As always, the build-vs-buy question is usually less about cost and more about specificity - you can tailor the build to a very specific project, whereas finding a pre-built version to meet some projects might be impossible. Plus the fun and learning of doing the build yourself, of course.
There's also the labor aspect--how much time did you spend prototyping, then building and testing the finished circuit(s)? I think this is the second half of the specificity trade-off. Your off-the-shelf components are ready to go, but might not be an exact match for your use case. A custom build can be made to your exact specification, but you have to spend the time to develop it to your specification.
It's not really difficult, i also explained how to make these leds in my channel, but using SMD components, just like the ones they sell (in my kit, the emitter was working at 180khz for a change) and even smaller (the comercial ones uses CD43 inductors, and i managed to use CD32 inductors, a bit lower range and dimmer light, but still working). I'm glad more people is spreading this knowledge arround for DIYers. Great video!
Take an ICM7555 with R1 = 1k8 and R2 = 4k7, parallel to R2 a diode from pin 7 towards pin 2 and 6. And a capacitor 1nF to GND. That should be about 222 kHz. Mayby a high Ohm poti.. Already has 5 volts level for the MOSFET. If it is not enough, two logic inverters in between, which usually have powerful outputs. First one inverter and the 5 inverter in parallel for the 4000 or 74xx series. The components are cheap and can be soldered onto a small circuit board. Then there are hardly any hidden capacities and loose contacts.
Definitely worth a go if you have a suitable logic level MOSFET that you can drive with 5v logic. And you can drive enough current to switch it quickly.
An ICM7555 is CMOS and will only drive a few milliamps, so isn't suitable for driving power mosfets directly. The CD4049UB isn't much better, although with five in parallel, you might drive about 50mA cleanly. The standard NE555 or similar should drive 100mA with a 5V supply, but you will still have problems with reduced voltage swing, probably to around 3V. A typical switching mosfet like the IRFZ44N will turn on with a gate voltage above 4V, but has a gate charge of up to 67nC, which will take 0.67μs to charge/discharge at a constant 100mA. The switching frequency used in the project is around 200kHz, which means an entire cycle takes just 5μs and each half-cycle is around 2.5μs. Having a switching mosfet in its linear region for a quarter of the time will cause problems. You just don't have enough drive without using a mosfet driver if you've only got a 5V supply to use.
This is nice. I was looking all over for how to do this like 5 years ago and came up pretty empty. I don't have the electrical skills to figure it out myself. I do love the very compact design of the original leds.
Amazing project!!!! Mainly because how you explore how evaluate the components needed in order to make work properly!!! Thank you for sharing!! Perfect for a science fare!!!!!
It would be really cool to have these wireless LEDs inside some dice. Then make a dice tower with the coil in the base. So when you throw the dice they light up! Awesome, excellent video! Love the thorough explanation
oh nice! Thanks! This is exactly what I was looking for - the plan is to 3d print a set of chess pieces (using clear for white and tinted resin for black) and put LEDs inside, powered from a coil in the board. This will certainly help to build this!
Perfect usage. They have been using these for years in hobby models. For example gun damn models. When you set the model on the stand. They glow. The eyes and shield and.... glows
Good stuff, but for those of us who don't have electronic parts lying around already, or who would have no use for the excess that comes with buying in bulk, it's actually cheaper to buy from the Chinese. Still, it's good to have the option!
This is fascinating, I have a very specific use-case where I will be buying the pre-made units because I need them to be very small, but I also want them to have light emitting all around, so I want to see how many LEDs each part can drive by just soldering extra SMD LEDs to the same leads as the existing one. I'll also have to check if I can do the trick of reversing some of them to use the other half of the waveform. Ideally I'd like 6 LEDs running off of a single inductor in a very compact little package. I"m really hoping there's plenty of overhead in the design, since LEDs really are very low-power devices.
