Interested in learning about wireless power? Subscribers can get up to 80% off Wireless Power to the People - Wireless Charging 101 on udemy using the coupon code "TH-cam" www.udemy.com/wireless-power-to-the-people-wireless-charging-101/?couponCode=TH-cam
What if my load is a battery? Will PMOSFET also protect me from plugging the charger to the battery the wrong way? Or will the voltage from battery keep PMOSFET open and everything will go in flames?
@@CaptainDangeax Aparently when the load can provide its own voltage, one mosfet is not enuough. You need at least two and two resistors to avoid blowing fuses or something more valuable. www.analog.com/media/en/technical-documentation/application-notes/AN-171_AN-1568.pdf
@@niceshotapps1233 My point is why would you use such a protection on a battery charger ? Of course it's perfectly fit for any electronic appliance with external power supply, but for a battery charger, I don't see the point as I already wrote
2023, 11 years on and this still helps a lot of us. Just dropping this comment to wish you well, and hope that you're living your best live. Warm regards from Manila, PH
Ok, I’m here. Sorry for the 4 year delay but please understand there’s a major backlog for TH-cam commenters requesting awards. With that being said... Hear Ye Hear Ye! We gather here today to recognize Afrotechmods for his stellar video tutorials that have gone above and beyond in helping others. His no bullshit approach combined with clear and concise dialogue alone was far more than his competition could come up with yet even that wasn’t enough for Afrotechmods. In a game changing manner, Afrotechmods set the bar higher than Willie Nelson at Snoop Doggs pool party by making a video without a single fucking digression, a feat many TH-cam viewers believed to be impossible. We Salute you, Afrotechmods!
Props to you for making electronics understandable to the rookie. Years ago, I took an electronics class at the college. First off, I didn't understand jack about what was really going on.... with a few exceptions. But I knew how to calculate stuff, and if I built the schematic right, it worked. Thus I got an A. Boy did I fool them. Later in life I'm exploring electronics again, and I'm pleased that there's lots of good videos that explain this stuff in a less technical fashion... so a rookie can get it.
Yours is a much better video for newbies like me, than the other popular channels which actually just zoom past the descriptions. Your formula and circuit descriptions were very lucid. Thank u..
or use a diode on the + and - backwards so that it will short circuit ur power supply if connected backwards for dirt cheap and effective reverse polarity protection :)
Professional Solution: 1)Use of data-sheet parameters to support design. 2)Didactic video that explains clear and detailed. 3)focus only in topic. Great Tutorial.
In the past, mosfets and co. was black magic for me. When I saw your tutorial videos, I understood it + the datasheets of mosfets make sense now. I really appreciate your time spending such good clips and hope to see more of them.
Very intuitive explanation. You make excellent videos!! Please make more! One of the channels on TH-cam that does make sense and is very educational down to the point you want. Lots of good examples and case scenarios.
Am I glad I stumbled on this tutorial! What a great trick. I never thought about that one. Furthermore, one could do the same with an N type MOSFET by just connecting the drain on the minus side of the power source.
sony10 , yes , I agree that at first glance, the circuit appear quite weird. But when you look closely pay attention to the polarity of a P channel Gate compare to Source. Keep in mind that a MOSFET is a variable resistor ( at least when utilized at low voltage ) . This means that , when ON, (short) , you can feed current both polarities, it still will act as a short circuit. This is why it appear weird when you look at a circuit that show a P channel connected as if it was a N channel. In this circuit the PMOS is utilized in reverse polarity. It need to be this way because of the internal inevitable substrate intrinsic diode that would be turned ON if you were to use an N channel and made a mistake of reversing polarity. You don't want any current circulating if in reverse battery connection. Hence, wether using a P or NMOS you want the internal diode in the direction of the current flow. Most circuits utilize a MOSFET configuration with current being blocked by the internal diode. In this circuit you want it configured such that when the battery polarity is correct the current would flow through the internal diode (if the gate was not energized) because in case of reverse battery polarity ( an undesirable error ) you certainly do not want any current flowing. So, you Then have to connect the PMOS in reverse, which is totally allowed according a MOSFET specification sheet. Once the battery is connected properly the internal diode will bring the Source to a higher potential then the Gate ( even if the Drain is of higher potential, it's not prohibited) you now have a Gate that could be connected to a voltage potential lower than the Source ( a condition to turn ON the PMOS) . Once the Gate is put to a lower potential it turns ON the PMOS and the voltage lost in the internal diode becomes almost zero because you now have a PMOS turned ON ( a short circuit ). Hope that make any sense to you. At least, that is how I see it.
sony10 , the short version. A MOSFET can be utilized both polarities. As long as the Gate is at lower voltage then the Source you have a ON (short) MOSFET , which is what you want here in this circuit. D higher voltage than S is totally allowed if you do not exceed the forward internal diode maximum current mentioned in the specification sheet of your device.
Very nice introduction to reverse battery protection technics. You also could have mentioned the use of NMOS (and its associated charge-pump circuit) if you are in the case of very high drain current (NMOS tends to have better Rdson characteristics than PMOS). But this method requires more complex circuitry of course ...
Handy to know tip! In certain cases where you're not bothered about power efficiency, just use a bridge rectifier then you can connect your power leads which ever way you like and it'll always work!
@mediaguardian Above -2V, (i.e. -1V, 0V, +2V etc.) the mosfet is in cutoff mode ("off"). Between -2V and -4V (specified min and max) the behaviour could easily vary from batch to batch and you could end up in the linear region where the Rds is several ohms to several kilohms. Then -4V or less (-5V, -6V etc) you enter saturation mode ("on") with lower resistance as you approach -25V. So in a switching application the designer knows to stay well beyond the -2...-4V range
Excellent explanation! There could be just one minor improvement - the contradiction proves the transistor cannot *stay* ON but it does *not* prove it wouldn't start to conduct from the initial state when it is still OFF. In other words, without proving the process of conducting could never *start* we cannot be sure if there aren't oscillations. Because, *if* in the initial moment the transistor was starting to conduct, we proved only it will turn OFF once after it *already* was turned ON. If the tranistor would then (when it is OFF) be able to start conducting (again) the process would repeat again and again. I am aware these conditional sentences are not grammatically 100% correct but I hope it could be understood what I wanted to say.
