Look up those MAX745 lithium charge modules! You can connect one directly to 20V and it steps the voltage down to 1-4S for your battery, has an adjustable charge current. Much easier and efficient that way. Also, with the heatsink inside and without fan it will definitely overheat the leds very quickly.
the IC supports this, the question is whether the board does. the input capacitors also need to be rated for that voltage, and since it's a buck convertor, a larger voltage drop means more ripple current for a given inductor value, so it affects the calculations for the inductor and output capacitor as well. the chip is only part of the picture
The supplier that these boards come from has 20v modules, but this project was rather time rushed, and as I live remote it would be hard to get them in time
6:43 so, a couple things: usually a MOSFET's gate threshold voltage is pretty low, it'll start turning on (just barely, meaning a high drain-source resistance) with just a couple volts, and then the drain-source resistance approaches its minimum at maybe 10 or 12 Vgs. You really need to read the datasheet of a MOSFET to have any idea what a given gate-source voltage will do. also: ideal MOSFETs don't automatically turn off when you disconnect the gate (see the end of the comment for why real MOSFETs sometimes do anyways, and why this shouldn't be relied on). the problem is, a MOSFET gate is basically a voltage sensor, it's a small capacitor that detects its level of charge and turns the switch on and off. When put your hand on a Van De Graaf generator and your hair stands up, it doesn't immediately flop down when you let go, because you are still charged; you have to touch something grounded, causing a spark of current equalizing your voltage with the voltage of ground. similarly, an ideal MOSFET gate doesn't consume the charge in any way, it just sits there, the mere presence of a charge difference between the gate and the source passively activates the switch. so when you disconnect the drain from the voltage supply, it'll keep chilling there. This is great for efficiency, it means that you don't need to consume any power to keep the MOSFET turned on, you put in a small amount of energy to *turn* it on and then it'll just stay on, modern computers would not extremely inefficient without this advantage. in your application, where you just want a simple circuit to make a switch turn on and off, this doesn't matter and is just an annoying problem you have to solve. The solution is to connect a resistor between the gate and the source, something in the 1k-10k range probably. That way the gate-source capacitance is constantly being discharged, and requires a constant stream of current to keep it topped up, so if you cut off the power it'll discharge at a predictable rate and quickly drop below the threshold voltage, turning it off. You can check the datasheet for the gate capacitance and gate threshold you can calculate how long it'll take to turn off for a given voltage and resistance, but practically for your use case 10k should turn it off plenty fast enough while only wasting a negligible amount of power (as per ohm's law, at 12V a 10k resistor draw 1.2mA of current, 1.2mA * 12V = 14.4mW). This is often called a pull-down resistor, because it's "pulling" the gate-source voltage towards zero. when you connect the gate to a voltage supply, it's a much stronger connection so it "wins", "pulling" the voltage up to 12. when you disconnect it, the pull-down resistor "wins" by default. In practice, nothing is a perfect insulator, so the imperfect insulation between the gate and source of a MOSFET can be thought of as a "built-in resistor" between the gate and the source. However, this is an undesirable property, and manufacturers are always trying to maximize this resistance. the only spec for gate-source resistance you'll find in a datasheet, if any, is a *minimum* value, so if you make something that relies on it being at most a particular value, it's either not gonna work, or be up to the silicon lottery, random environmental factors, the electricity gods, etc. essentially, a real-life MOSFET will turn itself off when the gate is disconnected, but you can only calculate the *minimum* amount of time it'll take, in practice you never know if it'll take 100us, or 3 seconds, or 3 hours. and if it does take 3 seconds, it's gonna spend a lot of time in the in-between ("linear") region, where it's switched on but its drain-source resistance is pretty high, causing a large share of the power of the circuit to dissipate in the MOSFET, and it'll blow up. if you don't like this weirdness about charge and voltage, you can use bipolar transistors, their "gate" (they use different terms, the "base", "emitter", and "collector" of a BJT roughly map to the "gate", "drain", and "source" of a MOSFET) is a *current* sensor rather than a voltage sensor. they're very different devices and not as well suited to this application, but for a simple device like this with relatively low performance needs, you can basically pick whichever type of transistor you find the math easier for.