Just some rough calculations on "without a gate driver" (sorry this one intrigued me). Looking at that trace, total time is ~660ns with fully on making up about a third of it. So at 222kHz you'd be spending about 5% of the time turning the thing on and off. Possibly not the end of the world but probably not superb either. (Apologies if I totally stuffed up on theory there - was just a guess)
iv just purchased 6 sets of these for my 1/14 scale truck trailers. all wireless. love em. would have known about them if it was not for you. thank you
Would love to see something like this become a wireless Christmas tree lights. Perhaps hot glue on to a tree branch, hang the coil like an ornament to power them and no more wires for the cats to chew. How many lights will one coil power? It looks like each set has about a 9 inch radius, so it seems like it would require a few coils to decorate a tree. Not to mention linking them up. Hmmm. Maybe it's not wireless after all.
Super surprised how much current it generates without the pull down resistor on the mosfet gate. This little design really makes me think how they work, cheers!
It's really interesting, if you leave the gate floating then it can easily get switched on - and if you're moving a lot of current around it's easy to damage something.
it's more that inductors resist a change in current, so if you keep things on, after a brief moment, you sort of have a short (but wire resistance etc). If you keep things going on and off quickly the inductor never quite gets up to fully conducting (it's busy resisting the constant changes in current). The resistor on the mosfet is more about just being certain it's at zero volts when off. Its a fun experiment to try (without driving big current) to try one without that resistor and see what you get. I've mistakenly forgotten them before and wondered why stuff doesn't turn off.
Actually quick demo why you want that resistor ruclips.net/video/8swJ_Bnsgl4/видео.html you can see even prodding the gate with your finger can be enough to get it (at least partly) on
I've been quite fascinated since I first saw your excellent video a couple of months ago. I have my version working with a 555 timer and a trimmer pot to vary the frequency. I couldn't manage to get the duty cycle close to 50%, so I used an 2N7000 mosfet to invert the signal, which also cleaned it up a bit, and then I fed this into a TC4420 mosfet driver. Maybe I'm pushing things a bit, but I've been using 12V across the power mosfet. My bench power supply tells me that I'm consuming about 2W of power, which seems reasonable. Everything seems to be working nicely, though I'm still tweaking things. One thing though: So far, I've been using a CBB film capacitor on the transmitter coil, but today I noticed that it was getting pretty hot. I suppose I could try turning the voltage down, but I'd rather not, since 12V improves the brightness. I see that you are just using a small ceramic cap, so I was wondering if you might have any thoughts on this? I've also been using SMD parts on the LED side. CD54/CD43 inductors, and 0805 LEDs and caps. I'm trying to find a way to solder the parts more efficiently. I'm thinking of making some tiny little DIY PCBs with 0.4mm copper clad, and using something like a paper punch to cut them into circles ... unless someone has a better suggestion.
That's great! You've certainly taken it further than I have. I would have a look at the ESR of the capacitor you are using. Maybe you can find one with a lower value that the one you have. I like the idea of the tiny PCBs.
The cost could be improved significantly, using bjt push pull for the gate driving, making an oscilator using an opamp (lm358 is dirt cheap) and you could probably get away with using cheapo mosfet like AO34xx which are cheap even in my country and can even do 5A and turn on even with 2.5V
Would extending the cables from the inductors, in order to have the LEDs at a greater distance from them, affect their operation? I was wondering if I could use such an array in order to "wirelessly" power some LEDs located inside scale models. The concept is as follows: 1)Make a diorama-base with the large coil. 2)Put small coils in appropriate places inside the model (located as low as possible) 3) Run wires from the coils to the LEDs you need to power. 4)Place the finished model on top of the base, making proper alignment with the main coil and the internal ones (located in a certain area inside the model).
That’s nice , using smd components, the leds can easly be made smaller. I see some comments of people wanting to implement this in a chess board… hmmm in that case tuned receivers could be a way to selectively lit up some leds while others don’t
Would these LEDs work through a wooden door of approx. 4 cm thickness? It would be interesting to have the LEDs wirelessly controlled from inside of e.g. a garage to lit the key hole.
I think so - they worked through some quite thick books. You would need to try it though. You can push more power through the driver coil - power it from 12 or 24 volts.
@@mrtechie6810 from outside using bluetooth? In lord of the rings hobbits, they couldn't find the keyhole in broad day light!! 🗝 🚪 If the keyhole itself shined then they could've save some time and suffering.