If the battery is connected in reverse as in 3:21 then the Drain of PFET is below ground, then the Source pin doesn't have much chance of ever going above ground so PFET should never turn on.
Thanks for the video. But when choosing the right MOSFET for some application, I think that the choice must also take in consideration the maximum current allowed through the transistor.
@CoolDudeClem I implied several times in the video that if Vgs is 0V the FET would be off. It needs -4V or less to turn on. The circuit at the end of the video with the zener diode will clamp a large gate voltage down to the zener voltage.
Good stuff. I've seen a few of your videos like this one. You don't overload them with information the average hobbyist doesn't need, but you don't leave it so vague that a part timer might still get it wrong.
1. schematic called "ideal diode". 2. n-mosfet better then p-mosfet (cuz Ron is smaller). it have simmetrical schematic but inserted to ground wire . 3. nice video :) best regards
Use a charge pump on the gate. Really simple to implement because its not drawing any current. You can buy a small IC that provides the voltage you need. To protect the charge pump IC you can just implement pMOS or a simple diode on its VCC terminal. Since it draws current in the uA range, the losses are negligible. The losses in implementing the charge pump to have a high side nMOS is usually much lower than using a pMOS
@@ericcartmann true but look at the datasheet again, it's spec'd at 0.026 ohms @ ~24 amps/10 volts, ( heat sunk to 25 deg C of course ) in any case at 2.0 amps it's much much less. personally i prefer Geranium diodes over Si... lower Voltage drop . . . & i have a ton of those laying around in metal cases.
@@robozstarrr8930 The problem with those metal cases is they are made of zinc, and zinc whiskers. Most of them have already shorted themselves out and they are no longer manufactured.
Be _very_ careful about that on a complex system (with multiple boards, for example) because you may short the "Protected" Ground with the real ground over multiple boards and fry them all T.T
If you think of protective fet as of part of the battery, it doesn't make any difference, is it pfet at positive battery terminal, or a nfet at negative one. In other words, just assume your ground and positive rail after fet, and never think of reaching battery's terminals, and you are good. So I would think of using nfet at bottom always, they often go under 10 mOhm of rds on. Also, nfet is very convenient when you want to control its gate with a MCU for soft on-off feature.
Thank you for this video. I have watched numerous videos all talking about reverse polarity and showing you a generic circuit. But very few bother to comment on how to choose your parts. Again, thank you.
Honestly, the only thing I really understood was at the very very end when u said "You are now hereby granted with powers of reverse voltage protection!" 😬👍
This one is really good for me as I have just finished making my 0v to 27 v regulated discreet power supply with variable current . The lecture is very clear and precise.Very good video
I know this works, but one thing is bugging me: if you are using a PMOS, the normal current flow should be from source to drain, and not the other way around. This should be OK if you had the symmetrical component (like JFET), but this is not the case. Can you explain this a bit?
Milan Lukic In fact, for PMOS, if Vds>0.7 or so, it just turn on and act as a diode (no matter what the voltage is on the Gate). you can get this info. from it's symbol...
Milan Lukic MOSFETs generally work on the fact that a channel is created between drain and source when the gate charge attracts charges into the channel, reversing its behavior. When a channel is created, it doesn't matter which way the electrons flow in it.
Milan, you’re right. The Drain-Source should be swaped. MOSFET has a internal diode that points from source to drain in N channel devices and drain to source in P channel devices. Therefore in a PMOS current flows from Source to drain not the other way around.
This limits Zener current. Máximun R value is calculated using Iz (reverse current when Vz is reached, typically a very small value). In this circuit: R = (30 - 10) / Iz As R= 100k, Iz must be 0.2mA In my opinion, you can use this 0.2mA value without fear of making a mistake, but you can take 1mA just to be sure
Thank you. You explained this really well! I like how you don't gloss over any of the important details and how you work up from a simple solution to a more complex one and explain all the hows and whys. Really excellent video!!
in a book i am reading, we just got from diodes to transistors and semitransistors and so i am glad to see this video from you. you can be a very great teacher for me to learn from
I have the best and simpliest solution to this( without any wasted energy) Get yourself a some round connector, that you cant plug in reversed ! It's drunk proof, dumb proof...... :D
That's funny. That is until I started to really thinking about it, it's not such a bad idea after all. If we made the whole world with drunk proof electricity then no more Boo Boo's with all of that going on No More, No More, as the Raven said, LOL.
Please, please make more vids! I have learnt more from you than any other. You have ultimate teachers talent. Use it to fulfill child's dream of better world.
I can't for the life of me figure out how I only now discovered your channel... TH-cam recommendations are lacking! I've been watching EEVBlog and W2AEW for years, but only found your channel recently.
To complete the analysis, you can show that the power lost by the 100Kῼ and diode is far less than the 1.7W across a forward-biased diode; power dissipated by 10V across a 100Kῼ = 0.4 mW? That's around a 3 decade improvement! Excellent by any standard.But at that voltage, the 100Kῼ only allows 200uAmps through the Zener. That might not constitute a high enough keep-alive current for the diode to work. At any rate, that was nicely done. Thanks for the crystal clear sequence of analysis.
Hello @Afrotechmods! At minute 3:00 in the video you subtracted 11 Volts off of 0 Volts. Could you please explain why you did it in this order and not the other way around (11V - 0V = 11V)?
That's just how you calculate the voltage between two points. The formula is at the bottom of the screen. More generally, V12 = V1 - V2. And this is of course always referenced to whatever the ground/common node in the circuit is.
Think of a little guy inside with a voltmeter and a rheostat. The voltmeter measures the voltage between the gate and source. As that voltage increases, he turns the resistance down on the rheostat. It's the same as the model for a BJT, except that the little guy there has an ammeter instead of a voltmeter.
You can do the same thing by using N-channel mosfet, and it will be more efficient. But you have to connect the mosfet to the low side. If you don't have problem with placing some circuit element between the circuit ground and outside ground, I suggest using N-channel mosfet.
@@robotboy3525 N-channel uses electrons as charge carrier, while P-channel uses holes. If you go to any semiconductor part seller (e.g.; Digikey, Farnell, Mouser, etc) and enter the same mosfet parameters (Rdson, Vdsmax, package type, Ciss, ect) as filter, you will see that a similar N-channel mosfet is very cheaper than the equivalent P-channel one.
