Q: Understanding MOSFET characteristics

I am trying to select a N-Channel MOSFET to switch a 12V power rail. The use case here is to switch on/off a couple of LEDS which are wired in series. The LEDs (datasheet) can draw a maximum of 700mA. I want to control the MOSFET via a MCU (3,3V logic level). I am now struggling to choose a suita…mosfetcomponent-selectiontlfong01For prototyping, I would suggest to try TO220 through hole 3V3 logical level Vgs(th) IRL540N N-channel power MOSFET: infineon.com/cms/en/product/power/mosfet/…. It might be risky to use SMD device. Bimpelrekkie72.1kWhat does xyz mean? What do you think it means? What do you think would happen of xyz was very small/large? Also it is strongly recommended to add a schematic so that we can see HOW you will be using the MOSFET. Your “What does xyz mean?” tells me you’re not that experienced so in order to give sound advice I need to see how you’re going to use the MOSFET. For that, include a schematic. Some xyz are more/less important depending on how you use the MOSFET. OYPS@tlfong01 thanks for your reply. I will try that one for prototyping.@Bimpelrekkie thanks again for your reply. I will add a schematic shortly if it helps to clarify the question Andy akaAn N channel MOSFET is an inappropriate choice for controlling a 12 volt power rail unless you are prepared to add a voltage boost circuit. It’s better to control the 0 volt rail with an N channel MOSFET. Bimpelrekkie72.1kI agree with starting with a commonly used MOSFET like the IRL540. Chances are that it will do the job. As a beginner you’re over-worrying about selecting the right component while experienced designers (like me) know that many commonly used devices can do the job (assuming you’re not doing anything extraordinary). Also beginners often over-worry about the components but then use the wrong circuit!What Andy mentions is a extremely common trap for beginners that 9 out 10 fall into (and then come here and ask why it doesn’t work). And that’s why we need to see a schematic. Are you making a source follower or not? Andy akaFeb 6, 0:37How are you going to solder the LEDs? What are you going to mount them on? tlfong015179#OYPS. I usually strongly recommend power LED lamp newbies, before starting to do any circuit design, to read the following: (1) One Watt LED – Components 101 2018mar17: components101.com/diodes/1-watt-led. After reading this short, newbie friendly tutorial, you might like to let me know if there was at least one important thing that you didn’t know that you didn’t know. Happy reading. Cheers.#OYPS, you might also like to watch how other guys are playing with power LEDs. (2) Power LED’s – Simplest Light With Constant-current Circuit – dan, Monkeylectric, instructables.com/… Happy reading. Cheers. OYPS175@tlfong01 thanks for your useful articles.@Bimpelrekkie I have added an example schematic to the question. I hope this helps to clarify things.@Andyaka they will be reflow soldered on to a pcb. Andy akaAre you hoping to drive the LEDs at around 100 mA @OYPS OYPS@Andyaka no, i am pretty sure they will draw more than that. Maximum of 700mA. Why? Andy aka330kWell, 700 mA through an 18 ohm resistor drops 12.6 volts and that’s more voltage than the supply. Whereas 100 mA through an 18 ohm resistor drops 1.8 volts leaving 10.2 across 3x LEDs or 3.4 volts each. You also need to specify which LED specifically you are using.And, 700 mA is the absolute maximum rating where they may get destroyed if you go a little bit more. You should aim for no more than 500 mA. OYPS175Feb 6, 0:37@Andyaka You are right. The schematic had the wrong Resistor value. It should be 1,8Ohm as calculated via this amplifiedparts.com/tech-articles/led-parallel-series-calcula‌​tor. The LEDs datasheet is listed the question. I can increase the value a bit to limit the current to 500mA. Thanks for pointing this out. BimpelrekkieYou didn’t try to use a source follower which is good. You’re switching the – side with an NMOS also good. Re-calculate that resistor and also how much power it is going to dissipate! It will need to be able to handle that or you will get a burned resistor. Hearth16.5kYou say that Vth should be smaller than your logic level. Good! That’s right. But also it must be much smaller than your logic level. A common pitfall people fall into is thinking that Vth is the voltage at which the FET can be considered fully on, when that’s not true at all. It’s the voltage at which it just barely begins to conduct, and depending on the FET you may need to be as much as several times Vth to turn it fully on. You want a FET with an Rdson specified at a Vgs of your logic level or less, and specified to be low. OYPS175@Hearth thanks for pointing that out. Is the following MOSFET suitable for this application: lcsc.com/product-detail/…. I selected it on behalf of your information, selecting low Rdson at the 3,3V logic level.I updated my question with a MOSFET suggestion based on your inputs.I updated my question with a MOSFET suggestion based on your inputs. @Bimpelrekkie I updated the resistor and calculated the power dissipation. The resistor should be able to handle 2W. Thanks for pointing that out again!  3 hours later… user289103386Feb 6, 4:03Regarding the DMN10H120SE MOSFET… If you look at figure 2 on the data sheet, “Typical Transfer Characteristics” you will notice that at 3.3V, the device is barely on, if at all. You really should have minimum 4.5V gate-source. Drain-source resistance for this device is spec’d at 6V and 10V, btw  8 hours later… tlfong015179Feb 6, 12:27@user28910 Yes, I fully agree. It is not just marginal, but out of margin.

