Why Your MOSFET Transistors Get Hot

[@cossiedriverrs]

But were you driving them with pulse width modulation? I had a P-channel mosfet powering a windshield wiper motor. (My RDS on was 3 milliohms). With the gate at full negative potential, didn't even need a heatsink. As soon as I drove it with PWM, heating was a big issue. The gate current to charge the internal parasitic capacitance meant that there was a "ramp-up (and down). A possible 200mA (from a 555) was not enough to drive the gate. Further research showed that there can be AMPS of gate current to charge the capacitance. (in my case, around 13,500 pF). And it appears that mosfet driver chips are available to overcome the problem. I am going down that road now... So, were your "dodgy" mosfets actual getting ramped on as a result of a high gate capacitance and insufficient drive, or were all your tests at DC? Regards, and good luck.

[@jurgenkruger3932]

This is the reason why in a bridge-configuration all MOSFETs are usually N-Channel MOSFETs. Combined with a dedicated MOSFET driver with an internal bootstrap circuit or with a discrete solution, those bridges can work with a much better efficiency. Also, it's necessary to add some dead-time of several nano-seconds between both MOSFET gates, to avoid the "shoot-through" effect.

[@LewisLoflin]

They have some very good P-channel MOSFETs with an on resistance in milli ohms. I use a +side power switch to turn off Vcc during the switch time. You are correct about shoot through, but I never had that problem.

[@kayquexavier]

What can we do if our load requires a voltage lower than logic level? For example I want to control a 3.7v motor with Arduino, will a bjt more effective?

[@boonedockjourneyman7979]

When you simply ramble a bit and share your real world experience you make significant contributions to my students. More than one has done the "see, I told you" thing. It is truly nice to have some unscripted stuff on KZbin. Yes, yes, the well scripted has its place too. But don't stop just talking about how things really happen.

[@Dc_tech386]

Am having heating is with my pwm the mosfet getting hot in 1 hour of turning on at 6 amp the capacitor valve is correct so am thinking the new components or manufacturing is bad all of my mosfet burn out on 1 amp and the pwm just never turn on with the second circuit I made and all 5 555 Ic just won’t turn on in pwm so am thinking if I use a pin holding Instead of the welding to the pcb so I will see if it’s the welding heat that burn something inside but I only have one working pwm for 10 Ic all 9 was bad the capacitor valve is 1 uf

[@sspence65]

For arduino's with 5v logic, I use IRL540's, with a rDS of .077 ohms and a Vgs of 5v.

[@T2D.SteveArcs]

Hi Lewis love your vids thank you... I have had this problem with E bay silicon fake irfz44 and the like...

[@random2059]

sir,will there any #permanent #change in threshold voltage or RDs ... if mosfet is running hot all the time ?? please reply

[@rosegold6031]

pls.. pls.. help to my project sir.. my mosfets gets hot... and my stepper motor vibrates.. does the switching speed of a mosfet needed to be considered? and how to check switching speed on the data sheet? my design it amplifying signal from a stepper driver allegro.. thanks a lot in advance..

[@LewisLoflin]

Several issues. What mosfets are you using and what is the current-voltage of the motor. First try switching the motor on off no pwm then go from there.

[@supyrow]

yep, do not source your parts from ebay for designs like this. Love your work Lewis!

[@Dc_tech386]

I bought a 50 pack mosfet and all of them or heating up fast in my inverter circuit the frequency is 60khz and i adjusted it all the way to from 1khz to 600khz and heating up as fast it’s on and burn up the transformer is a big 28-28 volt I try all type of frequency I’m my own drive and I can get the mosfet to work cool and stand on i even us a old inverter driver and the mosfet get hot as it on I bought some irfz44n and the same thing happens so am wondering why this having even with a manufacturer inverter drive it’s burn up

[@LewisLoflin]

Several issues here. Poor quality MOSFETs heat up when the frequency is above the rated value. In a way, there is "resistance" to the switching. Lowing the frequency changes the inductive reactance of the transformer. XL drops thus less resistance, and high current heats MOSFET. You need to find the resonant frequency of the transformer and go from there. Be aware circuit capacitance also effects XL.

[@Dc_tech386]

@@LewisLoflin ok can you please give me the right serial rod the good quality mosfet to use the resistance of the transformer is is 4 ohms to secondary and 3 ohms to primary and the capacitor use is 1000uf 25 volt rated and use 24 volt dc

[@LewisLoflin]

@@Dc_tech386 I said reactance not ohmic resistance. That varies by frequency.

