Sorry to hear that.

@@greatscottlab Why sorry , I really enjoyed you busting their videos. I had a feeling they were rubbish but who am I to argue with like you said half a million views.

The best of those crappy circuits I've ever seen, dunno if it was UKposts or just some random Internet page, was a power supply. 5V output, some ~5A at most, but you'll need extreme cooling on it, like, some people were adding water cooling to more powerful versions. Why? Oh well it was 7805 based. Rectify line voltage -> smoothing capacitor -> 7805s (or what is the version that can take line voltage, dunno). That's it. For real, a linear dropping of line voltage down to the supply you want. Not even talking about it not being insulated, so it was horribly inefficient and unsafe at that.

We need a blacklist of videos and channels providing crap and misleading content. Once loaded into appropriate browser extension (that blocks/filters out specific UKposts videos/channels from appearing), we would all be better off. I already have one started that blocks 40 or 50 of those 5-minute-bovine-excrement channels. For DIYers, this is extra important, to remove time wasters. One pet peeve of mine are videos with English title and thumbnail, but in an unrelated language and without any subtitles (manual nor generated), which should be a category of its own (given a red border or something?). A user-contributable github should be created.

@@cbs1710 pretty sure UKposts sometimes auto-translates video titles. I've seen videos I know have Japanese titles get translated to English on occasion. No idea why it only appears to happen occasionally, but I suspect that's what's happening with those English titles but foreign everything else ones.

I will see what I can do ;-)

I would even be interested in exposing the extremely dangerous electronics projects some of these channels post. A lot seem to be aimed towards kids, but almost never mention important and deadly aspects. Your knowledge on electronics would be an awesome perspective as I’m not nearly as knowledgeable lol

I do think the video is a good starting point for those who wish to make their own stuff.

I would not call it exactly a bust. It's more too simple to compensate for errors and stalls

@@greatscottlab Can you maybe do a 3phase motor and how to drive those?

Terrible when something like this is your first experience with electronics.

If you watched the video careful, you would have seen this device DID work.

@@janklas7079 If something "works" doesn't mean it'll keep working in the long run. Those MOSFETs will probably only last an hour before they go poof.

@@greatscottlab I would say this is a pefect first experience. It's simple and it works, but has a lot of room for improvements. A lot of stuff to experiment with.

@@janklas7079 sounds like a great way to burn out a motor.

Maybe hes using the Return Dislike Button browser extension

they won't as then you'll truly see who is pumping out "mis/dis-information"...which is why they hid it

yeah he must be a fellow return dislike appreciator

also YT removed dislike counter to avoid "bulling the creators" as they said, but it's hided for users and totally showed to the creators...

The only thing hiding it from users does is stop the mindset where people are now prone to bandwagon stuff, lots of dislikes makes them more prone to view things negatively, and more comfortable disliking it themselves. The problem is, youtubes like system doesn't lend itself to this behavior much naturally at all, and even then the large majority of people won't experience that pressure the vast majority of the time. It's not at all necessary and the only thing I can think of as a benefit is absurdly niche, you'd need content that has a large audience disliking it, and another audience portion in the middle, but also people to acting on a group mindset, and also it actually being detrimental enough to be relevant, which just isn't really common. Wanting to remove negativity from subjective stuff could be understandable, not everyone will like certain songs for example, and some may actively dislike it. And that can feel discouraging. But even then, it's almost never enough to be a relevant problem, only applies to opinion works, and typically is in the domain of discussion, not like and dislikes. And of course you'd only be influencing people who are going yo actively feel upset that 2% of people dislike something they love, and you probably get the point by now haha. It's a very strange decision to me. I can think of a few potential benefits like this, but they are so inconsequential and likely rare anyways, that it doesn't really effectively do anything. And on the flip side, you lose that credibility anchor, you lose people being able to dislike misinformation or misleading content. And you lose that feedback for a skeptical mind, you think hm maybe this is good? I mean it has 1k likes. But you don't see the 600 dislikes, so you miss that feedback, that a large chunk felt mislead or disappointed, and instead of just mindlessly clicking the button yourself you are much more likely to use that information to look into and think about things more yourself, had you still had access to it. Idk man. It's a very strange situation to me, the fact that the given justifications are worse than the junk I can whip up here is even more strange. Whatever. At least it's not something more horrible, like having the dislike ratio hide videos from search or something lol.