@@GhostsOfSparta use a motion-detector floodlight. Or a Bluetooth switch that turns the light on when your phone gets in range. You could also put a tiny light behind the keyhole, on the inside. If you use an LED, the power is so low that you could have it on all the time.
nice. I purchased the wireless led kit from Adafruit, and being inspired and added some components to a mouser order I was placing: 2.2mH inductors ($0.43 each), 240 pF capacitor, and red LEDs. I thought using 2 LEDs was a brilliant move, but it turns out not so brilliant. That is, 1 LED is much brilliant (~2X the brightness of an LED when using 2).
You shouldMaybe you could give it a go with the bigger coil version too since those have much better range. Great work. Will replicate for my projects.
OK, you reproduced it, now let's see some improvements! For example, can you get more range with a larger coil size, or more power? Does tweaking the frequency improve the range?
Are you sure the shrink wrap on the inductors shield anything? This is not the case on the shrink wrapped inductors we use. Works actually slightly better with the wraps since the permeability is 1.00002. The resonance frequency changes quite a bit with the load you put on. Can you test out what it is by varying the oscillator for max power?
Now I want to make this-ish but overkill...add a supercapacitor (there are some good cheap tiny options), and as someone else suggested, cast into resin...possibly with coils at right angles - the ali video showed the power drop off significantly when the pixels were rotated against the magnetic field) - add a bridge rectifier, and the cap, and you've got yourself some fun glowy marbles that ...do things.
That's an interesting idea - some kind of puzzle where you need to find the right place to put something so it lights up. Or when the lights are arranged so they light up they point to some clue.
What if I just use a 9 volt battery and a coil, skipping all the chips and capacitors and resistors and gates? Just current directly to the coil. Would that over power the LEDs?
I imagine a small metal dish (assuming good conductivity) would make the LEDs not light up or reduce the range at least. The time-varying magnetic field from the base coil would induce currents in the metal dish that aim to reduce the magnetic field from the coil. So it wouldn't interfere by muddling up the frequency, but by reducing the power transferred from the big coil to the LED coils.
Awesome work! I might have to try this project out. You can get some mcu's for pennies, or try using an oscillator chip or circuit. You can also get those SMD LED's and caps pretty cheap if you buy in bulk, so that will greatly shrink the size, but they're harder to solder. Now I just need to find a use for them.... I'm wondering if there's enough power for a wireless micro bugbot thing.
I think a low lower mcu could definitely be powered. You could also use the energy to charge up a capacitor for a more lower hungry device and then wake it for a brief time to do some work.
Would it be possible to use a simple 555 timer as oscillator? So you would not need a MCU and also no gate driver because you can run the 555 at 5v without problems.
Yes, no problem, and you dont need a FET driver, and if you keep the current under 200mA avg (check your 555 datasheet, not all 555 are built the same), you dont even need a FET.
@@alexstone691 In a standard 555 astable circuit the capacitor is charged through R1 and R2 and discharged through R2 alone. Putting a fast diode in parallel with R2 to short the charge path through R2 allows the duty cycle to approximate R1/R2 and that can be adjusted to give any desired duty cycle. Note that with a 5V supply, the diode drop has a significant effect on the rough calculations, so some experimentation or trimming of values may be necessary. Using a Schottky diode would bring the observed times nearer to the ideal ones. There's a exposition at en.wikipedia.org/wiki/555_timer_IC#Shorter_duty_cycle
@@Ozzy3333333 A standard 555 won't source or sink more than about 100mA from a 5V supply, and you'll find the high output voltage is going to be reduced to around 3V, which is very unlikely to turn on/off a power mosfet properly, even if it is "logic-level".
The SMD LEDs are not hard to create, just use SMD coils, they are easy to get hold of and definitely can work without the capacitors although they don't perform as well they work well enough, and have to be REALLY out of tune to not work actually, so it's hard to make them not shine at all. So DIY LEDs + an induction heating driver can work as a starting point for your wireless powered LEDs.
Thank you very much.This is absoluteley stunning. I was wondering how changing the coil Turns and diameter would affect this built and which physical laws could help me figure it out. I quite suck at understanding induction, but would like to experiment with this for different output voltages.
