Great video, excellent description and examples. I have been studying the basics of mosfets the past week and this was a great next step. Thank you for the time you give putting your videos together.
Baruch is using IEEE standard notation for the arrow direction (current flow). This is opposite (what Im used to seeing) that is shown in many circuit diagrams. He might have mentioned this.
Loads of times I've heard people talk about the existence of enhancement mode and depletion mode... I think this is the first time I've heard anyone explain what that actually means. :)
Tnx a lot, great video. In NMOS circuit, you always connected source to ground, led on seperate line. Is this common design or is it possible to connect resistor and led to the source of NMOS? So there is only "one line"?
I think if there insiries want the led be fluctuating in voltage because of the spike that happens when the gate is opened and closed, or I've completely misunderstood it which is more likely 😀
Great tutorials, thank you, Robert! Are you planning to go into other (non-logic) uses of FETs, or are you mostly interested in their application to logic circuits building?
The cmos explanation isn't clicking with me. In the diagram... You could totally remove the bottom nmos and the power would be the same whether the pmos is on or off. What am I missing here?
I don't think your comparison between the NMOS and PMOS configurations is accurate, because you are changing how you drive the load. In the PMOS configurations you are either turning on or off a switch in series with the load. This is like a traditional switch and light bulb. However your NMOS configuration has the transistors in parallel with the load which you are shorting out to turn off the LED. In this way you are adding an inverting component in one configuration but not the other. If you kept the LED in series between the resistor and the transistors, the series NMOS would be an AND gate compared to the PMOS's NOR. likewise, a parallel PMOS is a NAND, but the parallel NMOS is an OR. Basically switching only the gate channel type (N vs P), but leaving the load in series would swap the gate logic and invert the output (AND NOR, or OR NAND).
Adam Butt Yeah, that’s a sort of an unnecessary complication. You can drive LEDs that way in some applications where constant current is important (eg to avoid LED switching from polluting an analog supply), but even then what you’d need to do is a bit different since Vds is normally much smaller than the LED dropout. I had a micro power, precision all-analog circuit that had some led indicators and the on/off switching had to be done by having two identical LEDs in thermal contact, and switching between them. That way, the LEDs could be driven without a current source. The design was careful in keeping the “on” (LED A) and “off” (LED B) active for the same total time, so that they’d age identically - that was 2+ decades ago when LEDs aged much faster than they do today, since to get good light output you had to drive them much closer to their limits. Modern high brightness LEDs can be often driven at 1mA or less and still be plenty visible as indicators (although this is very application dependent of course).
Hi won't running the LEDs in series make the voltage spike in the led when the gate changes state or is there a way around this and don't MOSFETs only conduct when there switching state would having an led in series make them have a voltage all the time giving incredible power losses using them that way?
Robert Woods Mosfets really are electronic switches. When they open, they open pretty good. You have to understand that even mechanical switches aren’t perfectly open. Most unsealed switches (say a light switch on the wall) have enough contaminants accumulated on the “insulating” surface between the contacts, that even when “open” you can measure leakage current through them. It’s minuscule and requires good equipment to measure, but it’s there. Mosfets aren’t that good usually, but they still aren’t shabby. You won’t measure a mosfet leakage current at room temperature with a common multimeter set to 200mA range. So, you can think of building a circuit with an actual switch, and then replacing it with a mosfet. Of course all electronic switches will have some kind of transient no ideal behavior, but it’s not hard not to produce voltage spikes in a circuit that turns a LED on and off. Since you use the mosfet as a switch, you need something else to generate the current for an LED: LEDs are current driven and connecting them to an ideal voltage source either does nothing or destroys them in short order. The cheapest way to generate some current is to use a resistor. As long as the supply voltage and the LED diode drop are somewhat stable, the current will be “constant”. That resistor will also damp any inductive transient effects you may have due to fast mosfet switching interacting with lead resistance - especially if the LED circuit is a long. Finally, the mosfet can also be used as a voltage-controlled resistor. In that case, a feedback controller (can be just an op-amp) measures the current flowing through the LED, and adjusts the mosfet have the resistance necessary to generate that current. This is a much more ideal current source than a fixed resistor. The mosfet will dissipate the same power a fixed series resistor would (if it dissipates more, you’ve got a local oscillation going on and need to see what’s up on the mosfet’s pins with a scope). Further, Mosfets in integrated circuits are small enough that they become simple current sources. Usually, discrete Mosfets are much larger and you will destroy them before they start limiting current. But a digital CMOS chip - a microcontroller or a logic chip - will inherently limit current on the output pins. For example, 74AHS logic family drives about +/-8mA into a short. So, as long as you stay within the limits of the heat the chip can dissipate, and within the current limits on the pins (and the total limit), you can drive an LED directly from the digital pin, and you can parallel the outputs for higher drive current, too. You can think of these “inherent” mosfet current sources as self-controlling resistors. Thus, any such current source will dissipate the power equal to the product of the current it generates and the voltage across it. Example: a 74AHC chip running from 3.3V output drives an LED with 1.2V forward voltage. The output mosfet transistor will drive about 6mA (say), so the dissipation is (3.3-1.2)V*0.006A=0.012W=12mW. If that’s all the chip does, it’s well within it’s thermal dissipation limits and within its current limits too. No resistors needed. I find that many people are unaware of this inherent current limiting characteristic of small mosfet transistors, yet it is always taken into account when doing analog CMOS IC (chip) design.