Holes vs Carriers ??? So by using electrons as carriers how does that make it more efficient?? Also when you say cheaper, are you referring to the pricing ?? Based what I learned from an electrical engineer, n-channel is much faster since it paves the path to ground but it is the most dangerous. Pretty much all reverse voltage polarity integrated circuits on the market uses p-channel. Thus I think this would be the case because safety comes first.
This has made everything so clear, I have been trying to figure out p-channel mosfets for months now in a circuit simulator circuit simulator and getting really confused as to why things were not working the way I thought they would, and that's all because it doesn't show the ''parasitic diode'' in the mosfet.
This is a brilliant idea but one thing is not negligible though. In case of a high current load or even a short, the voltage drops at the drain port so that GS threshold voltage is not given anymore leading to a current flow through the diode and a resulting death of the fet. This is only useful with a following current limiter.
THANK YOU. CLEAR. EXACTLY WHAT WE WERE TALKING ABOUT AT BREAKFAST; AND YOU'VE ALREADY DONE IT !!! I WAS USING THIS IN SOLAR PANEL DESIGN ELIMINATING DIODE LOSS. LOOKING AT N CHANNEL FOR LOWER RDSS. NICE TUTORIAL, THANKS AGAIN.
There is also a way to connect a diode in parallel with the load, but this will damage the power supply. The method shown is a well-thought-out practical circuit.
@Afrotechmods that is exactly what I saw in the sim, and would expect. Thanks for the videos. Your time and effort making these videos are greatly appreciated. Thank you.
In old era devices relay also was used. You place contact of relay which is normally open in series with circuit under protection and connect relay coil through diode, so it turns on and closes the contact only if power is connected on the right polarity. It also could be used as undervoltage protection.
@colt4547 You absolutely can. But the downside is that you introduce a few milliohms of resistance on your ground rail, which may or may not be a problem depending on what you are doing.
Excellent. Just what I need. I'd heard of doing this but didn't know which FET to use or how to connect it. I previously used a reverse connected diode, by accident I connected the battery in reverse one day, sure it passed a lot of current, blew the fuse and saved my circuit but it blew the end off the diode. Another diode needed. This sounds ideal.
You alluded to it a bit, but I'm going to say it outright. The gate threshold voltage is just as important as the other parameters. Not only that, but the threshold voltage is specified for a certain amount of current. This is most important for low voltage supplies. If the load drops more current, the FET will act like a resistor, giving really bad losses. If it isn't heatsinked it can even overheat and fail. Usually you have to look at the transconductance curve. Hopefully there is one. Sometimes there is not, so it's a good idea to pick something that has a better datasheet unless you're sure the battery will have enough voltage at all times. Find where the load current and gate voltage meet on the curve and ensure that your battery or power source will never drop below that level.
In this very case, the MOSFET will never act as a straight up resistor, as it's always fully closed in reverse polarity, but in desired polarity, the body diode always limits the drain-source voltage and thus the losses for a given amount of current; additionally, the gate will partially open and reduce the voltage drop further. It's always an important consideration though. If you dimension the cooling of the MOSFET sufficiently, the battery voltage droop is not dangerous and it still beats a diode. Of course you also want to beware of the voltage-current behaviour of the circuit behind it - some will consume about the same current independent of voltage, many will consume a lot less current, the lower the voltage is, and then some, switching regulators in particular, will consume the more current, the lower the voltage is, and the latter causes a potentially dangerous corner case.
As in all cases, a suitably rated fuse at the battery is good. If your load can sink that much current at low voltage, then adding circuitry to switch the MOSFET completely off below a threshold is needed. Its not too difficult a design task. Discharging the battery to low voltage may kill it too!
love ur sense of humor. great vid too! I love to blow things up. ran 5v through a 3v led, lens cap got hot, exploded sending the cap through the ceiling tile AUSOME!!!!!!
It should be noted that if you want to use this in a high efficiency design that protecting the mosfet using a zener diode is much less efficient because the zener diode needs at least a few milli amps to safely limit the voltage over the gate-source junction. As I'm designing an 18650 Li-Ion battery based protection circuit that doesn't just protect the circuit against reverse voltage but also protects the battery against overdischarge, I can't use a zener diode to protect the mosfet. But I don't really need to either, because the Si2371EDS I got a while ago has a maximum Vgs of +/- 12V, so it will be able to handle the 4.2V of the 18650 cell easily. And the 3.1V at which I'm planning to switch it off, minus the body diode maximum forward voltage drop of 1.2V, gives 1.9V which is still more than enough to switch it on and let the TLV3012B comparator handle disconnecting the load.
I love coming to your channel and videos to "wrap up" my understanding of various topics. I'm newbie and just getting back into using p-channel mosfets. I had an ok understanding of how to read the p-channel data sheet to choose the correct transistor for my app however after listening to this it makes complete sense - I wish all my teachers were like you "back in the day" (or at least I wish they had youtube when I was in college lol) - thank you!
When the diode in reverse mode we don't have 2A current flowing thru . Only very tiny leakage current can go thru it, that's normally considered as 0A . Therefore, the power dissipation across the diode in this circuit is also considered as 0W . The first circuit is good enough for reverse voltage protection . However, when the voltage is in forward direction, we will loose 2A X 0.7V = 1.4W if using silicon diode, or we will loose 2A x 0.3 A = 0.6W if using germanium diode .
@LiKBeAsT "So when it's not ... choose which ever one we want (without being too supid:P) ?" Yes. The mistake you are making with the battery is that you are viewing the + terminal as an absolute value of +12V. Voltage is always relative, between two points. The batt's + terminals is 12 volts higher than the batt's -ve terminal. So if we declare GND to be at the batt's + terminal, the -ve terminal has to be at 12 volts less than that.
With a bit more "tuning" the circuit could be set up to limit the current drawn by the load. Another cheap way to prevent reverse voltage from damaging a circuit is put a relay in series. The coil is powered by forward biased diode. There will by more current/power absorbed by the relay's coil than by a P-FET, but in some cases it may be cheaper especially if the load is high and the source has power to "burn".