In addition to adding a LED in reverse to the main one, try adding 3, 4, 5 or more led's in series (instead of one), the inductive kick voltage can power many led's in series, and no need for the cap across the led imo (actually its not helping at all at that freq), and the same with the drive coil cap, replace it with a string of led's + resistor, its free energy in the since your cap is just snubbing the drive coil kickback voltage, why not use it to power some leds. Try it and report back.
Hmm, tricky, aside from the power from the device and the manner in which it's transmitted, to achieve the resonant frequency of 2.4GHz, even with the smallest value capacitor of 0.1pF, you'd need to find an inductor value of 0.04uH which would be challenging...
@@mrtechie6810 maybe, but realistically even a perfect 2.4GHz patch antenna may only achieve at best a -30dBm sensitivity which is 0.001mW of receive power crudely speaking - even the small LEDs shown need more than that to light up :/
Let’s say I wanted to make some liquid core resin dice, and I wanted to cast a a wireless led in a bit of shield of resin (maybe just 5-6mm in diameter), put it in the liquid core vessel with a weight on the bottom so that it always sinks to the bottom of the liquid core and shines upward, illuminating the top 70% of the dice. Now let’s say I wanted the LED to slowly flicker so that it looks like candlelight is radiating from the dice when rolled in a dice tray with the coil in the bottom of it. Would something like that be possible? I’ve seen people on Etsy put flickering leds in miniature terrains, like wall torches, fireplaces and campfires, burning debris ect, but it’s wired to something that I assume has the flickering element “programmed” into it and then connected to a battery obviously. What would you have to do with the coil and its power supply (if anything) to cause that flame flicker effect I’m interested in? Thanks in advance.
Yes, but this tech isn't really that bright. I got it to work, but you would need to super tune the transmitting coil with what is used in this video, run at 24v high power. There are tons of dead spots around the coil. The coil varies output power (brightness) a Lot rotating in 3 axis. I was gonna use this to power minis with light up lanterns, torches, etc. It works, but a coin battery (2032) is just better. Better go with a Big D20, and the coil under a THIN dice mat at the bottom of a dice tower. If you really can cast dice perfectly, let me know, I can send you what makes the flickering circuit work.
Given that component prices for LEDs and similar are always in freefalll, the main factor missing here was the time spent prototyping, making mistakes, hitting walls, lessons learned and then the actual production of consistently 100% functional pieces in a mass-production mindset for whatever one's given project is: whether one LED is needed or fifty per project. Factor in whatever your time is worth financially (If you earn $15 at your day job, your down-time is worth the same or more) and then subtract those hours lost in other life-pursuits and I expect that mass-market LED is far cheaper than suggested here. The real questions are always "How much is your time worth?" and "What are you sacrificing by spending that time needed on 'this' instead of 'that'?" In these terms those LEDs could cost a dollar each on the mass-market and still be worth the value brought home when factored vs. your personal time and peace of mind.
I know right, instead of building DIY versions, they could buy the commercial versions and then be spending thier time watching, and commenting on others doing it DIY.
@@notahotshot Or spend that free time commenting on commenting or even commenting on comments commenting the original comments. One wonders where all the time comes from.
They will light up, but tent to flash on and off. Most modern charging pads will detect if there is a phone on the other side and will power off if they detect a foreign object. So you get periodic flashes.
I don’t get it. Why use ESP32 if the original didn’t have it? What does it replace here? Why is a different microcircuit used here than in the original? I really want to repeat this project, but I have very little money and I want to do it as cheaply as possible, so i don't want to buy nothing extra.
It's interesting .. forgetting the resonance .. it looks like an inductor that receives the power .. and a capacitor that rectifies the output with the diode
please do a video on how to track human using ESP 32 cam with servo motors, I try to make a robot like a vector robot so I use your DIY Alexa video for robot talking system now I want how to detect objects and how to track a human.I waiting for your video, please reply
Is it possible to build a modular (linear) charging platform so that the wireless LEDs would be individually addressable via their position in the sequence? (E.g. a strip of 4 "slots", configured so that each slot would light up individually in increasing sequence. In such an arrangement, a person could insert/remove different LEDs in those slots but they would always light in the same sequence) Source: I have many RGBIC led strips at home, but am curious if a more modular platform could be made for repurposable needs.