@@absurdengineering hi many thanks for your reply, I'm not trained in electronics so trying to grasp it from these videos is difficult sometimes. I'm really grateful for any new knowledge I receive and you have explained a few more things for me. I only started to learn because when looking at a circuit board it p#@'s me off that I don't understand what's going on and the more I seem to learn the more I realise how little I know, so sorry if I do seem abit ignorant on the subject I don't mean to be but with little understanding I'm going to sound it anyway. Thank you and I will keep everything you said in mind when watching other videos 🙂👍
Robert Woods No problem. I don’t know a whole lot either - and it took me a good while before discrete transistors weren’t “black magic”. I could build fairly complicated circuits as a teen yet I was really lacking in understanding of some basics, and had to fill in the gaps much later while already working as an engineer. The resources we have today make it easier, because you’re more likely to find materials with teaching style that matches your needs and preferences.
Not impressed with this video. You are talking about digital switching and digital logic - and touching on MOSFET function. I learned nothing on how you use MOSFETs, part one was better in this regard. I want to know how to choose a MF how to read and interpret their data sheets,any quirks of these devices how to use with a uP how to cascade them to switch high currents etc NOT how to make a logic gate.
Great video, excellent description and examples. I have been studying the basics of mosfets the past week and this was a great next step. Thank you for the time you give putting your videos together.
Thank you for explaining so well and showing me everything.
Thanks! I see I'm a late comer but I'm so glad to have found you! The old brain needs that!
Well done!!!! Simplifying concepts that otherwise would be difficult to understand!!!
Baruch is using IEEE standard notation for the arrow direction (current flow). This is opposite (what Im used to seeing) that is shown in many circuit diagrams. He might have mentioned this.
you are the best. thank youuuuuuu
Loads of times I've heard people talk about the existence of enhancement mode and depletion mode... I think this is the first time I've heard anyone explain what that actually means. :)
Thank you
thanks for that insights :)
what is the correct symbol of NPN and PNP FET? I think it is opersit of transistor
Tnx a lot, great video. In NMOS circuit, you always connected source to ground, led on seperate line. Is this common design or is it possible to connect resistor and led to the source of NMOS? So there is only "one line"?
I think if there insiries want the led be fluctuating in voltage because of the spike that happens when the gate is opened and closed, or I've completely misunderstood it which is more likely 😀
thanks a bunch
superb
Great tutorials, thank you, Robert!
Are you planning to go into other (non-logic) uses of FETs, or are you mostly interested in their application to logic circuits building?
Pretty much exclusively logic.
The cmos explanation isn't clicking with me. In the diagram... You could totally remove the bottom nmos and the power would be the same whether the pmos is on or off.
What am I missing here?
a bit confusing, but overall very useful video, thanks
I don't think your comparison between the NMOS and PMOS configurations is accurate, because you are changing how you drive the load. In the PMOS configurations you are either turning on or off a switch in series with the load. This is like a traditional switch and light bulb. However your NMOS configuration has the transistors in parallel with the load which you are shorting out to turn off the LED. In this way you are adding an inverting component in one configuration but not the other.
If you kept the LED in series between the resistor and the transistors, the series NMOS would be an AND gate compared to the PMOS's NOR. likewise, a parallel PMOS is a NAND, but the parallel NMOS is an OR. Basically switching only the gate channel type (N vs P), but leaving the load in series would swap the gate logic and invert the output (AND NOR, or OR NAND).
Adam Butt Yeah, that’s a sort of an unnecessary complication. You can drive LEDs that way in some applications where constant current is important (eg to avoid LED switching from polluting an analog supply), but even then what you’d need to do is a bit different since Vds is normally much smaller than the LED dropout.