@OYPS Some more references you might find useful:(1) What’s the voltage used to power LEDs inside an LED light bulb?
electronics.stackexchange.com/…(2) Control a 3W RGB LED with an Arduino
electronics.stackexchange.com/…(3) Constant current source for resistance measurement
electronics.stackexchange.com/…(4) Constant Current Source Learning Notes V0.1 – tlfong01 2021feb02
penzu.com/public/bda5986a  6 hours later… OYPS175Feb 6, 18:56@tlfong01 thanks for the links. I did not read all of them yet, but they were already pretty useful. After getting all the answers to my question i was able to filter the existing MOSFETs much more precise. Especially in regards of Rdson. Is an Rdson of 250Ohm@2.5V suitable for my application. I was not able find one at 2.7 or 3.3V.For example this one from Infineon: mouser.de/datasheet/2/196/…OYPSFeb 6, 19:15250mOhm@2.5V of course.  20 hours later… tlfong015179Feb 7, 15:27@OYPS Ah, 2.5V is just a “benchmark” for you to “extrapolate“. I think your choice is already very good. It would be perfect if you can find something similar, but with a DPAK package which has a SMD heatsink. 🙂

  3 hours later… tlfong015179Feb 7, 18:10@OYPS So far we are discussing on low side current switching. You might like to explore high side switching, as shown below.

Yet there is one more switching configuration, or constant current switching/limiting/source called “True Floating CCS”, which sits between Vcc and ground, in other words, and does not connect to Vcc or ground, as referred above:(4) LM334Z CCS Learning Notes V0.1 – tlfong01 2021feb02
penzu.com/public/bda5986atlfong015179Feb 7, 18:29@OYPS This is P-channel power MOSFET, also very low level Vgs(th) which you might consider high side switching.

  9 hours later… OYPS175Feb 8, 3:39@tlfong01 Thanks for your reply. What is the advantage/disadvantage of high or low side switching ?On which use case would I choose on over the other?Are the 250mOhm a problem? Some answers on the original question suggested, that Rdson could/should be lower than 200mohm. As far as I understand it, I should just try to* minimize this  8 hours later… tlfong015179Feb 8, 11:38This might help：High-side versus low-side switching – Lednique
lednique.com/gpio-tricks/…And for ninjas only: GPIO high-side driver fail – Lednique
lednique.com/…tlfong015179Feb 8, 11:59Well, comparing to the power NPN BJT such as Darlington TIP120/122 with high Vce(sat), power MOSFET’s ridiculously low Rds(on) is too good to be true. You might like to read the following to get a rough idea of what I am talking about:
Determine if a heatsink is required (Darlington TIP120)
electronics.stackexchange.com/…. tlfong015179Feb 8, 12:12So for low currents, say, less than 1A, you don’t need to bother Rds(on). What you need to care is the current limiting series resistor of the power LED in your inefficient, power wasting circuit. And that is why I always strongly recommend the power LED newbies to at least use the much more efficient “Two NPN BJT current mirroring, negative feedback CCS (See Wiki on ‘Current Source’)”.  4 hours later… tlfong015179Feb 8, 16:15@OYPS Moreover, you might need first to decide your constant current spec, say NiChia 500mA at 500mW radiant flux, or start with the Cree commercial/industrial pseudo standard of 350mA 3W. I only play with 1W toys. So no guarantee I am not speaking nonsense. 🙂 tlfong015179Feb 8, 16:26And do you have any (1) Current sensor amplifier, eg MAX4172 electronics.stackexchange.com/… (2) UV light sensor: AliEXpress 2.7V-5.5V Uv Sensor Module
nl.aliexpress.com/item/…, to do CCS negative feedback experiments?(see full text)  8 hours later… OYPS175Feb 9, 0:09@tlfong01 thanks.I will work through your links ans articles to get hang of what you are talking. Of course I am always trying to optimize my circuit, but just to a certain extent. As my general design is connected to an external power supply and not power by anything like a battery, I am not very concerned about power consumption.But if my design is inefficient and generally wasting power, I will try to change that. I will report back after looking at your material! OYPS175Feb 9, 0:30@tlfong01 @tlfong01 you are talking about this?