It's quite hard to get high frequency on the 555 with a breadboard circuit. The mark to space ratio on the standard circuit is also always >50%. Using an MCU gives a nice easy way of generating the required frequency with an easily adjust mark to space ration.
@@KeyCe as a relative beginner to electronics, I've found getting my head round a 555 to be one of the hardest parts so far. Compared to programming microcontrollers to kick out a certain duty cycle, theres a lot to learn for a 555 even if you are just following some configuration instructions. Understanding how they work involves a lot more. Sorta ironic given they were what was churned out when I was school... but then again I didn't go to school in the days of cheap microcontrollers.
Because driving power mosfets at switching speeds of 200kHz from a 5V supply is hard work. They need their gate capacitance to be well charged or discharged to avoid excessive dissipation. That means you need to pick a mosfet that has a total gate charge low enough to be efficiently charged or discharged by the 100mA or so that you can extract from a 555. Because even 100mA will pull the high output of the 555 down to around 3V, you'll also need your mosfet to have a gate threshold voltage well below that. Something with a total gate charge of no more than 10nC and a Vt of around 1.5V would probably be okay at 200kHz, but the only mosfet I can think of with that sort of spec is the AO3400A. You won't get away with just any old "logic-level" mosfet.
What if you wanted somthing that would light up when on the wireless charger, but could be connected to a battery pack/on off switch so you could use it away from the coil as well?
There is a misscalcution in the overall cost. The cost for the MCU is not added. This is OK for testing builds, but if you like to build this fix into a project, the cost of the MCUs had to be added. Then the Chinese version is cheaper. If you need 10 or more kits, then the transportation cost is only 1 or 2$ each. Then the really nice homebrew version is much more expensive. The "drama" of the most DIY projects :-) :-) By the way: Looking for a cool project? If you build houses with LEGO the wiring is often terrible for lightning. So adding the big coil underneath the 1st floor allows a wireless lighting. I'll give it a try. It seems that the small LEDs may be fit into a manipulated clear 2x2 brick.
Indeed. It’s very easy to look at the price and think - that’s just a few components that only cost a few cents. But that ignores things like labour costs - even if you’re doing it yourself you could be spending you time on something else. The DIY option is definitely just for the fun of it. Nice idea for a practical use!
Sure. I think more turns on the drive coil will go a long ways in increasing EFF and distance. Every turn multiplies the magnetic flux, this is why relays/solenoids/etc have hundreds/thousands of turns, to make them more EFF.
Could a circuit be designed to drive LED strip or tape lights so that they could be attached with cable ties or clear tubing to 40vac extension leads so they light up when the cable has mains power applied? That would be useful!
I have thought about building a chess set with wireless LED lights built into the pieces and the coil/s in the board with 2-4 x 18650 batteries.
Thinked just about the same!
Gonna make and sell this, hope you’ve got a patent pending. 😂
that is so genius
Chinese are reading this. Be careful !
That's a lot of batteries.
As always, the build-vs-buy question is usually less about cost and more about specificity - you can tailor the build to a very specific project, whereas finding a pre-built version to meet some projects might be impossible. Plus the fun and learning of doing the build yourself, of course.
There's also the labor aspect--how much time did you spend prototyping, then building and testing the finished circuit(s)? I think this is the second half of the specificity trade-off. Your off-the-shelf components are ready to go, but might not be an exact match for your use case. A custom build can be made to your exact specification, but you have to spend the time to develop it to your specification.
It's not really difficult, i also explained how to make these leds in my channel, but using SMD components, just like the ones they sell (in my kit, the emitter was working at 180khz for a change) and even smaller (the comercial ones uses CD43 inductors, and i managed to use CD32 inductors, a bit lower range and dimmer light, but still working). I'm glad more people is spreading this knowledge arround for DIYers.
Great video!
Thanks! You have steadier hands and better eyesight than me!
What about in £'s sterling.
@@m101ist If you buy the components in reasonable quantities with surface mount components it would be about 33p sterling.