oh. reaching the end of the video, you did use a BJT? i think you missed that change in your narration, because where you last left off you were using a depletion-mode MOSFET?
also, since you didn't mention datasheets at all, and idk what caused all the MOSFETs you got to not work, i should also mention that the pinout of transistors is not standardized, you have to check the datasheet for the specific model. for most packages I'm aware of, i don't think there's even a de-facto default most transistors will use, so you really can't get away with assuming the pinout. I would expect some aliexpress transistors could be totally non-functional, but the majority of the "non-legit" transistors you get will be counterfeits. counterfeit components usually try to fly under the radar, they don't want them to be obviously fake at first glance which means they need to work and not instantly blow up. the ideal counterfeit component is one that doesn't cause any problems at all, and the next best thing is one that doesn't fail until it's too late to reject the batch. most counterfeit parts are either salvaged parts that have been cleaned up and sold as new (very problematic for electrolytic capacitors which wear out over time, while semiconductors theoretically might still work fine), or a cheaper part (or even a clone) that has the same basic function, package, and pinout, that has had its markings changed to make it look like a more expensive part. that's why i'm suggesting possible circuit issues, it seems weird that all the MOSFETs you got were fully non-functional (unless you bought them all from the same seller)
The issue with the mosfets was Purley that they wre from aliexpress, I have actually since recreated the project with new mosfets and everything is working fine. I just had mosfets that didn't match the data sheet. All the mosfets were same pack, I nought it specifically for this project.
Good luck the man who predicts everything. Keep going lol🥰
Look up those MAX745 lithium charge modules! You can connect one directly to 20V and it steps the voltage down to 1-4S for your battery, has an adjustable charge current. Much easier and efficient that way.
Also, with the heatsink inside and without fan it will definitely overheat the leds very quickly.
the IC supports this, the question is whether the board does. the input capacitors also need to be rated for that voltage, and since it's a buck convertor, a larger voltage drop means more ripple current for a given inductor value, so it affects the calculations for the inductor and output capacitor as well. the chip is only part of the picture
@@spambot7110 It does obviously. Also it basically needs to, since 4S is 16.8V that means you need to supply ≥18.8V for it to work at all.
The supplier that these boards come from has 20v modules, but this project was rather time rushed, and as I live remote it would be hard to get them in time
you predicted everything
Mna HE PRWDICT IT.
i subbed.
6:43 so, a couple things: usually a MOSFET's gate threshold voltage is pretty low, it'll start turning on (just barely, meaning a high drain-source resistance) with just a couple volts, and then the drain-source resistance approaches its minimum at maybe 10 or 12 Vgs. You really need to read the datasheet of a MOSFET to have any idea what a given gate-source voltage will do.
also: ideal MOSFETs don't automatically turn off when you disconnect the gate (see the end of the comment for why real MOSFETs sometimes do anyways, and why this shouldn't be relied on). the problem is, a MOSFET gate is basically a voltage sensor, it's a small capacitor that detects its level of charge and turns the switch on and off. When put your hand on a Van De Graaf generator and your hair stands up, it doesn't immediately flop down when you let go, because you are still charged; you have to touch something grounded, causing a spark of current equalizing your voltage with the voltage of ground. similarly, an ideal MOSFET gate doesn't consume the charge in any way, it just sits there, the mere presence of a charge difference between the gate and the source passively activates the switch. so when you disconnect the drain from the voltage supply, it'll keep chilling there. This is great for efficiency, it means that you don't need to consume any power to keep the MOSFET turned on, you put in a small amount of energy to *turn* it on and then it'll just stay on, modern computers would not extremely inefficient without this advantage. in your application, where you just want a simple circuit to make a switch turn on and off, this doesn't matter and is just an annoying problem you have to solve.
