high side and low side version, from a couple of amps up to dozens. They take 3.3V or 5V logic input and tend to have an open drain fault output. Some of them have integrated current sensing. They’re basically magic indestructible MOSFETs.

    The only problem is that most of them have a lower voltage bound above 5V, so they don’t quite make the requirement here.

  5. I got a bag of MTA3055 Enhanced N-channel FETs from one of the surplus places and have successfully switched 12v 4-5A inductive loads using 5v Arduino-atmega logic with no intermediate drivers. I believe the “enhanced” part is important, but I can’t tell you what has been “enhanced”.

    Looking at the specs it seems the average gate threshold voltage is 3.0 and max is 4.0, so they might not work reliably with your 3.3v logic. It never occurred to me to look at the threshold since I am an old-skool 5v kinda guy. It appears that others have posted alternate parts with somewhat lower gate voltages, so I expect that there are a number of choices.

    Of course, using a FET driver chip is probably more in keeping with modernity and flexibility for higher drive voltages/currents.

    1. Enhancement mode v.s. depletion mode FET’s. Simply put Enhancement mode fets are “OFF” until you put a voltage on the gate, and depletion mode FET’s are “ON” until you put a voltage at the gate – kind of weird.

      Both can be N-Channel or P-channel.

  6. IRLML6244, 20Vds max, 27mR at 2.5Vgs. Tiny SOT23 thing so 100 degrees (Kelvin or Celsius ;) ) per Watt, better keep dissipation below 1W.

    6,3A max, will heat it 107 degrees above ambient, so at 25 degrees Celsius ambient you’ll burn you fingers but not the FET. Best keep it well below 5A and give it a big copper plane when switching large currents.

    For larger currents look at larger packages.

      1. Something in DPAK or SOP8?

        I think you are looking for a logic level mosfet with low Rds_on, but there is so many of them… I bought a few for experimenting with small, low power switchers, so shorter pulses.

        But, what currents are we talking about? A few 100mA shouldn’t be a problem at 27mR… Say 1A continuous, so 27mW, so 2,7degrees temperature rise.

        For higher power, maybe IRLR8743PbF? (Just looking at what I happen to have, it might not be the best for your application)

        For things like this I usually play around a bit with parametric selection tools at a supplier, then pick whatever is the cheapest that fits the purpose with room to spare.

        With mishaps with those small switchers I already cooked an inductor, and the ‘6244 was still fine… (Software mishap, FET stayed on, its a boost converter so that shorted the (4.2V) battery over the (100uH/0R7) inductor. Inductor overheated, insulation melted, and it’s now a low-inductance short). So a small motor’s stall current might not be too much of an issue for the IRLML6244 as long as it is below 5 / 6A. You’d risk burning the motor before burning the FET if it’s a small enough motor.

        (Measure stall current / locked rotor current, pick a fet that can handle it. Maybe even include a PTC fuse…)

      2. I know this is a seriously old thread, but I came across it in searching for _common_, low Rds(on) at low Vgs mosfets. Theres some great info here on readily available parts which is hard to identify in a parametric search. However this assertion about the safe operating area is incorrect and may be misleading people.

        The safe operating area chart is describing operation within a linear region, ie with the mosfet in varying states of ‘on’. What seems to be getting missed here is that it’s describing the mosfet’s state when on, which includes the Vds.

        If you are switching a 5V 4A load, with lets say for argument a 2.5V gate drive voltage, then when the mosfet is on, ie when Ids is 4A, then Vds is *not* 5V. It is roughly Ids * Rds, in this case 4A * 0.027ohms = 0.108V. This is still within the safe operating area.

        It is not at all limited to 100mA as implied here, and can in fact handle several amps without issue.

        Also, thats with a Tj of 150C, which is higher than it would reach with this load. 4A*0.108V = 0.432W, with an Rja of 100C/W, 0.432W * 100C/W means it would be approximately 43C above Ta (ambient).

        Hopefully this helps straighten things out for others who come across this thread in future.

    1. Done the calculation for the full on dissipation at 6.3Aand the answer is:
      Voltage across transistor:
      R=0.027 ohms
      I=6.3A
      V=IxR
      V= 0.1701 across the MOSFET unless the load is shorted
      Power (heat-loss) in transistor:
      P=IxV
      P=6.3×0.1701
      P=1.07163 watts of heat.

      The transition between on and off for that component is where things get tricky and maths-TL;DR. What is the peak current for how many mS/uS? that will give you a rough idea of the transition speed at the 6.3A loading if at all possible due to gate capacitance.
      Keeping the load at 5A or less would extend the life of the MOSFET (captain obvious moment)

      Also think of the rice grain sized mosfet turning on and off your WiFi card, SSD, display, etc… and the ones that switch low current battery voltage into very high current low voltage CPU and logic bus VCC. Tens of amps in rice sized switches (mostly high side switched)

      1. Totally good call. The VNS3 or VNS7 would fit the bill very nicely. PWM up to 50 kHz is a sweet bonus, as is over-temp, over-current limiting. They’re a bit more expensive than a straight FET, but they have a bunch more features. I’ll get some to try out. Thanks.