@R4ND0M1C That is, why you have to use at least the resistor in the gate, but best even the Zener, even for lower voltages then Vgsmax. That serve as a voltage limiter clamp (together with the MOSFET on-chip gate protection clamp and the gate capacitance, witch goes in nF range), so even with the ESD discharge, the gate voltage is clamped to way safe value.
In p-channel.. load is always connected to source.. when gate is pulled down to ground, the current flows from source and passes through the drain into the load. Mosfet is bi directional. Your video is trying to tell the opposite way if I am not wrong. Your explanation is pretty good. @ 2:39 if the mosfet is ON , then body diode turns OFF. @ 2:39 both are conducting which is a big flaw.. Body diode is ON when the FET is OFF... I am just trying to explain it. Your videos are awesome...
Awesome explanation, we do this at work but many an engineer just take it as tribal knowledge, it is refreshing to have the principal explained as the teacher would explain it to the pupil.
The easiest way to protect your circuit from reverse polarity is to add a rectifier bridge before the circuit input. Cheap and effective. Nice video though.
Very nice bit of info here. I always felt the simple diode (K) to input between B+ and a second diode also (K) to input from output, used for rev Pol. protection on Vreg IC's was suitable, still is of course since these are usually used in such low power circuits. However, since battery packs such as Lipo are now such a common supply for circuits, I really like knowing about this method. Now I need to put up a stock of P channel FET's and Zeners.
@LiKBeAsT I've thought about doing such a video. There are just too many things to make videos on... Anyways when we are not talking about ground meaning the earth's ground, ground can be anywhere you want in the circuit. It is a reference point. You can choose the ground node to be anywhere you want and do your design & analysis based around that. However, using certain nodes makes things easier than others. Most of the time it's easiest if GND = battery's -ve terminal.
MOSFETS are a great way to do this. The on resistance can be as low as 7mOhm for a P-Channel (Fairchild FQP27P06). Power = (0.01ohm) X 2^2 = 40mW loss (heat).
You can do it with an N fet on the low side in a similar way just be aware that the resistance you are introducing on the ground path can cause problems in some circuits
Great for small currents and voltages drawn from batteries.For higher currents a bridge rectifier can be used or if the voltage is known and stable a relay with two independent contacts for making a dont care polarity input.
Thank you very much, someone FINALLY!!! did a video on this. I've been looking for this particular answer for a week or two by now. First one was the SINE WAVE-MADE PURE was my first & of coarse the second one fell right into order with what I needed. I subscribed so hopefully I'll get any new ones on these subjects. Thanks again, Tom.
Your videos are awesome. I was watching a few of them trying to connect the dots for a simple reverse voltage protection circuit that doesn't have a big forward drop (I'm splitting the output from a computer power supply to try and power a few ultra-low-power boards from one). Then I saw this, which makes those leaps for me. AWESOME. Also, if you think there's any huge red flags I should watch out for in my project, please say something. :) Otherwise, keep making videos for the rest of time.
Afrotechmods Iv heard of this before, now I know how to do it, thanks! I will be needing this in the future. You make great tutorials. I never thought of doing this as I didnt know how, I knew about using a diode which is a bit lossy as you said. Of course I one used this to my advantage when something REQUIRED a 8.4V supply and I only had 9V. I believe I will watch all your tutorials, I could learn a lot form them.
Great video! I literally just ordered some assembled PCBs with diode protection just this morning before watching this. should have implemented the mosfets instead doh! thanks for the great explanation
While its a good video I think the best protection is a physical one. Just make your power source unable to plug in backwards. Also the diode is more cost effective and fine for most DIY projects if the user is worried but its great to know other options as well and thats where this comes in.
@xanokothe cell phone chargers rectify mains voltage and pulses it through a really small transformer(somtimes just a little bigger than your thumb) with a transistor. This is just based on what Ive seen when I took one apart.
Interested in learning about wireless power? Subscribers can get up to 80% off Wireless Power to the People - Wireless Charging 101 on udemy using the coupon code "TH-cam"
www.udemy.com/wireless-power-to-the-people-wireless-charging-101/?couponCode=TH-cam
thank you for this very usefull video. I know what to do with all those useless mosfets I got on dead PC motherboards
What if my load is a battery? Will PMOSFET also protect me from plugging the charger to the battery the wrong way? Or will the voltage from battery keep PMOSFET open and everything will go in flames?
@@niceshotapps1233 The charger already provides a diode bridge rectifier and a fuse. Plug it the wrong way a the fuse will simply blow
@@CaptainDangeax Aparently when the load can provide its own voltage, one mosfet is not enuough. You need at least two and two resistors to avoid blowing fuses or something more valuable. www.analog.com/media/en/technical-documentation/application-notes/AN-171_AN-1568.pdf
@@niceshotapps1233 My point is why would you use such a protection on a battery charger ? Of course it's perfectly fit for any electronic appliance with external power supply, but for a battery charger, I don't see the point as I already wrote
2023, 11 years on and this still helps a lot of us. Just dropping this comment to wish you well, and hope that you're living your best live. Warm regards from Manila, PH
Same here, only 2024.
This series is excellent. Easily the best electronics tutorials on TH-cam that I've run across. Someone please give this man an award.
Ok, I’m here. Sorry for the 4 year delay but please understand there’s a major backlog for TH-cam commenters requesting awards. With that being said...
Hear Ye Hear Ye! We gather here today to recognize Afrotechmods for his stellar video tutorials that have gone above and beyond in helping others. His no bullshit approach combined with clear and concise dialogue alone was far more than his competition could come up with yet even that wasn’t enough for Afrotechmods. In a game changing manner, Afrotechmods set the bar higher than Willie Nelson at Snoop Doggs pool party by making a video without a single fucking digression, a feat many TH-cam viewers believed to be impossible. We Salute you, Afrotechmods!
True
Why dont YOU give him an award?? Allways the others should do it...
Here's the award : 🏅🏆
Props to you for making electronics understandable to the rookie. Years ago, I took an electronics class at the college. First off, I didn't understand jack about what was really going on.... with a few exceptions. But I knew how to calculate stuff, and if I built the schematic right, it worked. Thus I got an A. Boy did I fool them. Later in life I'm exploring electronics again, and I'm pleased that there's lots of good videos that explain this stuff in a less technical fashion... so a rookie can get it.