The solution is to connect a resistor between the gate and the source, something in the 1k-10k range probably. That way the gate-source capacitance is constantly being discharged, and requires a constant stream of current to keep it topped up, so if you cut off the power it'll discharge at a predictable rate and quickly drop below the threshold voltage, turning it off. You can check the datasheet for the gate capacitance and gate threshold you can calculate how long it'll take to turn off for a given voltage and resistance, but practically for your use case 10k should turn it off plenty fast enough while only wasting a negligible amount of power (as per ohm's law, at 12V a 10k resistor draw 1.2mA of current, 1.2mA * 12V = 14.4mW). This is often called a pull-down resistor, because it's "pulling" the gate-source voltage towards zero. when you connect the gate to a voltage supply, it's a much stronger connection so it "wins", "pulling" the voltage up to 12. when you disconnect it, the pull-down resistor "wins" by default.
In practice, nothing is a perfect insulator, so the imperfect insulation between the gate and source of a MOSFET can be thought of as a "built-in resistor" between the gate and the source. However, this is an undesirable property, and manufacturers are always trying to maximize this resistance. the only spec for gate-source resistance you'll find in a datasheet, if any, is a *minimum* value, so if you make something that relies on it being at most a particular value, it's either not gonna work, or be up to the silicon lottery, random environmental factors, the electricity gods, etc. essentially, a real-life MOSFET will turn itself off when the gate is disconnected, but you can only calculate the *minimum* amount of time it'll take, in practice you never know if it'll take 100us, or 3 seconds, or 3 hours. and if it does take 3 seconds, it's gonna spend a lot of time in the in-between ("linear") region, where it's switched on but its drain-source resistance is pretty high, causing a large share of the power of the circuit to dissipate in the MOSFET, and it'll blow up.
if you don't like this weirdness about charge and voltage, you can use bipolar transistors, their "gate" (they use different terms, the "base", "emitter", and "collector" of a BJT roughly map to the "gate", "drain", and "source" of a MOSFET) is a *current* sensor rather than a voltage sensor. they're very different devices and not as well suited to this application, but for a simple device like this with relatively low performance needs, you can basically pick whichever type of transistor you find the math easier for.
oh. reaching the end of the video, you did use a BJT? i think you missed that change in your narration, because where you last left off you were using a depletion-mode MOSFET?
also, since you didn't mention datasheets at all, and idk what caused all the MOSFETs you got to not work, i should also mention that the pinout of transistors is not standardized, you have to check the datasheet for the specific model. for most packages I'm aware of, i don't think there's even a de-facto default most transistors will use, so you really can't get away with assuming the pinout.
I would expect some aliexpress transistors could be totally non-functional, but the majority of the "non-legit" transistors you get will be counterfeits. counterfeit components usually try to fly under the radar, they don't want them to be obviously fake at first glance which means they need to work and not instantly blow up. the ideal counterfeit component is one that doesn't cause any problems at all, and the next best thing is one that doesn't fail until it's too late to reject the batch. most counterfeit parts are either salvaged parts that have been cleaned up and sold as new (very problematic for electrolytic capacitors which wear out over time, while semiconductors theoretically might still work fine), or a cheaper part (or even a clone) that has the same basic function, package, and pinout, that has had its markings changed to make it look like a more expensive part. that's why i'm suggesting possible circuit issues, it seems weird that all the MOSFETs you got were fully non-functional (unless you bought them all from the same seller)
I used a depletion mode mosfet
The issue with the mosfets was Purley that they wre from aliexpress, I have actually since recreated the project with new mosfets and everything is working fine. I just had mosfets that didn't match the data sheet. All the mosfets were same pack, I nought it specifically for this project.
@@Boris_Woodwards ok, that's good. i'd still recommend adding a resistor to discharge the gate for long-term reliability though
What did u actually predict
Bro predicted everything🙏🙏🙏🙏🙏🙏
I am he
@@Boris_Woodwards nah, your him
The predictions are unreal 🙌🙌🙌🙌🙌🙌🙌🙌
lol you should get a aliexpress sponsership
Would be pretty nice
What is going on who r u and what are you predicting on this video 😅
@MAHDERTsegay-p4i I'm tending on tiktok, because of a meme, where people say I predicted everything.