Congratulations on producing an incredibly clear, well informed and crisp video -really a masterpiece of technical explanation.
Thank you:)
Take an ICM7555 with R1 = 1k8 and R2 = 4k7, parallel to R2 a diode from pin 7 towards pin 2 and 6. And a capacitor 1nF to GND. That should be about 222 kHz.
Mayby a high Ohm poti..
Already has 5 volts level for the MOSFET.
If it is not enough, two logic inverters in between, which usually have powerful outputs. First one inverter and the 5 inverter in parallel for the 4000 or 74xx series.
The components are cheap and can be soldered onto a small circuit board. Then there are hardly any hidden capacities and loose contacts.
Great information - makes this a lot more accessible for people who don't have a handy MCU.
So it'll be possible to use it without mos driver??
Definitely worth a go if you have a suitable logic level MOSFET that you can drive with 5v logic. And you can drive enough current to switch it quickly.
An ICM7555 is CMOS and will only drive a few milliamps, so isn't suitable for driving power mosfets directly. The CD4049UB isn't much better, although with five in parallel, you might drive about 50mA cleanly. The standard NE555 or similar should drive 100mA with a 5V supply, but you will still have problems with reduced voltage swing, probably to around 3V.
A typical switching mosfet like the IRFZ44N will turn on with a gate voltage above 4V, but has a gate charge of up to 67nC, which will take 0.67μs to charge/discharge at a constant 100mA. The switching frequency used in the project is around 200kHz, which means an entire cycle takes just 5μs and each half-cycle is around 2.5μs. Having a switching mosfet in its linear region for a quarter of the time will cause problems.
You just don't have enough drive without using a mosfet driver if you've only got a 5V supply to use.
@@wallkiller8080 Not with any common power mosfet on a 5V supply.
This is nice. I was looking all over for how to do this like 5 years ago and came up pretty empty. I don't have the electrical skills to figure it out myself. I do love the very compact design of the original leds.
Sometimes it's just cute to see people using age-old techniques asif it's something totally new and amazing :-)
Oh hey look, it's a douchebag.
Amazing project!!!! Mainly because how you explore how evaluate the components needed in order to make work properly!!! Thank you for sharing!! Perfect for a science fare!!!!!
It would be really cool to have these wireless LEDs inside some dice. Then make a dice tower with the coil in the base. So when you throw the dice they light up!
Awesome, excellent video! Love the thorough explanation
How will you balance the weight?
@@mrtechie6810 9 X 9 X 9. Some could be disabled? 🥰👋
oh nice! Thanks! This is exactly what I was looking for - the plan is to 3d print a set of chess pieces (using clear for white and tinted resin for black) and put LEDs inside, powered from a coil in the board. This will certainly help to build this!
Nice idea! When you've finished send me some pictures or do a little video!
Was thinking the same thing, but for dice. Would have to make a custom jig to center then though for weight.
Perfect usage. They have been using these for years in hobby models. For example gun damn models. When you set the model on the stand. They glow. The eyes and shield and.... glows
Im just happy there are still people, who like me can still enjoy this side of life.
God bless you.
Good stuff, but for those of us who don't have electronic parts lying around already, or who would have no use for the excess that comes with buying in bulk, it's actually cheaper to buy from the Chinese. Still, it's good to have the option!
Indeed - the build vs buy is always a lot more nuanced than simple cost.
Your channel was a find on RUclips. A pearl. Very good ! Very good ! If it's up to me... everyone I know will know your channel. (from Brazil)
This is fascinating, I have a very specific use-case where I will be buying the pre-made units because I need them to be very small, but I also want them to have light emitting all around, so I want to see how many LEDs each part can drive by just soldering extra SMD LEDs to the same leads as the existing one. I'll also have to check if I can do the trick of reversing some of them to use the other half of the waveform. Ideally I'd like 6 LEDs running off of a single inductor in a very compact little package. I"m really hoping there's plenty of overhead in the design, since LEDs really are very low-power devices.