Yours is a much better video for newbies like me, than the other popular channels which actually just zoom past the descriptions. Your formula and circuit descriptions were very lucid. Thank u..
Just wanted to say you make the best tutorials !!!!!!
+George Kot Thanks!
Agreed!
or use a diode on the + and - backwards so that it will short circuit ur power supply if connected backwards for dirt cheap and effective reverse polarity protection :)
If your 'power supply' is a non protected lithium polymer battery, that's not going to be pretty
With a proper rated fuse or polyfuse this could work
Professional Solution: 1)Use of data-sheet parameters to support design. 2)Didactic video that explains clear and detailed. 3)focus only in topic. Great Tutorial.
I've been doing electronics for 50 odd years and how that simple trick evaded me I'll never know. Well presented.
lol
In the past, mosfets and co. was black magic for me. When I saw your tutorial videos, I understood it + the datasheets of mosfets make sense now.
I really appreciate your time spending such good clips and hope to see more of them.
Very intuitive explanation. You make excellent videos!! Please make more! One of the channels on TH-cam that does make sense and is very educational down to the point you want. Lots of good examples and case scenarios.
You're born to be a teacher! The explanation is easy to follow
Am I glad I stumbled on this tutorial! What a great trick. I never thought about that one. Furthermore, one could do the same with an N type MOSFET by just connecting the drain on the minus side of the power source.
Hey man. The circuit in this video. Why does D and S revesed?
Usually, PMosfet, S to + and D to load!!! This is crazy
Plz help 😣
sony10 , yes , I agree that at first glance, the circuit appear quite weird. But when you look closely pay attention to the polarity of a P channel Gate compare to Source. Keep in mind that a MOSFET is a variable resistor ( at least when utilized at low voltage ) . This means that , when ON, (short) , you can feed current both polarities, it still will act as a short circuit. This is why it appear weird when you look at a circuit that show a P channel connected as if it was a N channel. In this circuit the PMOS is utilized in reverse polarity. It need to be this way because of the internal inevitable substrate intrinsic diode that would be turned ON if you were to use an N channel and made a mistake of reversing polarity. You don't want any current circulating if in reverse battery connection. Hence, wether using a P or NMOS you want the internal diode in the direction of the current flow. Most circuits utilize a MOSFET configuration with current being blocked by the internal diode. In this circuit you want it configured such that when the battery polarity is correct the current would flow through the internal diode (if the gate was not energized) because in case of reverse battery polarity ( an undesirable error ) you certainly do not want any current flowing. So, you Then have to connect the PMOS in reverse, which is totally allowed according a MOSFET specification sheet. Once the battery is connected properly the internal diode will bring the Source to a higher potential then the Gate ( even if the Drain is of higher potential, it's not prohibited) you now have a Gate that could be connected to a voltage potential lower than the Source ( a condition to turn ON the PMOS) . Once the Gate is put to a lower potential it turns ON the PMOS and the voltage lost in the internal diode becomes almost zero because you now have a PMOS turned ON ( a short circuit ). Hope that make any sense to you. At least, that is how I see it.
sony10 , the short version. A MOSFET can be utilized both polarities. As long as the Gate is at lower voltage then the Source you have a ON (short) MOSFET , which is what you want here in this circuit. D higher voltage than S is totally allowed if you do not exceed the forward internal diode maximum current mentioned in the specification sheet of your device.
@@jimviau327 tks alot
And tks again!!!
Im thinking.... This one is crazy
@@jimviau327 now I can die in peace ✌️
I am impressed by the clarity in your English and pronunciation.
Very nice introduction to reverse battery protection technics. You also could have mentioned the use of NMOS (and its associated charge-pump circuit) if you are in the case of very high drain current (NMOS tends to have better Rdson characteristics than PMOS). But this method requires more complex circuitry of course ...
If your battery is isolated from 0V a N MOSFET on the negative works just as well😊
Handy to know tip! In certain cases where you're not bothered about power efficiency, just use a bridge rectifier then you can connect your power leads which ever way you like and it'll always work!
@mediaguardian Above -2V, (i.e. -1V, 0V, +2V etc.) the mosfet is in cutoff mode ("off"). Between -2V and -4V (specified min and max) the behaviour could easily vary from batch to batch and you could end up in the linear region where the Rds is several ohms to several kilohms. Then -4V or less (-5V, -6V etc) you enter saturation mode ("on") with lower resistance as you approach -25V. So in a switching application the designer knows to stay well beyond the -2...-4V range
Great video, no god-awful background music (which is always too loud), or sound effects. Thanks Afrotech, you do good work.
Yeah Bob you're right for once somebody mentioned something about the crappy generic music that people put on these electronic videos..thanks :)
Excellent explanation! There could be just one minor improvement - the contradiction proves the transistor cannot *stay* ON but it does *not* prove it wouldn't start to conduct from the initial state when it is still OFF. In other words, without proving the process of conducting could never *start* we cannot be sure if there aren't oscillations. Because, *if* in the initial moment the transistor was starting to conduct, we proved only it will turn OFF once after it *already* was turned ON. If the tranistor would then (when it is OFF) be able to start conducting (again) the process would repeat again and again. I am aware these conditional sentences are not grammatically 100% correct but I hope it could be understood what I wanted to say.
If the battery is connected in reverse as in 3:21 then the Drain of PFET is below ground, then the Source pin doesn't have much chance of ever going above ground so PFET should never turn on.
The world needs tons of people like you Sir....!
very well detailed explained...!
Thanks..!
Thanks for the video. But when choosing the right MOSFET for some application, I think that the choice must also take in consideration the maximum current allowed through the transistor.
@CoolDudeClem I implied several times in the video that if Vgs is 0V the FET would be off. It needs -4V or less to turn on. The circuit at the end of the video with the zener diode will clamp a large gate voltage down to the zener voltage.
Just as your other videos, a clear and concise explanation of an electronics concept. Thank you.
Good stuff. I've seen a few of your videos like this one. You don't overload them with information the average hobbyist doesn't need, but you don't leave it so vague that a part timer might still get it wrong.
1. schematic called "ideal diode".
