What a fantastic tutorial … as an old crusty 56 year old engineer, you have taught and consolidated some really important stuff that I haven’t seen much in my career being focused mainly on the maintenance and fault investigation side of things. I’m currently having a bash at designing a high current boost converter and it’s no walk in the park. I think as a designer, a good analogy is if this was woodwork, my design would probably work but would resemble a load of roughly sawn bits hammered together with a few rusty nails with over sized parts but structurally weak so don’t use it to sit on!!! As opposed to a beautifully crafted product with smooth lines and a beautiful finish that is light enough to fly at supersonic speeds but can also be used to stand on. 😂😂😂😂
@@EEforEveryone Its clear you are very passionate about your work and therefore its gets me all hiked up. Especially switchers. I know so little but damn sometimes it feels like i can't gulp enough info about them fast enough for my thrirst. Are you working for a company at the moment or is you channel your fulltime job?
@@gerhardgroenewald6560 Hey Gerhard, thank you for the encouragement! That's awesome to hear. SMPS are pretty awesome, and we're going to play around with them a lot tomorrow! (Stream) I have an awesome job at the moment, and I hope it will stay that way for a long time. I have a lot of fun with the channel, but to create big systems that have a huge impact on the world.. it takes a big team. I love teaching, sharing knowledge, and trying crazy things in the shop... but keeping EEforEveryone a hobby is pretty important to me, and a big part of keeping it fun.
@@EEforEveryone I actually read on a comment about a stream sometime? At what time are you streaming? I would love to tune in but i think the time zones will be a problem... I agree with your idea of having an impact on the world. I work with all things mechanical in various sectors for a living so i get to see a lot of ideal opportunities where one could have instead implemented a "smarter" solution going electromechanical. Unfortunately there is still a big gap from primary to secondary industries throughout most parts of the world. Even in smaller secondary sectors. My opinion from being " in the field" would be a lack of knowledge and and not being informed about technology. Well thats a good thing for entrepreneurs. I wish you well with your profession, that your ambition will never die and thank you for your contributions. Please let me know in what time zone you are...
@@gerhardgroenewald6560 Yeah! We'll be live in 5 hours. You can see our scheduled streams here! ru-vid.comlive youtube translates for differences in time zones, which is very helpful. I'm glad that you are able to make big strides in the world through your work as well! I wish the same for you with regard to your ambition and passions! Thank you for being an awesome part of our community!
You are a very clear communicator. As a change of style, I think you would make high quality educational content if you did slower paced demonstrations (hands on with oscilloscope or simulation).
Exceeding vds does destroy fets, on the contrary, it's actually quite useful if you know how to use it. Avalanche mode is allowed as long as the heat generated does not exceed the failure point of the junction. Time and current while in avalanche is the concern. I regularly design switching supplies to use avalanche as means of controlling emi/emc, just need to make sure your inductor spikes time in avalanche and the current flowing doesn't push the junction over its dissipation abilities. The capacitor on the gate is silly, just use a fet driver, or if looking for low cost solution a totem pull of bjts and a pull down resistor will make induced turn on a non issue. Keep making great video's
@@easyamp123 Oops, yes I'm meant doesn't destroy. Just like using a pn junction diode in avalanche, you just have to watch the heat because the junction will now carry both the voltage and current simultaneously but is a useful mode.
Hi Dawn, great point! I certainly wasn't thinking of avalanche-rugged fets when making this video. I suppose that would be a great qualifier here. If the design is meant to use avalanche mode, and you add a couple asterisks about not violating the other limits you mentioned, that is certainly a useful lever to keep in mind! Thanks for the great comment/reminder, useful tips, and thank you for watching!!
Hi, thanku for the info. We're using the NMOS FET in our circuit as we apply voltage across Vgs, Id continously starts increasing and it's not getting stable. What to do?
I wish you touched on the dV/dT subject more because it is something I didn't know and it is rarely mentioned in beginners guides on how to use a MOSFET.. If I understand things correctly, you are saying that in a "totem pole" (don't know that is a good term to use) like one side of an full H bridge, turning on the top MOSFET too fast can result in the lower MOSFET turning on due to the dV/dT at the drain of the lower MOSFET?
Hello wgm! Thanks for watching! Yeah, dv/dt is one of those things where it doesn't always become significant - it can be safely ignored for some applications if best practice is followed. You understood correctly! I observed this while simulating a 1:1 Full Bridge Power Converter we're using in the PFC module for our DIY UPS Project (ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-24PXGbT_0Fs.html) where we talk about this effect, what It meant for our design, and how we worked around it. Maybe this will help by providing another example?
Hello, does this data sheet specification "Vgs - Gate-Source Voltage: - 10 V, + 10 V" mean what I think it does? As in, do not supply more than plus or minus 10V to the gate from the driver? Gate-Source is confusing due to the MOSFET having a "source" pin. Does it really mean Driver-Gate Voltage?