Yeah and thats... sorta what the circuit in the video does, poorly, a bldc will work with just 3 phases like that but a proper feedback system and all that from a proper controller makes them a lot more useful

It might be a modified sine wave at best or just a rounded off square wave. Real sinusoidal AC takes a lot of components and usually a transformer and filters. Something like an audio amp. Then something like capacitors or a microcontroller to have a 120 degree phase shift for each phase.

Well said!

It's annoying how UKposts loves recommending fake, clickbait and viral crap more than proper stuff. You gotta really go out of your way to make the proper stuff show up instead.

I wish I could get UKposts to display only the things I’m interested in and block the rest - I’m so sick and tired of being fed clickbait , reaction videos and fantastic new idea or technology videos 🤢

Thanks. Not sure if I will ever get back to the VESC video. Sorry.

Some ridiculing is deserved though especially when the hacks are extremely dangerous, which this one isn't but I have seen to many playing around whit main voltage stuff

@@link7417 Perhaps in some cases, but not when almost the entire airtime of a "debunk" video is just ridicule, especially of an idea that seems pretty good at first when written down on paper but starts to fall apart the moment you squint and start to look into it. At that point, the debunk deserves to be less a ridicule fest and more a "Okay, this is what works in this idea, and this is why this idea as a whole isn't going to work."

@@calyodelphi124 oh I definitely agree whit that,

What the fok

How can you comment 9-days ago 😳😳😳, the video is uploaded just now😂😂

@@Xavier_Everwhere patreon

There is. Mostly error conditions and timing. To get timing going, and stuff like braking, one needs a CPU. This increases the number of components a lot. Also, most ESC's also have a BEC function, providing power to receiver/servo's, and so forth.

@@janklas7079 I don't think a CPU is a hard requirement. You could do what the CPU does with comparators etc. But if you did it would be hard to tune and probably only work with one type of motor. Using a microcontroller reduces the amount of components because you move most of the problem from the hardware domain to the software domain.

True

You are welcome :-)

Haha true :-)

@MuhammedG Exactly most of those prototype BLDC ccts looks like an amp schematic lmao More Test & Tune is needed to make a pro-compact ESC!

So it isn't good as a guitar amp cause not enough distortion, neither as a music amp cause too much. 10/10.

The main difference is that the commercial sort have a microcontroller scanning across the coils, activating one at a time. This simple DIY sort is stable just activating one coil, and only switches to another one if the rotor moves. I guess it relies on induced current from the magnets, to switch states. It would probably be possible to make a better one, with more components, by adding capacitors to turn it into an oscillator with three states - similar to a two transistor, two capacitor oscillator, but with three stages instead of two, with each one triggering the next. You wouldn't be able to adjust the speed easily though.

@@TooSlowTube in this vid it is a joule thief or a skin effect consor

With a few more bipolar transistors you could have actual feedback gating and zero crossing detection, as well as anti-stall kicking. Sure, it’s easiest to do with an MCU if you don’t know a lot of analog building blocks, but those MCUs for the most part implement simple algorithms that come directly from older industrial servo drives that were 100% analog.

@@TooSlowTube I'd love to see that video. Even better if one started with this one, an showed how to solve problems as they turn up. That would be a Great Scott -video I'd watch over and over again.

Hmmmmm well. Fine. I ok with that.

Without going into the math? Is a tad sad but they are way more complicated than one thinks. Salliant poles ect... The math in the control schemes is mad too

@@matthewmaxwell-burton4549 for start your journey you dont want math, you just need see that thing work to start beliving its for you

@@JTKK9 fair point, I guess everyone is different.

