An important extra note about 11:50 is that you have to be careful of that resistor value depending on the frequency you're operating at. When the transistor activates and the gate capacitance is discharged it is through the very small impedance of the saturated transistor. So it turns on quickly. But when the transistor turns off and the gate charges, it is through the resistor, and the time constant becomes something you have to worry about. The turn off will be slowed down. As you approach an operating period equal to that time constant the gate won't fully charge and the FET won't entirely cut off. It will spend more time in the linear region while cutting off and will warm. At a high enough frequency it will essentially never charge and the FET will permanently stay on. (Assuming a 1nF gate) Basically if it's anything higher than 10kHz I'd go no higher than 10k for a snappy gate response. And lower it as the frequency goes higher. Better yet add a BJT push-pull stage after the level-shifter and then the gate sees a low impedance to both the high and low side. Really that's what I would do for anything 100kHz or beyond. You're gonna reach a point where to keep the gate responsive at cutoff you'll need such a low resistor value that it will begin to stress the transistor. Plus the average current draw of the resistor-transistor branch could become high enough that it will be problematic for the longevity of battery operated devices. However, at the same time, if you're operating on a seconds/minutes/hours timescale and the transistor will be on or off for a long period of time, 1M should be fine.
A great way to remember which way the "arrow" points is this: "NPN" is Not Pointing iN. Then, of course, "PNP" would be just the opposite. Just remember NPN is Not Pointing iN, and then you'll have no trouble remembering......
Thank you. P-FETs make great Reverse Polarity Protectors = Worthy of a video. Would a PNP/3906 pull-up transistor be needed, for a RPP arrangement? Or, use a CMOS PFET?
You don't need pull-up transistors when using a p-channel mosfet for RPP. The only thing you need to ensure is that the gate doesn't get pulled down by more than its rated voltage, so you may have to clamp it with a 10V Zener diode. There's a circuit illustrating its use as an "ideal diode" in Stack Exchange at electronics stackexchange com/questions/588808/reverse-polarity-voltage-protection-using-p-mosfet (put the dots in to get a url), and the answers explain the operation.
The common problem on youtube when poeple explain transistors they only explain npn transistors and sometimes whe have a load grounded and only a pnp can be used and the curcuitry is different to have a constant supply of current when used with a battery But you explained both sides of transistors helping us young players
@@blackarrow8683 No, you can't swap drain and source on a mosfet. The construction of all modern mosfets means that there is a "body diode" (not a "protection diode") between the n-type substrate or body (which is then connected internally to the source) and the p-type drain. If you make the drain more positive than the source in a p-channel mosfet, the body diode will conduct regardless of the gate voltage. Therefore you only have normal fet operation ('triode' region) when the drain of a p-channel is more negative than the source.
An important extra note about 11:50 is that you have to be careful of that resistor value depending on the frequency you're operating at. When the transistor activates and the gate capacitance is discharged it is through the very small impedance of the saturated transistor. So it turns on quickly. But when the transistor turns off and the gate charges, it is through the resistor, and the time constant becomes something you have to worry about. The turn off will be slowed down. As you approach an operating period equal to that time constant the gate won't fully charge and the FET won't entirely cut off. It will spend more time in the linear region while cutting off and will warm. At a high enough frequency it will essentially never charge and the FET will permanently stay on.
(Assuming a 1nF gate) Basically if it's anything higher than 10kHz I'd go no higher than 10k for a snappy gate response. And lower it as the frequency goes higher. Better yet add a BJT push-pull stage after the level-shifter and then the gate sees a low impedance to both the high and low side. Really that's what I would do for anything 100kHz or beyond. You're gonna reach a point where to keep the gate responsive at cutoff you'll need such a low resistor value that it will begin to stress the transistor. Plus the average current draw of the resistor-transistor branch could become high enough that it will be problematic for the longevity of battery operated devices.
However, at the same time, if you're operating on a seconds/minutes/hours timescale and the transistor will be on or off for a long period of time, 1M should be fine.
You have a great way of explaining and visualizing things. Making it very easy to understand what is happening. Thank You !
Nice video! I use the logic-level MOSFETs like the IRLZ44 that can be driven with 5V.
A great way to remember which way the "arrow" points is this: "NPN" is Not Pointing iN. Then, of course, "PNP" would be just the opposite. Just remember NPN is Not Pointing iN, and then you'll have no trouble remembering......
I think of PNP as "Point iN Please"
Heard one youtuber say it and never had problem drawing or knowing which is emitter or if it is NPN or PNP
Always interesting and learning at the same time.👍🤓
Still loving your videos! Appreciate the detailed explanation on the differences between NPN and PNP, this video and explanation definitely helps! 👍🏻
Great examples!
Thank you. P-FETs make great Reverse Polarity Protectors = Worthy of a video. Would a PNP/3906 pull-up transistor be needed, for a RPP arrangement? Or, use a CMOS PFET?
You don't need pull-up transistors when using a p-channel mosfet for RPP. The only thing you need to ensure is that the gate doesn't get pulled down by more than its rated voltage, so you may have to clamp it with a 10V Zener diode. There's a circuit illustrating its use as an "ideal diode" in Stack Exchange at electronics stackexchange com/questions/588808/reverse-polarity-voltage-protection-using-p-mosfet (put the dots in to get a url), and the answers explain the operation.
@@RexxSchneider Thank you!!!
The common problem on youtube when poeple explain transistors they only explain npn transistors and sometimes whe have a load grounded and only a pnp can be used and the curcuitry is different to have a constant supply of current when used with a battery
But you explained both sides of transistors helping us young players
As shown, under what conditions would the p-Chan’s drain to source diode conduct?
if you drive an inductive load you can get back EMF
This package cannot dissipate much power. To drive high current, you need to turn it on completely and you need to transition very quickly.
A question because I don't know the answer. In the example with the arduino, could you not just have a 10k pull up resistor rather than a BJT?
not good to connect 12v to a 5v chip
What is the little PCB board that the potentiometer is mounted to, and where can I get one on line?
just a protoboard that I cut down on the bandsaw.
13:01 I am stilla newbie, but I think in your drawing the source should be at the top (so a connection from the middle to the top)?
Or, does it really matter? Can you swith Drain and Source? Hmm..
some JFET you can swap D and S but not the normal fets
@@blackarrow8683 No, you can't swap drain and source on a mosfet. The construction of all modern mosfets means that there is a "body diode" (not a "protection diode") between the n-type substrate or body (which is then connected internally to the source) and the p-type drain. If you make the drain more positive than the source in a p-channel mosfet, the body diode will conduct regardless of the gate voltage. Therefore you only have normal fet operation ('triode' region) when the drain of a p-channel is more negative than the source.
@@RexxSchneider Thank you. That makes sense!
The reverse current protection diode is drawn as a zener, right? Is it ever used that way?
drawn as Schottky
@@IMSAIGuy oops
@@IMSAIGuy wait a minute. commons.wikimedia.org/wiki/File:Zener_diode_symbol-2.svg#/media/File:Zener_diode_symbol-2.svg
commons.wikimedia.org/wiki/File:Schottky_diode_symbol.svg#/media/File:Schottky_diode_symbol.svg
"Big load"
May you please explain MagAmp theory ?
don't now anything about them