2. n-mosfet better then p-mosfet (cuz Ron is smaller). it have simmetrical schematic but inserted to ground wire .
3. nice video :) best regards
+nRADRUS 2. True, but we want a common ground, if somehow possible.
lg, couka
Use a charge pump on the gate. Really simple to implement because its not drawing any current. You can buy a small IC that provides the voltage you need.
To protect the charge pump IC you can just implement pMOS or a simple diode on its VCC terminal. Since it draws current in the uA range, the losses are negligible.
The losses in implementing the charge pump to have a high side nMOS is usually much lower than using a pMOS
@@ericcartmann true but look at the datasheet again, it's spec'd at 0.026 ohms @ ~24 amps/10 volts, ( heat sunk to 25 deg C of course ) in any case at 2.0 amps it's much much less. personally i prefer Geranium diodes over Si... lower Voltage drop . . . & i have a ton of those laying around in metal cases.
@@robozstarrr8930 The problem with those metal cases is they are made of zinc, and zinc whiskers. Most of them have already shorted themselves out and they are no longer manufactured.
Be _very_ careful about that on a complex system (with multiple boards, for example) because you may short the "Protected" Ground with the real ground over multiple boards and fry them all T.T
This channel is a Carrier and by extension life saver,tons of credit to you Boss.
Great video. It's much easier/cheaper to find low Rds on n-channel mosfet than p-channel. Once you get to
If you think of protective fet as of part of the battery, it doesn't make any difference, is it pfet at positive battery terminal, or a nfet at negative one. In other words, just assume your ground and positive rail after fet, and never think of reaching battery's terminals, and you are good.
So I would think of using nfet at bottom always, they often go under 10 mOhm of rds on. Also, nfet is very convenient when you want to control its gate with a MCU for soft on-off feature.
Thank you for this video. I have watched numerous videos all talking about reverse polarity and showing you a generic circuit. But very few bother to comment on how to choose your parts.
Again, thank you.
Honestly, the only thing I really understood was at the very very end when u said "You are now hereby granted with powers of reverse voltage protection!" 😬👍
This one is really good for me as I have just finished making my 0v to 27 v regulated discreet power supply with variable current . The lecture is very clear and precise.Very good video
Maxx Smaxx the zip
I know this works, but one thing is bugging me: if you are using a PMOS, the normal current flow should be from source to drain, and not the other way around. This should be OK if you had the symmetrical component (like JFET), but this is not the case. Can you explain this a bit?
Milan Lukic In fact, for PMOS, if Vds>0.7 or so, it just turn on and act as a diode (no matter what the voltage is on the Gate).
you can get this info. from it's symbol...
Milan Lukic MOSFETs generally work on the fact that a channel is created between drain and source when the gate charge attracts charges into the channel, reversing its behavior. When a channel is created, it doesn't matter which way the electrons flow in it.
Milan, you’re right. The Drain-Source should be swaped. MOSFET has a internal diode that points from source to drain in N channel devices and drain to source in P channel devices. Therefore in a PMOS current flows from Source to drain not the other way around.
ElectroBOOM Medhi! You are one of my heroes
@@SakarPudasaini10 you are a godsend
I'm just starting to study electronics and these are some of the most fun vids on YT.
Please continue to make videos!
You are simply an AWSOME teacher. Fantastic work! Thank You.
Wow. It is something awesome and cool. Great work uploaded.
This limits Zener current. Máximun R value is calculated using Iz (reverse current when Vz is reached, typically a very small value). In this circuit: R = (30 - 10) / Iz
As R= 100k, Iz must be 0.2mA
In my opinion, you can use this 0.2mA value without fear of making a mistake, but you can take 1mA just to be sure
Thanks man, I always learn something watching your vids
Afrotechmods come back, I miss how informative and easy to follow your videos are for learning. :(
amazing tutorial. thank you very much for the fine explanation. thoroughly clear
Thank you. You explained this really well! I like how you don't gloss over any of the important details and how you work up from a simple solution to a more complex one and explain all the hows and whys. Really excellent video!!
"...and you can be as drunk and irresponsible as you want." 😂
in a book i am reading, we just got from diodes to transistors and semitransistors and so i am glad to see this video from you. you can be a very great teacher for me to learn from
I have the best and simpliest solution to this( without any wasted energy) Get yourself a some round connector, that you cant plug in reversed ! It's drunk proof, dumb proof...... :D
That's funny. That is until I started to really thinking about it, it's not such a bad idea after all. If we made the whole world with drunk proof electricity then no more Boo Boo's with all of that going on No More, No More, as the Raven said, LOL.
Poka yoke, always best indeed. Thought a protection can't do any harm...
Until the drunk installs the batteries backwards.
Please, please make more vids! I have learnt more from you than any other. You have ultimate teachers talent. Use it to fulfill child's dream of better world.
Thanks for the great video, now I can make a more efficient killing machine. :-)
I can't for the life of me figure out how I only now discovered your channel... TH-cam recommendations are lacking! I've been watching EEVBlog and W2AEW for years, but only found your channel recently.
To complete the analysis, you can show that the power lost by the 100Kῼ and diode is far less than the 1.7W across a forward-biased diode; power dissipated by 10V across a 100Kῼ = 0.4 mW? That's around a 3 decade improvement! Excellent by any standard.But at that voltage, the 100Kῼ only allows 200uAmps through the Zener. That might not constitute a high enough keep-alive current for the diode to work. At any rate, that was nicely done. Thanks for the crystal clear sequence of analysis.
Great I am a master at this...now I just have to learn electricity in physics...lolz
Hello @Afrotechmods! At minute 3:00 in the video you subtracted 11 Volts off of 0 Volts. Could you please explain why you did it in this order and not the other way around (11V - 0V = 11V)?
That's just how you calculate the voltage between two points. The formula is at the bottom of the screen. More generally, V12 = V1 - V2. And this is of course always referenced to whatever the ground/common node in the circuit is.
Mosfets are soo strange
Think of a little guy inside with a voltmeter and a rheostat. The voltmeter measures the voltage between the gate and source. As that voltage increases, he turns the resistance down on the rheostat. It's the same as the model for a BJT, except that the little guy there has an ammeter instead of a voltmeter.