how would this make my circuit any more power efficient?OK ignore my question. I now understood the wiki entry. This seems to be more complex but more efficient approach to my initial problem. But i never used such a circuit before and therefore do not know how to adapt it to my use case. especially how to calculate all the resistor values. I will research on this a bit more OYPS175Feb 9, 1:03I additionally found this circuit example electroschematics.com/… with a digital control, which seems to be best fit for me. any thoughts?

  8 hours later… tlfong015179Feb 9, 9:16@OYPS I know you are using an external power supply, but I think you missed the point. I would suggest you to read Components 100’s short article on 1W LEDs one more time, and then read all the rest references. One more thing is important is the user requirement and functional spec of the UV power LED lamp. Let me casually and randomly scribble something.Suppose I want to DIY a UV LED lamp for my cat’s home. My preliminary cat’s requirement and functional spec is summarized below: (1) 350nm, (2) 300mW, (3) 200mA, (4) 15 min daily, (5) 12V Vcc, …As told by Components 101, if you use the amateurish one transistor, one current limiting resistor approach, if the voltage across the LED varies by 0.1V, current would vary drastically, perhaps from 200mA by 50% to 400mA (for exact numbers, see my I-V test results), so it is UV overdose problem, which might shorten my cat’s life. Or you are using UV kill germs, viruses, or make green fruit to ripe faster, then over UV exposure might make the fruit ripen too soon, …Errata – 200mA by 50% would be 300mA. Note – I am too slow to make corrections to my chat comments, so be warned that there are some important errors I could not correct in time, or delete.  2 hours later… tlfong01Feb 9, 11:29@OYPS Yes, you see two NPN BJT transistors, me don’t see no transistor, me only see current mirror and negative feedback CCS.  1 hour later… tlfong015179Feb 9, 12:46@OYPS Yes, your video explaining CCS is good.LED Constant Current Source – ElectroSchematics
electroschematics.com/…Led constant current source►The tutorial gives the following equations:R (Ω) = I (mA)/0.5

The power dissipation of R is

P = I2 (A) x R (Ω)So you know how to calculate Rsense = I/0.5VIn your case I = 700mA, so Rsense = 700mΩ / 0.5V = 1400mΩ = 1.4ΩThe tutorial uses digital control. But we can use PWM control. To make things as loosely coupled as possible for prototyping and troubleshooting/testing/calbration, we can use PWM high/low side switching complimented by with low/high side CCS, also MAX4172 current sense amplifier, and UV sensor for feedback. tlfong015179Feb 9, 13:10@OYPS Yes, the handsome guy P Marian is indeed a good story teller. 🙂

  5 hours later… OYPS175Feb 9, 18:31OK. I think i am slowly getting a hang of it. Using the article above i created the following circuit. As far as i understand it, i am able to control my LEDs via my MCU. PWM is not an option for me, as the MCU has not enough PWM channels to support this.