@@greatscottlab Starting 25 years ago, for purely teaching purposes, I have been showing the principle of driving a BLDC motor using 3 H-bridge based DC motor drivers. In practice it seems that people who want to drive 3 wire motors already have experience with H-bridges. Many H-bridges don't provide an easy way to control each half independently so I use half of each of three of them. I first did this with the LEGO Mindstorms RCX (the model introduced in 1996). It made a cool demo and showed useful principles of operation. The maximum speed was quite slow because the waveform steps were generated with the commands that started, stopped, and set the direction of each motor in the language interpreted by the RCX. This removed a lot of the confusion about BLDC driver theory and builds on what they already know. After seeing what amounts to bit banging a motor controller, they see the benefits of a more traditional controller with the 85 parts.

Fundamentally, "Brushless DC" motors are just 3 phase motors. The controller is really a 3 phase inverter with smarts to change the speed and handle the different bad things that can happen. Your example would be the same as a phase being out. It "works", but you really need that 3rd phase.

then add 3rd audio channel with correctly shifted signal and itll work. ;) thinking about that sinusial signal and timing logically is some thing ;)

Just then, I only had a stereo amplifier and a generator from the phone application, so I could only get two signals. I was just curious if something like that is possible at all and at what maximum frequency the motors I had will not be able to spin anymore. Everything was checked without load.

What's the point, you have no control over the torque which esc's give. They allow you to decouple speed and torque which is a godsend.

Considering it doesn't even have a full 3 phase h-bridge, it is rather impressive.

@@matthewmaxwell-burton4549 So you're telling me there isn't another way to control torque without microcontroller esc? I don't believe it. This thing certainly could be improved with right knowledge. Only thing is it won't be a "universal" type esc, rather a "motor specific" once. Still such controllers have some niche applications. And not all motor applications need torque. I.e - The spinning reflector of a laser printer

@@Jtretta Yep, my exact thoughts. Spinning a BLDC is bloody hard, it's impressive such simple design can do that.

First comment which actually makes some sense.

Well. In that regard I do agree :-)

theoretically you could hold it under a faucet to get it to spin

@@Blox117 How does that matter tho? The question is not "how can I make it spin in general?" the question is "how can I make the motor spin with electricity?". I don't know how your comment factors in that discussion. Its not like the circuit is not doing what a normal bldc controller wouldn't do, its just that it is quite simple and does it poorly.

@@Ramog1000 you could get it to spin by connecting the phases at different times manually. ta da

Would u mind sharing it?

There were some sparks in the video ;-)

Electroboom doing those for entertainment purposes and intentionally. And he doing well of this.

Take 3 of the DIY inverters and feed it by a signal that's delayed by 120 degrees each. Of course, at that point, it's just easier to use a $0.50 cent microcontroller and PWM the signals with a smoothing capacitor. It still won't detect stall conditions, but it will at least drive the thing reliably.

@@arthurmoore9488 absolutely

Thing is "simple" is often simple* when you start cutting so much out that the circuit is reliant on certain stray values or other unspecified/variable component characteristics to work properly, which makes it difficult to duplicate. There are some popular simple circuits that are simple and easy to get going with close enough parts, joule thief comes to mind, but there aren't many of them and they don't do anything particularly complicated, e.g. properly drive a BLDC.

Aka muntzing

True :-)

This goes for damn near all electronics.

Unless they are free and abundant if you know where to look.