Hello Anond, great question!! My apologies for the slow response. Vgs, is the voltage at the gate, measured with respect to the source pin of a MOSFET. A rating of +/-10V is measured at the part, so typically some margin is left when designing the driver. (Parasitic capacitance can lead to voltages measured at the gate that are above or below the driven voltage. That's more necessary for higher-voltage parts and parts that have high dV/dT events (rapid changes in voltage or current) Side note, but driving an nmos fet with a negative gate voltage will almost always reduce off state leakage current! I hope this helps!
Absolutely! ESD events can kill a mosfet very easily if it strikes the gate pin directly. This can charge the gate pin beyond the VGS rating, destroying the part. In fact, mosfets are most sensitive to this before soldered on to the PCB! (The board will add an element of shielding, additional components, and parasitic capacitance that help to minimize peak Vgs stress during ESD events.)
I am looking for the best mosfets for an induction heater (2 - 4 kW power, input voltage up to 52 V). I tried the power mosfet IRFP4768 from infineon but got terrible ringing problems at the gate. Tried to minimize loop induction and used gate resistor up to 68 ohms - problem dampened but not solved and getting slow slopes - generating even more switching losses. The IRFP4768 has very high gate capacitance of more than 11 nF. Should i better go with the IRFP260N? Has less gate capacitance but also less power handling capability so i probably need 2 or 3 in parallel. Or should i better go with IGBTs? The last tend to have more losses on lower amp ratings, but are more effective on very high power levels but need extremly efficient cooling. I would probably need water cooling to get rid of 100+ W from the to-247? Any ideas?
Hello Richardnis, I think we chatted in a live stream a little while ago, but yes, sinking heat from MOSFETs is a crazy complex problem. Transients, shoot-through, and ringing can cause a lot of problems. Gate drive isn't simple, and the added capacitance doesn't help! It sounds like you're trying the right things... you must be close. :) High-power designs are always a balancing act. In general, spread the heat out, or suck it out as fast as you can. I've had good luck with the "A14692-30" material from Laird - part of the K52 series. If you need a great thermal interface, this is a beautiful combination of polyamide film (great voltage isolation) with a phase change material coated on either side. Thanks for watching!
Hello ?wchen, you'll get it! If you stick to standard to- style packages, there should be many drop-in replacement parts if something isn't working as expected!
I need a mosfet to work in DC region for long time as a DC load battery 60v test . What type you suggest to use ? I need mosfet for test battery as dummy load 60v 20A for 4hr
Hi abed, thanks for reaching out! A linear mosfet is what you're looking for. These devices are designed to be used exactly how you described. I personally bought a cheap electronic load from Amazon, which uses a device like this with a control loop built in. I hope this helps!
I buy one from AliExpress and it's 20A 200v 180W ( DIY 300W ) for short time as they say . You can find it in AliExpress for 45$ have a big fan like a computer cpu fan . But unfortunately when I put a 60 v even in 2amp it just exploded with fire under the fan . The mosfet they use is irfp260 it supposed to handle more than 60v and more than what I applied . But I ask some one to review the dummy load reverse engineering he say that the problem is not for the mosfet . The problem that the driver for the mosfet need to be work in linear compatible with the what mosfet can handle in high voltage with out any nose desorbed the drive . Some other one say that I need to change the dummy resistance loud with a bigger number . I replaced the mosfet with same one from China but it's not working as the original and exploded at 30v 3amp. I make some exploring in Google and found something call irfp90n20d . I install it and it worked at 60v 2 amp as the testing for 15min I put 3amp for more than 60 second and short out the mosfet . So I give up on the dummy load and start to test my battery with car LED ( old lamp 60 / 100w lamp ) serial and parallel way. I was really hope to find some mosfet can really handle max 80v 10a 800W for a more than 4hr with out pressed the mosfet and not let it work like a crazy www.google.com/search?q=dummy+load+20A+180W&sa=X&ved=2ahUKEwipyMPTkrjnAhVSrxoKHRVjCdEQ7xYoAHoECAsQAg&biw=360&bih=512&dpr=3
Hello again, I see. You'll need to pay very close attention to the safe operating area for these parts. Dissipating that much power for that long will be a challenge! Getting that much heat out of a single part may require something exotic like liquid cooling *gasp*. Another option may be to use multiple mosfets in parallel, as a way to share the load. It might be possible to use a linear mosfet like a voltage-controlled resistor, but I'd reccomend having some form of closed-loop control to regulate output current.