Is this circuit correct? Via this, I am gaining a more power efficient circuit ?  15 hours later… tlfong01Feb 10, 9:18@OYPS (1) Not sure, at least no forum’s reputation would be damaged. (2) If you are not sure, that means you have not understood the stuff.  1 day later… tlfong015179Feb 11, 15:15[1] ZXLD1350 30V 350mA LED DRIVER with AEC-Q100 Product Info – Diodes.com
diodes.com/part/view/ZXLD1350

[2] ZXLD1350 30V 350mA LED DRIVER with AEC-Q100 Datasheet – Diodes.com
diodes.com/assets/Datasheets/…  5 hours later… OYPS175Feb 11, 20:05@tlfong01 thanks for your reply. I think your provided LED driver is quite good, but I do not feel comfortable with the maximum rating of 350mA. I have to be aware of the absolute worst case pulse forward current of up to 1A the LEDs can draw. I am currently prototyping the last schematic i send to verify it, but on behalf of LED drivers i would rather go with something like this: datasheet.lcsc.com/szlcsc/…  1 hour later… tlfong015179Feb 11, 21:33@OYPS Yes, I very much agree with you about the 350mA limitation. On the other hand, I hesitate to support you on the P Marian circuit for a couple of reasons, including the following: (1) The driving transistor is power NPN BJT, which is very power inefficient, comparing with N-channel power MOSFET.(2) The circuit design looks a bit amateur or casual, eg, using MOSFET 2N7000 to control bipolar circuits, looks weird. (3) I am pretty sure that this seemingly not too professional circuit can be improved.Coming back to the buck concurrent source chip you suggested:LM3405 1.6-MHz, 1-A Constant Current Buck Regulator For Powering LEDS – TI, 2016
datasheet.lcsc.com/szlcsc/….@OYPS I am very glad that you have found this LED driver. I skimmed the datasheet and I am very impressed. All in all, for 1W/350mA, 3W/1A power LED applications, I would give your LM3405 five stars, and my ZXLD1350 only 4 stars.Both your and my recommendation have all the constant current source LED driver features I want, as highlighted in the following picture:

My quick and dirty conclusion is that even my recommended ZXLD1350 can do 1A max, I still think your recommended LM3405 is the best, or almost perfect for your application.Ah, bed time. See you later. Cheers.  21 hours later… tlfong015179Feb 12, 18:22@OYPS More references:[1] How can I figure out what the forward current and forward voltage will be for an LED given a resistor size Asked today Active today Viewed 69 times
electronics.stackexchange.com/…

[2] ***What is the formula for the dependence between modulating input voltage – output current for two transistor constant current sink*** Asked 4 years, 1 month ago
Active 5 months ago Viewed 779 times
electronics.stackexchange.com/……(see full text)  15 hours later… tlfong015179Feb 13, 9:26@OYPS References to 70mA LED with built in series current protecting resistor.LED with built in resistor: electronics.stackexchange.com/…70mA LED notes: (1) … The LED is sort of super bright with a built in resistor, so that you can directly apply 5V and get say, 70mA without any problem. If you supply 3V3 the current might be smaller, but might not be much smaller. (2) the built in resistor might be much bigger than 100R, so with or without the series resistor does not make much difference to human eyes. – tlfong01 20 hours ago
References: (1) Red LED 5mm with Resistor (2V ~ 5V, 2000 ~ 3000 MCD, 620 ~ 625NM) Product Sheet – SparkFun $9/25pcs sparkfun.com/products/14560, (2) YSL-R531R3c-5V Red LED 5mm with Resistor Datashe…(see full text)Errata – References: [1] sparkfun.com/products/14560,
[2] cdn.sparkfun.com/assets/c/9/b/…,
[3] imgur.com/gallery/GkKqwEo.  1 day later… tlfong015179Feb 14, 18:56@OYPS I am jealous that more and more guys are using 700mA LEDs. So I am thinking of going to the 700mA grade LM3405.

  11 days later… tlfong01517912:41@OYPS Our discussion seems coming to a close. I am glad to conclude that we have learned something, and found 350mA and 700mA LED drivers. I have ordered both drivers from RS and might test them out later. But my long term plan is to DIY a over 1A LED driver using N-channel power MOSFET. One thing I still need to dig deeper before starting to design is the Rds(on) and the real turning on process. One interesting relevant article I am reading is the following:Power MOSFET Basics: Understanding the Turn-On Process – Vishay:
vishay.com/docs/68214/…. Have a great project. Cheers.