The current to the gate of the MOSFET is nearly immaterial. It is a voltage-controlled device. A huge current flows for an instant when switching but then it quiesces (falls essentially to 0). Many will put a resistor in the connection to the gate to keep current down, to keep that huge switching pulse down. Really any resistor will work. 1K, whatever. You're going to have a lot of trouble running a P-FET to switch that resistor. Specifically to turn the FET off. This is because for the PFET to work you have to put it in the circuit between the resistor and the positive voltage rail with the source connected to the positive rail, the drain to the resistor and the gate toward the Arduino. You cannot put it between the resistor and ground because then the source voltage will drop to 0 when the FET is off, with no constant source voltage you can't manipulate the gate to source (critical turn on/off) voltage properly. you cannot turn the FET around so that the source is at Gnd because then the body diode will conduct all the time and it'll never turn off. So the PFET will be above the resistor and to turn the FET (resistor) on you simply need to output any voltage below about 23V to the gate. That's easy. The problem is that to turn it off you need to get the gate voltage up to about 24V (above 23.4V). And you can't do that from your Arduino. The highest voltage it can output is 5V (IIRC). The reason for this is the turn on voltage for a PFET is a negative number referenced to the source. And the source in your case is at 24V since it i. If the turn on voltage is (say) -1.1V you add that to the source voltage of 24V and then that means putting 0-22.9V on the gate turns it on and 23.0V and up turns it off. Your easiest out for this is just to turn the circuit over and do low side switching with an N-MOSFET. Like in this figure. en.wikipedia.org/wiki/MOSFET#/media/File:Mosfet_n-ch_circuit.svg Where you just leave the LED out. And note the source is to ground, the drain to resistor. Other end of resistor to 24V. Then the wire that goes to the switch there from the gate you connect that to your Arduino GPIO. Now to turn off you have to get 0V onto the gate. That's easy, either drive 0V out or just don't drive out and let the resistor pull the gate to Gnd. To turn on you have to get over about 1.1V to the gate. That's easy, just drive out high (5V) and it'll turn on. Now you can turn your resistor on and off easily from your Arduino. There are some difficulties, like currents. You can work that out. The other is that whether the resistor is on or off it still has at least one end at 24V. So you have to design your enclosure so that that resistor doesn't come into contact with ground OR PEOPLE. If you can't do that, then you have to go back to the PFET design again and instead design a gate driver. I personally do this by using an NFET to drive the PFET gate. You hook the PFET up above the resistor as mentioned above, put a weak (10K, 100K) resistor from the PFET gate to 24V so that the PFET is off. Now to turn on the PFET you have to pull the gate down. You do that using an NFET. You hook the drain of the NFET to the gate of the PFET. The source of the NFET to ground. And the gate of the NFET to the Arduino. Also hook a resistor to gate of the NFET and to ground. Now, as the NFET explanation above driving the Arduino GPIO out high (5V) will turn on the NFET. But when the NFET turns on now it pulls the PFET gate down and turns that on. Driving out low or not driving out at all will turn it all off, although a bit slower than it turned on (still in tens of milliseconds at most). Note these NFET circuits (and thus the NFET driver to PFET) will also work fine if you have a 3.3V Arduino (do such things exist?). As long as your gate threshold on the NFET is below 3V. And there are many which are.