I had a micro power, precision all-analog circuit that had some led indicators and the on/off switching had to be done by having two identical LEDs in thermal contact, and switching between them. That way, the LEDs could be driven without a current source. The design was careful in keeping the “on” (LED A) and “off” (LED B) active for the same total time, so that they’d age identically - that was 2+ decades ago when LEDs aged much faster than they do today, since to get good light output you had to drive them much closer to their limits. Modern high brightness LEDs can be often driven at 1mA or less and still be plenty visible as indicators (although this is very application dependent of course).
Hi won't running the LEDs in series make the voltage spike in the led when the gate changes state or is there a way around this and don't MOSFETs only conduct when there switching state would having an led in series make them have a voltage all the time giving incredible power losses using them that way?
Robert Woods Mosfets really are electronic switches. When they open, they open pretty good. You have to understand that even mechanical switches aren’t perfectly open. Most unsealed switches (say a light switch on the wall) have enough contaminants accumulated on the “insulating” surface between the contacts, that even when “open” you can measure leakage current through them. It’s minuscule and requires good equipment to measure, but it’s there. Mosfets aren’t that good usually, but they still aren’t shabby. You won’t measure a mosfet leakage current at room temperature with a common multimeter set to 200mA range. So, you can think of building a circuit with an actual switch, and then replacing it with a mosfet.
Of course all electronic switches will have some kind of transient no ideal behavior, but it’s not hard not to produce voltage spikes in a circuit that turns a LED on and off. Since you use the mosfet as a switch, you need something else to generate the current for an LED: LEDs are current driven and connecting them to an ideal voltage source either does nothing or destroys them in short order. The cheapest way to generate some current is to use a resistor. As long as the supply voltage and the LED diode drop are somewhat stable, the current will be “constant”. That resistor will also damp any inductive transient effects you may have due to fast mosfet switching interacting with lead resistance - especially if the LED circuit is a long.
Finally, the mosfet can also be used as a voltage-controlled resistor. In that case, a feedback controller (can be just an op-amp) measures the current flowing through the LED, and adjusts the mosfet have the resistance necessary to generate that current. This is a much more ideal current source than a fixed resistor. The mosfet will dissipate the same power a fixed series resistor would (if it dissipates more, you’ve got a local oscillation going on and need to see what’s up on the mosfet’s pins with a scope).
Further, Mosfets in integrated circuits are small enough that they become simple current sources. Usually, discrete Mosfets are much larger and you will destroy them before they start limiting current. But a digital CMOS chip - a microcontroller or a logic chip - will inherently limit current on the output pins. For example, 74AHS logic family drives about +/-8mA into a short. So, as long as you stay within the limits of the heat the chip can dissipate, and within the current limits on the pins (and the total limit), you can drive an LED directly from the digital pin, and you can parallel the outputs for higher drive current, too.
You can think of these “inherent” mosfet current sources as self-controlling resistors. Thus, any such current source will dissipate the power equal to the product of the current it generates and the voltage across it. Example: a 74AHC chip running from 3.3V output drives an LED with 1.2V forward voltage. The output mosfet transistor will drive about 6mA (say), so the dissipation is (3.3-1.2)V*0.006A=0.012W=12mW. If that’s all the chip does, it’s well within it’s thermal dissipation limits and within its current limits too. No resistors needed. I find that many people are unaware of this inherent current limiting characteristic of small mosfet transistors, yet it is always taken into account when doing analog CMOS IC (chip) design.
@@absurdengineering hi many thanks for your reply, I'm not trained in electronics so trying to grasp it from these videos is difficult sometimes. I'm really grateful for any new knowledge I receive and you have explained a few more things for me. I only started to learn because when looking at a circuit board it p#@'s me off that I don't understand what's going on and the more I seem to learn the more I realise how little I know, so sorry if I do seem abit ignorant on the subject I don't mean to be but with little understanding I'm going to sound it anyway.
Thank you and I will keep everything you said in mind when watching other videos 🙂👍
Robert Woods No problem. I don’t know a whole lot either - and it took me a good while before discrete transistors weren’t “black magic”. I could build fairly complicated circuits as a teen yet I was really lacking in understanding of some basics, and had to fill in the gaps much later while already working as an engineer. The resources we have today make it easier, because you’re more likely to find materials with teaching style that matches your needs and preferences.
U said Play. Not Formulate!!!!!
Not impressed with this video. You are talking about digital switching and digital logic - and touching on MOSFET function. I learned nothing on how you use MOSFETs, part one was better in this regard. I want to know how to choose a MF how to read and interpret their data sheets,any quirks of these devices how to use with a uP how to cascade them to switch high currents etc NOT how to make a logic gate.
Ray Well, if you want to design power circuits - why would you watch a mos logic design video? You need to watch power mos design stuff :)