  7. An AO3400 or AO3401, usually. They’re SOT-23, cost about 5c each and can push an amp or two without getting warm.

    None of this TO-220 BS, it isn’t 1995 anymore. You don’t need big packages until you get well past 5A, and for that there’s the IRL series.

  8. “a MOSFET device that’s able to move a handful of amps of low-voltage current without losing too much to heat, that is still drivable from a 3.3 V microcontroller, with bonus points for PWM ability at a frequency above human hearing. I”

    Go bookmark and read:

    “Design And Application Guide
    For High Speed MOSFET Gate Drive Circuits
    By Laszlo Balogh ”

    Before writing “with bonus points for PWM ability at a frequency above human hearing. I””
    –as that has nothing to do with the MOSFET, but rather with the MOSFET driver.

    MOSFETs and IGBTs are used in particular power drive niches (think variable frequency drives for motor-machine tools ore 3 phase power generation from single phase AC power.) IGBT’s losses are linear with current. MOSFET’s increase with the square of current. IGBT’s are not so good above 20kHz. MOSFETs are.

    You like darlingtons?, try the Sziklai pair instead, it has a 0.7V turn on.

    “Loosing too much heat” would need to define the available heatsinking. Is there 4 square inches of 1oz copper available to a TO-262 package with 62.5 degrees C / Watt and convection cooling?

    Is your power source power limited? What happens when the motor’s rotor locks? Will the wires just burn?

    For motor driving in a real product that has actual safety requirements you might want to consider parts (Half / H-Bridges) that are meant for that purpose that also include things like short circuit protection, over voltage protection, built in diodes, over current protection, over temperature protection all with a built-in hystersis and feedback pin to identify the problem. ST, Infineon and ON Semi all make parts for this purpose. Think automotive.

    If you are building a home-built design consider an OMNIFET or self-protected “load switch” that will have built in ESD protection to its gate and drain. Regular MOSFETs are awfully static sensitive for such a noise application.

    All that said.

    Diodes Inc is the king of cheap ass MOSFETs.

  9. I seem to use a lot of DMG1012UW in my designs when I want to drive moderate loads from 1.8V logic with a small (SOT-323) package.
    They’re only a few cents each (in large quantities) at the usual online stockists.

    VDS max 20V
    VGS(th) 0.5V to 1.0V
    RDS (ON) 750mohm max at VGS = 1.8V, ID = 350mA

  10. I think MOSFETs are used a lot more because there much easier to design for. You can easily get MOSFETs that are specifically designed to be driven from a GPIO pin.

    Bipolar transistors don’t need to dissipate power in their on state. The voltage drop is a result of the Base-Emitter junction when the supply voltage is connected to the collector. You can use the opposite polarity transistor and supply voltage to the emitter – that’s how LDOs work. In this config you need another transistor to invert the control signal and drive the main current transistor.

    The biggest considerations come down to frequency and whether you want switching or linear control.

    MOSFETS aren’t good for high frequencies and they are not very ‘linear’ in there so called linear region.

    Bipolars are very ‘linear’ but they have a large capacitance between base and emitter that you need to bleed off for higher speed application. This is the trap that turns people off bipolar. If you want to use a bipolar as a switch any rate of toggling then you need to put a resistor from base to emitter to help turn it off or it will spend too much time in the linear region generating heat. This situation normally ends with a pop.

    Because of the need for a Base to Emitter resistor to turn off a bipolar, darlingtons are useless as a high toggle rate switch because you have no connection access to the base of the main transistor – *However* some darlingtons have this resistor built in.

    1. I don’t agree with most of your comment. MOSFETs gates are inherently capacitive and therefore have slew-rate issues regarding turn on and turn off.

      BJTs are indeed slow to turn off, but is easily improved if you don’t fully saturate them. Look up Baker Clamps.

      Also, why are you saying that BJTs don’t dissipate power when on? The two huge advantages MOSFETs provide are the lack of DC base current and Vce saturation voltage. Even when a BJT is fully on, its collector-emitter voltage drop is at minimum around 0.2 volts. At 2 amps, that’s 400mW, very significant for all but large package BJTs. Further, a high power BJT will have *low* hFE/Beta, requiring maybe 1/40h or 1/20th of the collector current through the base, multiplied by .7 to 1v, that is a significant loss in itself. MOSFETs, on the other hand, when fully enhanced have an RDSon characteristic that can reduce that drop to the millivolts range, and require no DC gate current.