You can do the same thing by using N-channel mosfet, and it will be more efficient. But you have to connect the mosfet to the low side. If you don't have problem with placing some circuit element between the circuit ground and outside ground, I suggest using N-channel mosfet.
AhmedHan,
How is N channel efficient ?? care to elaborate ?
@@robotboy3525 N-channel uses electrons as charge carrier, while P-channel uses holes.
If you go to any semiconductor part seller (e.g.; Digikey, Farnell, Mouser, etc) and enter the same mosfet parameters (Rdson, Vdsmax, package type, Ciss, ect) as filter, you will see that a similar N-channel mosfet is very cheaper than the equivalent P-channel one.
Holes vs Carriers ??? So by using electrons as carriers how does that make it more efficient?? Also when you say cheaper, are you referring to the pricing ??
Based what I learned from an electrical engineer, n-channel is much faster since it paves the path to ground but it is the most dangerous.
Pretty much all reverse voltage polarity integrated circuits on the market uses p-channel. Thus I think this would be the case because safety comes first.
"lol, typo" ha nice ^^
Still there
Very nice explanation. I teach electronics at a local community college, and I will definitely suggest your TH-cam channel to my students.
This has made everything so clear, I have been trying to figure out p-channel mosfets for months now in a circuit simulator circuit simulator and getting really confused as to why things were not working the way I thought they would, and that's all because it doesn't show the ''parasitic diode'' in the mosfet.
This is a brilliant idea but one thing is not negligible though. In case of a high current load or even a short, the voltage drops at the drain port so that GS threshold voltage is not given anymore leading to a current flow through the diode and a resulting death of the fet.
This is only useful with a following current limiter.
what a gift to sell knowledge to others ! - world needs such teachers
THANK YOU. CLEAR. EXACTLY WHAT WE WERE TALKING ABOUT AT BREAKFAST; AND YOU'VE ALREADY DONE IT !!!
I WAS USING THIS IN SOLAR PANEL DESIGN ELIMINATING DIODE LOSS.
LOOKING AT N CHANNEL FOR LOWER RDSS. NICE TUTORIAL, THANKS AGAIN.
There is also a way to connect a diode in parallel with the load, but this will damage the power supply.
The method shown is a well-thought-out practical circuit.
Thanks for taking the extra time and effort to make your videos concise and succinct.
One of the best videos that I have seen about electronics in a while
Thanks
I use this trick for all my portable projects. Excellent protection from the effects of alcohol.
@Afrotechmods that is exactly what I saw in the sim, and would expect. Thanks for the videos. Your time and effort making these videos are greatly appreciated. Thank you.
In old era devices relay also was used. You place contact of relay which is normally open in series with circuit under protection and connect relay coil through diode, so it turns on and closes the contact only if power is connected on the right polarity. It also could be used as undervoltage protection.
Excellent in all aspects. You have more than earned my admiration, and subscription.
@colt4547 You absolutely can. But the downside is that you introduce a few milliohms of resistance on your ground rail, which may or may not be a problem depending on what you are doing.
Excellent solution, I was wondering about the wasted power of just using a diode. Thanks for a great answer at the right time
Excellent. Just what I need. I'd heard of doing this but didn't know which FET to use or how to connect it. I previously used a reverse connected diode, by accident I connected the battery in reverse one day, sure it passed a lot of current, blew the fuse and saved my circuit but it blew the end off the diode. Another diode needed.
This sounds ideal.
And I was sober at the time. It was moment of madness.
You alluded to it a bit, but I'm going to say it outright. The gate threshold voltage is just as important as the other parameters. Not only that, but the threshold voltage is specified for a certain amount of current. This is most important for low voltage supplies. If the load drops more current, the FET will act like a resistor, giving really bad losses. If it isn't heatsinked it can even overheat and fail. Usually you have to look at the transconductance curve. Hopefully there is one. Sometimes there is not, so it's a good idea to pick something that has a better datasheet unless you're sure the battery will have enough voltage at all times. Find where the load current and gate voltage meet on the curve and ensure that your battery or power source will never drop below that level.
In this very case, the MOSFET will never act as a straight up resistor, as it's always fully closed in reverse polarity, but in desired polarity, the body diode always limits the drain-source voltage and thus the losses for a given amount of current; additionally, the gate will partially open and reduce the voltage drop further. It's always an important consideration though. If you dimension the cooling of the MOSFET sufficiently, the battery voltage droop is not dangerous and it still beats a diode.
Of course you also want to beware of the voltage-current behaviour of the circuit behind it - some will consume about the same current independent of voltage, many will consume a lot less current, the lower the voltage is, and then some, switching regulators in particular, will consume the more current, the lower the voltage is, and the latter causes a potentially dangerous corner case.
As in all cases, a suitably rated fuse at the battery is good. If your load can sink that much current at low voltage, then adding circuitry to switch the MOSFET completely off below a threshold is needed. Its not too difficult a design task. Discharging the battery to low voltage may kill it too!
love ur sense of humor. great vid too! I love to blow things up. ran 5v through a 3v led, lens cap got hot, exploded sending the cap through the ceiling tile AUSOME!!!!!!
It should be noted that if you want to use this in a high efficiency design that protecting the mosfet using a zener diode is much less efficient because the zener diode needs at least a few milli amps to safely limit the voltage over the gate-source junction. As I'm designing an 18650 Li-Ion battery based protection circuit that doesn't just protect the circuit against reverse voltage but also protects the battery against overdischarge, I can't use a zener diode to protect the mosfet. But I don't really need to either, because the Si2371EDS I got a while ago has a maximum Vgs of +/- 12V, so it will be able to handle the 4.2V of the 18650 cell easily. And the 3.1V at which I'm planning to switch it off, minus the body diode maximum forward voltage drop of 1.2V, gives 1.9V which is still more than enough to switch it on and let the TLV3012B comparator handle disconnecting the load.
I love coming to your channel and videos to "wrap up" my understanding of various topics. I'm newbie and just getting back into using p-channel mosfets. I had an ok understanding of how to read the p-channel data sheet to choose the correct transistor for my app however after listening to this it makes complete sense - I wish all my teachers were like you "back in the day" (or at least I wish they had youtube when I was in college lol) - thank you!