i like ur funny words magic man so are u trying to turn on a resistive wire or are u trying to make a complementary transistor circuit or are u trying to make a buck converter? if bjt doesnt work, u could try cmos but u dont really need driver for switching a heating wire if u want to switch something on and off, u don't even need a complementary configuration, normally u only need 1 mosfet but the tricky thing here is actually the 24v source, most mosfets probably cant handle vgs more than 20v difference even trickier the that the mosfet is p type if u wanted a complementary configuration at the gate of the p-mosfet, u would have needed 2 different voltage source that supply high current and more than 2 extra transistors, and the arduino 0v and 5v would have to be mapped to 12v and 24v, this is much more complicated and an opamp might be necessary a buck converter is quite simple to make, u can feedback using opamp to an oscillation circuit, or u can directly feedback to microcontroller to control pwm another way to control voltage for the wire is using linear regulator, lm317 or build ur own with an opamp like the buck converter building a linear regulator with opamp is easier than building buck converter, just need some resistors and a transistor but a lot of power loss in the transistor, and linear power transistors are more expensive than switching power transistors and a buck converter doesn't really "control the voltage", if u truly wanted to control the voltage, u would need the p-mosfet to operate in it's linear region that's a lot of power loss, most mosfets today are switching mosfets, u can get one expensive beefy switching mosfet and strap a heatsink to it to operate in linear region, but it is still 50% chance that it is gonna burn up in flames and explode because they are just not designed to do that, u have to consider the power lost in the transistor using P=VI most switching transistors say they could handle crazy currents like 49A 100A, but that's for switching at high frequency, very momentary current, they gonna explode if u tried to conduct 49A in saturation but why are u even trying to control the voltage? fun fact, u dont need to u are just trying to heat up the wire and why do u care about the current in the gate? for dc application, this is like non existent only for very high frequency do u actually need to care about the gate capacitance and crap u are overcomplicating a very simple problem and there is a very simple solution, it is called PWM as for the p-mosfet part, everything is reversed from n-mos but for both, load are typically driven at the drain side and source pin are connected directly to ground for an n-mos this make the mosfet have very low and negligible resistance and voltage drop(see rds(on)), even lower than bjts, bjts normally have 0. something of vce(sat) and vce(sat)>1v for power bjts, but mosfets just has rds(on) that means when turned on, mosfets are effectively short circuit while bjts are like diodes(in terms of voltage drop) and p-mosfet has reversed everything, that means current directions and voltages, it can do the same thing as n-mos, just reversed for p-mosfet, u would want ur source voltage instead of ground on ur source pin and since positive voltage on gate turns on an n-mos, that means p-mos will turn on if u apply a more negative voltage than the source pin but dont directly connect the gate to ground, mosfets have max rated vgs, and for those irf stuff i have seen is + -20 for n and p mosfet respectively, connecting directly to ground means vgs=-24 and that means it blows up and this is very inconvenient because arduino only outputs 5v or 0v so we will need to use another component, either another transistor or opamp, since using opamp is overcomplicating a simple problem, we will use transistor we will use an npn bjt, connect emitter to ground, connect arduino output in series with a resistor to base now make a voltage divider where one side is connected to 24v, middle is connected to gate of p-mosfet and another side is connected to the collector of npn bjt use equal value for both resistors so when the npn bjt is turned on, the gate voltage becomes about 12v so now when output is 0v, bjt turns off, gate is pulled up to 24v, vgs=0v, p-mosfet is turned off and when output is 5v, current goes through bjt and turns it on(base resistor is only there to limit current to prevent arduino and bjt from being damaged, so make sure base resistance is low enough that it can still make bjt go into saturation, turning it completely on), and voltage gets divided at the gate, vgs=-12v, turning it on and perfectly safe for the last part, we connect the heating wire to the drain of p-mosfet, then the other side of the heating wire to the ground voila, we made it when arduino outputs 5v, p-mosfet turns on and wire heats up, when arduino outputs 0v, p-mosfet turns off and wire cool down now this is where PWM comes in, imagine if you switched ur output on for some time then switched it off for some time let's say 60% of the time u switched on and 40% time it is off that means 60% of the time u r applying 24v which is same as applying 14.4v constantly if 50% of the time u switched on, that is equivalent to applying 12v constantly this is almost like a buck converter but with less extra steps and analogWrite in arduino does exactly just that, it isnt actually outputting a real analog voltage, it is outputting a 5V PWM, at around 500 or 1000 something Hz that frequency is fine so u dont have to manually to change the timer register to adjust PWM resolution and frequency at this frequency, u dont really have to care about gate current or capacitance remember in ac circuit, Xc is lower when frequency is higher, but Xc is higher when frequency is low, that means current will be lower at low frequency, negligibly low just assume everything is 0 for heating a wire, u can even use tens of Hz, mybe even 10Hz, 100Hz is also fine any frequency will produce the same result and same heat, frequency isnt really critical for heating a wire in conclusion, this is a very simple problem, u just needed 3 more resistors and an npn bjt, base resistor could be something like 2.2k or 680 which will almost certainly ensure saturation, both resistors for voltage divider can be 10k or 2.2k no calculations at all, dont need to overcomplicate stuff

​@u r noob I love your answer. I know that I might be overcomplicating heating a resistive wire. My initial thoughts were that I need a mosfet, that will open power to the wire and then I continued with I need to limit that voltage so I can control the amount of the heat that will be produced. So I jumped to the buck converter and immediately thought about fast switching the mosfet. My first tries were kinda too hot, because I was almost burning the MOSFET with slowly charging the gate and wrong voltages at the gate. Then I learned about push-pull (totem pole) driver and tried that. I also posted a question about this with a schematic on a stackexchange but instead of some constructive critique or some thoughts I got almost shamed without any explanation. I was trying to learn everything I can from all kinds of videos so when I learned about push-pull I was eager to use it. Also I want to learn everything about mosfets and how to properly use them. But I believe that I was overcomplicating that. Thank you for your answer. That helped me understand a lot. I'll follow it.