      1. Vbe is always going to be 0v2 – 0v6 and that isn’t what I am talking about.

        In common collector config the highest Vb can be is Vss so Ve is going to be lower than Vss by Vbe to give Vce so you have a drop and power dissipation in your current path.

        In common emitter config the Vb just has to be lower than Vss by Vbe but Vss and Vc can be the same so you have Vce = 0 and no real dissipation like on common collector mode.

    2. Normally the MOSFET has the capacitance, although the BJT also has to be “turned off” hard to be fast in this case. Yes darlingtons have some drawbacks, I see not much reason to use them these days.

    1. Hah, read through *all* these comments, and noticing nary a mention of looking at everyday examples like switching-power-supplies/buck/boost-converters on old motherboards as references, until quite literally the very last comment.

      Ground well-covered in these comments.

  11. Question for those in the know:

    Why are many MOSFET’s advertised with huge current handling capabilities on TO-220?
    To pull a random example: IRLB3034.
    Can those pins ‘really’ handle 195A at all, let alone continuous?

    1. IRLB3034. Note 1:
      “Calcuted continuous current based on maximum allowable junction
      temperature Bond wire current limit is 195A. Note that current
      limitation arising from heating of the device leds may occur with
      some lead mounting arrangements.”

      They mention that:
      Id (silicon limited) is actually 343A
      Id (package limited) is 195A

      Rds (on resistance) is only 1.7m-ohm

    2. It is specmanship.

      If you look at the datasheets closely and read appnotes on MOSFETs, you’ll learn that those specs are only the case if the package temperature is locked at 25C. Thus, the assumption is you have a PERFECT, infinitely large heat sink attached.

      Translation. You’ll get no-where near that performance in the real world.

    3. Instantaneous current handling usually comes with a duty cycle rating. The limit is the bond wire for MOSFETs constructed with bondwires.

      TO-220 can be water cooled if necessary.

      TO-247 is a better package, it has more surface area.

      Look up Flip-Chip MOSFETs. They do not have bondwires, but rather directly connect their drain and source to their mounting points. They were designed for lower inductance and higher current handling, think motherboard VRM’s that have to power a 50 Amp 3.3V rail.

  12. BJTs are usually thought of as current controlled current sources/sinks, FETs as voltage controlled current sources/sinks. Still, with a BJT you’ll need 0.7-1V or 1.4-2V (as someone else mentioned, check the graphs) to reliably handle the current you require. This follows the typical diode IV curve. Still, high power BJTs tend to have low hFE (Ic/Ib), perhaps on the order of 20-40. In a darlington arrangement, this current may not come from your MCU but it is still present as loss and additional heat. I feel the article is misleading. There are minimum base voltage requirements with BJTs too, it just happens to be lower generally, and does not have the options in threshold voltage you find with FETs. All normal silicon BJTs have similar, typical, base voltage requirements.

    There are some options with logic level FETs, but as always there are trade-offs. You might be able to get by with a couple amps with a carefully chosen logic-level FET. Still, if unbridled power is your goal, it’s better to use a step-up supply and proper MOSFETs of an appropriate package size. This is a similar issue when you require a high-side switched FET. Generally, that means the gate voltage at your FET must be 10-15V higher than your greatest rail supply. There are multiple techniques to accomplish this, e.g. flying capacitor (<100% duty cycle), boost-up power supply, transformer isolation, etc.

  13. Any really significant power switching is going to require a significant gate drive at say 10-12V or so. Even if the gate-source threshold voltage is small enough for the microcontroller, running the gate right down at the threshold will give high resistance – not good for high power jobs.

    Also, how much current can the microcontroller pin supply? Adding a series resistor damps oscillation and also limits the gate inrush current to a safe level – but depending on the gate capacitance this limits the switching speed.

    For single-channel gate drivers, the Microchip MCP1402 is a neat little choice.

  14. A Cree Silicon Carbine mosfet. Cree, Inc. (Nasdaq: CREE) a market leader in silicon carbide (SiC) power products, has introduced its latest breakthrough in SiC power device technology: the industry’s first 900-V MOSFET platform. Optimized for high-frequency power-electronics applications, including renewable-energy inverters, electric-vehicle charging systems, and three-phase industrial power supplies, the new 900-V platform enables smaller and higher-efficiency next-generation power conversion systems at cost parity with silicon-based solutions.

      1. GaN is better. Capacitance near zero. But requires a different drive mechanism. If you want to drive it from a standard silicon source with a 0 reference then you have to run it Cascode with another MOSFET. IR makes that all in once package.

        SiC’s main benefit is extremely high temperature rating and simple drive mechanism. I don’t seem them as “cost-parity” with regular IGBT’s at all though.

Leave a Reply

Please be kind and respectful to help make the comments section excellent. (Comment Policy)

This site uses Akismet to reduce spam. Learn how your comment data is processed.