When the diode in reverse mode we don't have 2A current flowing thru . Only very tiny leakage current can go thru it, that's normally considered as 0A . Therefore, the power dissipation across the diode in this circuit is also considered as 0W . The first circuit is good enough for reverse voltage protection . However, when the voltage is in forward direction, we will loose 2A X 0.7V = 1.4W if using silicon diode, or we will loose 2A x 0.3 A = 0.6W if using germanium diode .
Finally the physical explanation I was looking for. Many Thanks🙏
@LiKBeAsT "So when it's not ... choose which ever one we want (without being too supid:P) ?" Yes.
The mistake you are making with the battery is that you are viewing the + terminal as an absolute value of +12V. Voltage is always relative, between two points. The batt's + terminals is 12 volts higher than the batt's -ve terminal. So if we declare GND to be at the batt's + terminal, the -ve terminal has to be at 12 volts less than that.
With a bit more "tuning" the circuit could be set up to limit the current drawn by the load.
Another cheap way to prevent reverse voltage from damaging a circuit is put a relay in series. The coil is powered by forward biased diode. There will by more current/power absorbed by the relay's coil than by a P-FET, but in some cases it may be cheaper especially if the load is high and the source has power to "burn".
@R4ND0M1C That is, why you have to use at least the resistor in the gate, but best even the Zener, even for lower voltages then Vgsmax. That serve as a voltage limiter clamp (together with the MOSFET on-chip gate protection clamp and the gate capacitance, witch goes in nF range), so even with the ESD discharge, the gate voltage is clamped to way safe value.
Man, I'm just seeing this wondering how to do this earlier today, and I already sent out my PCB design for manufacturing! Such timing.
ONE of the BEST videos .Crisp and Clean.. Well done
In p-channel.. load is always connected to source.. when gate is pulled down to ground, the current flows from source and passes through the drain into the load. Mosfet is bi directional. Your video is trying to tell the opposite way if I am not wrong. Your explanation is pretty good. @ 2:39 if the mosfet is ON , then body diode turns OFF. @ 2:39 both are conducting which is a big flaw.. Body diode is ON when the FET is OFF... I am just trying to explain it. Your videos are awesome...
Awesome explanation, we do this at work but many an engineer just take it as tribal knowledge, it is refreshing to have the principal explained as the teacher would explain it to the pupil.
dazzling presentation. you're the electronics prof i never had. long line afrotechmods
You do such a great job with the presentation: comprehensive and thorough.
Thanks for covering the P-MOSFET VGS part as well. I corrected my design at the last moment. Thanks again.
Awesome tutorial, very clear and definitely not a waste of time, like many others out there! Thanks a million!
very good video. Everything explained was to the point. You must be the expert in Electronics. Thanks
Y aquí estoy yo 12 años después, disfrutando de este maravilloso vídeo por segunda vez. Es una lástima que este canal esté tan parado.
Good demonstration. Nice clear voice that is easy to understand. Looking forward to more of your videos. You have a new subscriber.
The easiest way to protect your circuit from reverse polarity is to add a rectifier bridge before the circuit input. Cheap and effective. Nice video though.
Very nice bit of info here. I always felt the simple diode (K) to input between B+ and a second diode also (K) to input from output, used for rev Pol. protection on Vreg IC's was suitable, still is of course since these are usually used in such low power circuits. However, since battery packs such as Lipo are now such a common supply for circuits, I really like knowing about this method. Now I need to put up a stock of P channel FET's and Zeners.
@LiKBeAsT I've thought about doing such a video. There are just too many things to make videos on... Anyways when we are not talking about ground meaning the earth's ground, ground can be anywhere you want in the circuit. It is a reference point. You can choose the ground node to be anywhere you want and do your design & analysis based around that. However, using certain nodes makes things easier than others. Most of the time it's easiest if GND = battery's -ve terminal.
MOSFETS are a great way to do this. The on resistance can be as low as 7mOhm for a P-Channel (Fairchild FQP27P06). Power = (0.01ohm) X 2^2 = 40mW loss (heat).
True. However I'd like to use N-channel-MOSFETs with their lower RDS-on and cheaper costs. Any idea?
You can do it with an N fet on the low side in a similar way just be aware that the resistance you are introducing on the ground path can cause problems in some circuits
have you found a good n-fet?
also since this is generally used in a floating charge circuit the ground is not a problem
Great for small currents and voltages drawn from batteries.For higher currents a bridge rectifier can be used or if the voltage is known and stable a relay with two independent contacts for making a dont care polarity input.
Thank you very much, someone FINALLY!!! did a video on this. I've been looking for this particular answer for a week or two by now. First one was the SINE WAVE-MADE PURE was my first & of coarse the second one fell right into order with what I needed. I subscribed so hopefully I'll get any new ones on these subjects. Thanks again, Tom.
Your videos are awesome. I was watching a few of them trying to connect the dots for a simple reverse voltage protection circuit that doesn't have a big forward drop (I'm splitting the output from a computer power supply to try and power a few ultra-low-power boards from one). Then I saw this, which makes those leaps for me. AWESOME. Also, if you think there's any huge red flags I should watch out for in my project, please say something. :) Otherwise, keep making videos for the rest of time.
super hard to find information oh HOW to use a diode but this was very clear and helpful and I finally understand what it does
Afrotechmods
Iv heard of this before, now I know how to do it, thanks! I will be needing this in the future. You make great tutorials. I never thought of doing this as I didnt know how, I knew about using a diode which is a bit lossy as you said. Of course I one used this to my advantage when something REQUIRED a 8.4V supply and I only had 9V. I believe I will watch all your tutorials, I could learn a lot form them.
Great video! I literally just ordered some assembled PCBs with diode protection just this morning before watching this. should have implemented the mosfets instead doh!
thanks for the great explanation
While its a good video I think the best protection is a physical one. Just make your power source unable to plug in backwards. Also the diode is more cost effective and fine for most DIY projects if the user is worried but its great to know other options as well and thats where this comes in.
@xanokothe cell phone chargers rectify mains voltage and pulses it through a really small transformer(somtimes just a little bigger than your thumb) with a transistor. This is just based on what Ive seen when I took one apart.
Very good video. Thanks! What happens if the load feeds back some current, e.g. a motor? Is the circuit reverse current proof?