[@Dc_tech386]

@@LewisLoflin ok the resistance is 0.88 ohms I rebuilt the 555 timer pwm to a pure sine wave 50hz the transformer turn on nicely but mosfet still get hot in 20 seconds I try IRf740 and irf840 and irf9540 the same keeps having I did check the duty cycle is 50 % and i inverter the signal to get two signal with bjt npn and the voltage at the gate is 15 voltage just 4.9 voltage from fully turn on I did try turn a load on one of the mosfet differently with a 5watt light burn the mosfet heat up so fast at if fully volt of 19.9 volt to the gate so can you please explain what the hell is going on with those mosfet I did repair a inverter and I saw the same irf840 in it burn out I replaced them with IRf740 and the circuit heat up so am think the all the 100 pieces of bad mosfet the Amazon sell send me

[@LewisLoflin]

Because of questions using MOSFETs I tested a lot of them. Here are the results. Added the following related videos: N-Channel MOSFETs Test Results, Observation kzbin.info/www/bejne/qIvCi519ityod9U Parallel P-Channel MOSFETs Results, Observations kzbin.info/www/bejne/e161mIlol9xolc0

[@DAVIDGREGORYKERR]

What about High Power High Voltage MOSFET Modules by Vishay

[@LewisLoflin]

Just looked at the spec sheets. I have never used them but see a lot of potential.

[@DAVIDGREGORYKERR]

@@LewisLoflin Xraytony uses them

[@valentinaselektrikas]

isnt Rds on dependent of temperature?

[@SootySweep22]

night3x Datasheets usually specify RdsOn at 25°.

[@nicholas.hassett9860]

very well explained

[@Tecnotronics223]

Here is solution: use voltage 20 v on gate of transistor mosfet .... and before that use pc817 uptocopler .... to control transistor.... it’s very easy I try it many time and working great up to 40 amp .....

[@Mosfet510]

The IRFZ44N has come in handy for me. Rds on of .02! or less (.0175) Not bad, there are much bigger etc, but it covers a lot. Just my opinion, good video.

[@Dc_tech386]

Queen bro did your z44n get hot because am having that issue with them burn out alot of them in a pwm device with only 5 amps in seconds I fit the pwm capacitor valve that was causing some issues but even at the right valve the mosfet get up 70df even though the data sheet said frequency and 1 uf capacitor the circuit get ready heat up not smoking but hot

[@stephanc7192]

Good video

[@chudyf]

where to buy non fake mosfet

[@electronic7979]

👍👍👍

[@ianbertenshaw4350]

Yes ebay stuff is usually but not always factory seconds . Buy from a reputable supplier and if you find that the components are out of spec you should get a refund or replacement.

[@SianaGearz]

Would be nice if they were merely rejects. They are often enough fakes, some other component with same case, pinout and general function, and if you're "lucky" they chose something with a similar gate capacitance so you don't get immediately suspicious the moment you stick it into a transistor tester. Then the marking side is resurfaced and re-labelled but there's a completely different die inside. Even Chinese domestic semiconductor brands aren't safe from fakes once they gain a bit of popularity.

[@alexkatsaros6180]

Hey! I have many components purchased from EBay I hope I didn’t burn that doe for rubbish!

[@LewisLoflin]

Most of my parts are from Ebay and this is the first problem I have had. The ones that worked were from Ebay.

[@Dc_tech386]

Well I don’t agree if your mosfet is fully on and fully off the voltage drop is 0.9 volt your circuit is not is order the circuit missing a lot of temperature control when you using high and low gate drivers Ic or emitter follower use a pnp bjt to sink the gate fully on for n channel mosfets to fully on and fully off npn bjt to turn off p channel fully of and fully on I don’t have this issue, because when I was learning about gate drivers and low side gate drivers and high side drivers what i discovered is that when I use the driver only the mosfet would heat up badly and even if I use a 10k resistor the mosfets would still get hot and for months I try different methods and the only method that works was bjt as a voltage sink and it work to my surprise the mosfets was turning on fully and off fully with dead time in sink with rising and the mosfet never get even warm at 200 watt without heat sink for 3 minutes and then it get warm because of amperage passing and with 1000 watt and heat sink with fan cooling the temperature rises to 48.6 degrees c so just remember resistor won’t work only bjt can do the job

[@LewisLoflin]

This is circuit theory, not a finished product as I warn on my website.

[@Rev22-21]

AT 3:40....you claim that is a P channel. Shouldn't that be a N channel because the arrow points away from the gate

[@LewisLoflin]

The arrow points away from the gate in a P-channel.