@@urnoob5528 I have one question tho, wouldn't the MOSFET burn? I tried your solution in a simulator and there was a 20V drop over the MOSFET.

And with said limitations its possible turn the motor in the first place. In my opinion this is valuable information, but its niche market for places that can utilize this stone age esc. Years back i would loved to know about this, for random test contraptions that just need to be thrown together and run in constant speeds. Hell it could simple be used to test brushless motors if those even work. It really does beat expensive and fancy ESC's specially when time and cost are priority over features like "cold start". The real issue with these videos are that they dont explain the pros and cons of the setup at all with leads to mentioned scammy feeling coming from it.

Thank you! Cheers!

It's funny you mention code - I'm not really a programmer (though I do code), but I am a fairly senior systems engineer. An instant display of naivety is, to me, when I see anybody say "Pfft, I could build that in [x]". You see it so so often in engineering, electronics, IT. Nobody ever seems to stop and think "Maybe it took clever people so long because it's more complex than it looks to get right"

​@@hermand Over confidence happens even to the best scientists. Not too long ago, and even now, many scientist still use the word "junk DNA" for non-coding DNA thinking they are garbage. That is like me looking at a computer mother board, and say all these resistors are junk components.

Exactly, every part in any mass manufactured product is there for a reason. Whether or not you know and understand that reason, they didn't waste money putting it there for lulz.

Could it perhaps be due to their lack of brushes?

@@SpeccyMan They don't run on DC, though.

Could it also perhaps be because they are run from a DC power source instead of a synchronous 3-phase AC power source?

@@kimsmoke17 But they're not. They're a motor that requires a controller. How the controller derives its source power doesn't classify the motor it's driving.

Use an arduino starter kit.

There are these experimental / educational kits which have a very good handbook, explain every step all the parts and the physics, start easy and then get more and more complex. Like the ones of KOSMOS (for example Electronic XN 1000). But there are many more like these.

I recommend watching videos explaining the individual components, like the electronics-101 videos from electroboom. This way, you might understand individual components enough to build circuits.

"The art of electronics" is the book you are looking for

I know that feel. GreatScott! does some really cool shit. His videos inspire me while simultaneously making me feel like I suck at eLeCtRoNiCs. Remember you're only 19 and GreatScott is an electrical engineer. There are a hundred youtube channels, where some barefoot Indian kid with bugs and chickens walking around everywhere, make some really cool simple projects. You can learn a lot from those videos. Try watching all different types of electronics videos, keep experimenting, keep learning, and keep having fun with it.

fax

There is a chrome extension that allows us to see it

There are Chrome extensions for that ;-)

You can see it if you have vanced on android

Patreon supporters can watch way earlier ;-)

@@greatscottlab awesome! ;) ;) I might just have to comment 9 days earlier...

When your German is as good as his English........😄

That wont happen he is german Its not that bads

People talk how they talk. Get over it.

Is it any worse than your ending a sentence with a preposition?

This is how I talk. I try to improve every time. But this is the way it is. End of story.

My pleasure

Hope you enjoyed it!

Yup bro try to increase the upload speed and decrease the time taken to create an video ..... don't try to make it too perfect.

Waiting

Can you please elaborate ? What's the input/output signals, how does it work to be a phase controlled system in a control loop ? I have basic understanding of PLL but i can't make sense of your comment

@@weistonaski6924 it is a fixed frequency phase shift oscillator in which the drain pin on one MOSFET phase feeds though a capacitor to the gate on the adjacent fet. Repeat 3x and you have three fets operating in parallel. Drain on each FET is tied to a resistor to source voltage. Like an astable multivibrator which is basically a 2 phase shift oscillator the three FET circuit is a 3 phase shift oscillator. Frequency cannot be adjusted without varying each drain-to-gate capacitance equally otherwise you will have inequal phase shifts due to inequal capacitance time constants.

Any time!