Hello Sanel! Good point, I typically like to add a reasonable pull-down on the gate as well. Some gate drivers have sufficient dead-time or adaptive dead-time control, which can help regardless of if this resistor is in the design. In general: I'd rather have the pads on a PCB and not need them than find out later that it was necessary. :) Resistors are small and cheap. Thank you for contributing, I neglected to mention this, and a reminder is always good. As always, thanks for watching!
High dv/dt level causes a current flow through internal Cgd and Cgs caps. I think this resistor helps to damp unwanted gate voltages which crossing the threshold voltage during high dv/dt presence.
@@erhandemirok I think in simple manner. If there's no gate to source resistor, it is very much probable that bottom MOSFET stays turned on, so that when high side command comes to gate of upper switch, it turns it on, while bottom MOSFET is still on.
O: I am not electrician, this information sounds really well made though. I was wondering if you can guess what happened to my spot welder mosfets. I was using 2 12v batteries in parallel for powering the spot welder because 1 was too weak for making a welder. So it worked Very well but the cables were getting too hot. So I decided to use thicker copper wires. But after the first welding the mosfets exploded. What happened? The same setup, just thicker wires.
I need your help about a mosfet gate control . 1 : what is the best (OP) ic to control mosfet gate in DC region . 2 : in high voltage / current What filter I can put in the gate ( because the mosfet when it work in high voltage and the gate get switching voltage so fast . I can hear some time sound like nose ( very high frequency can be detected by oscilloscope between gate and ground ) And if I want a complexity gate driver for a 2 mosfet in parallel ( every one with his own shunt ) but same ground and same voltage but separated gate drive and a feedback from this 2 mosfet to make the current same between this 2 shunt )
Hello Abed, Thanks for watching, and thanks for the comment! You are asking a lot of great questions, and I wish I could answer them completely here. I'll do my best to give a brief overview, but know that we have an upcoming E4E electronic load project! Stay tuned for that series, which will dive into these concepts in a lot more detail! 1. You'll need build some form of compensator. (see TI app note) www.ti.com/lit/an/slva662/slva662.pdf This can be done in hardware, or software, with many different types of devices! 2. Controlling slew rate, etc, is important, and having a gate driver that can supply adequate current will be very important! for multiple FETs in parallel, you may need to have one control loop for each FET, since temperature effects and tolerances in the two linear MOSFETS may lead to imbalanced current sharing. I hope this helps!
Hi Luca, Thanks for watching! I'm glad you enjoyed it. Small update on the giveaways you mentioned a while back... those are going to start happening more often... very soon! I'm working on that.. :)
hello brother... I have made one Boost PFC circuit from UCC28019A by TI. I have made the circuit on general purpose PCB and interconnected the components with wires. I have used the same circuit as TI has given in their evaluation board. All the values are same. But upon providing Vcc to the controller IC in my board, transistor is getting burnt and immediately fuse is getting blown away. But this is not in the case of TI board. It is giving 390VDC as output. Many of my friends are saying because I used general purpose PCB for the circuit and provided loose contacts using teflon wires, the circuit is failing. But I don't feel the same way. I have a doubt that may be because I have not used any snubber, my circuit is failing. Please confirm me whether my speculation is right or Am I just wasting my time...!!!!!
Hello Pasumarthi, Thanks for watching! First off, I'd love to see some pictures of your setup. That sounds awesome!!! You're braver than I to build a complex circuit like this with point-to-point wiring. :) If you have truly duplicated their schematic exactly, the wires may be an issue. Wires will tend to be longer than traces on a PCB, so you'll tend to have more parasitic inductance in your "traces". For critical nets like the switch nodes in a power supply, gate drive, etc, that extra parasitic inductance can have a BIG impact! This extra inductance could be the cause of ringing which is effecting the gate and/or drain of the mosfet in your circuit, so a snubber may help. More likely, or perhaps equally likely, is that there is a small bit of stray wire somewhere that is letting high voltage somewhere it shouldn't be. That is always possible. Without some pictures it's difficult to say more, but think about parasitic inductance you may be adding with your wires and stay safe!
@@pasumarthivenkatanagajayan9284 Hello Pasumarthi! Apologies for the run-around, but I'm trying to use Googles system to avoid spam so I can keep on top of emails from awesome people like you. Here's how to find our email: support.google.com/youtube/answer/57955?hl=en Let me know if you have issues! (P.S.) You can also post your pictures on twitter and tag us with @EEforEveryone to share with the larger community!
Hello Sollertia, MOSFETs usually get warm under load.. but if they get too hot, that's a problem! Somewhere around 80-100C at the junction is usually the hottest I'll push a part... unless I really need to. I hope this helps!
Hmmm, Great Question Sadick, Probably not a discrete MOSFET, but almost every modern IC (processor, microcontroller, etc) is made of many, many small Field Effect Transistors. It's much more likely a signal integrity, memory, or software issue leading to the blue screen!