Just some rough calculations on "without a gate driver" (sorry this one intrigued me). Looking at that trace, total time is ~660ns with fully on making up about a third of it. So at 222kHz you'd be spending about 5% of the time turning the thing on and off. Possibly not the end of the world but probably not superb either. (Apologies if I totally stuffed up on theory there - was just a guess)
iv just purchased 6 sets of these for my 1/14 scale truck trailers. all wireless. love em. would have known about them if it was not for you. thank you
Thank you! I am really like your videos with the "deep dive"
explained well enough to go DIY ... thanks a lot
Your teaching technique is amazing, I think I found my favourite channel SUBSCRIBED❗
Would love to see something like this become a wireless Christmas tree lights. Perhaps hot glue on to a tree branch, hang the coil like an ornament to power them and no more wires for the cats to chew. How many lights will one coil power? It looks like each set has about a 9 inch radius, so it seems like it would require a few coils to decorate a tree. Not to mention linking them up. Hmmm. Maybe it's not wireless after all.
In a pinch you could drive the MOSFET with a 555 and save the cost of your processor board and driver chip.
Super surprised how much current it generates without the pull down resistor on the mosfet gate. This little design really makes me think how they work, cheers!
It's really interesting, if you leave the gate floating then it can easily get switched on - and if you're moving a lot of current around it's easy to damage something.
it's more that inductors resist a change in current, so if you keep things on, after a brief moment, you sort of have a short (but wire resistance etc). If you keep things going on and off quickly the inductor never quite gets up to fully conducting (it's busy resisting the constant changes in current). The resistor on the mosfet is more about just being certain it's at zero volts when off. Its a fun experiment to try (without driving big current) to try one without that resistor and see what you get. I've mistakenly forgotten them before and wondered why stuff doesn't turn off.
Actually quick demo why you want that resistor ruclips.net/video/8swJ_Bnsgl4/видео.html you can see even prodding the gate with your finger can be enough to get it (at least partly) on
@@WistrelChianti That totally reminds me why we have RC snubber circuits for inductive loads on solid state relays!
@@TradieTrev haha next level (to me anyhow)
I've been quite fascinated since I first saw your excellent video a couple of months ago. I have my version working with a 555 timer and a trimmer pot to vary the frequency. I couldn't manage to get the duty cycle close to 50%, so I used an 2N7000 mosfet to invert the signal, which also cleaned it up a bit, and then I fed this into a TC4420 mosfet driver. Maybe I'm pushing things a bit, but I've been using 12V across the power mosfet. My bench power supply tells me that I'm consuming about 2W of power, which seems reasonable. Everything seems to be working nicely, though I'm still tweaking things. One thing though:
So far, I've been using a CBB film capacitor on the transmitter coil, but today I noticed that it was getting pretty hot. I suppose I could try turning the voltage down, but I'd rather not, since 12V improves the brightness. I see that you are just using a small ceramic cap, so I was wondering if you might have any thoughts on this?
I've also been using SMD parts on the LED side. CD54/CD43 inductors, and 0805 LEDs and caps. I'm trying to find a way to solder the parts more efficiently. I'm thinking of making some tiny little DIY PCBs with 0.4mm copper clad, and using something like a paper punch to cut them into circles ... unless someone has a better suggestion.
That's great! You've certainly taken it further than I have. I would have a look at the ESR of the capacitor you are using. Maybe you can find one with a lower value that the one you have. I like the idea of the tiny PCBs.
Great info, I am now your big fan
The cost could be improved significantly, using bjt push pull for the gate driving, making an oscilator using an opamp (lm358 is dirt cheap) and you could probably get away with using cheapo mosfet like AO34xx which are cheap even in my country and can even do 5A and turn on even with 2.5V
Would extending the cables from the inductors, in order to have the LEDs at a greater distance from them, affect their operation? I was wondering if I could use such an array in order to "wirelessly" power some LEDs located inside scale models. The concept is as follows:
1)Make a diorama-base with the large coil.
2)Put small coils in appropriate places inside the model (located as low as possible)
3) Run wires from the coils to the LEDs you need to power.
4)Place the finished model on top of the base, making proper alignment with the main coil and the internal ones (located in a certain area inside the model).
Great video, definitely a 'DIY' option....cheers.
Product idea: A gameboard with led game pieces.
This could have been done with a 555 timer with the appropriate passive components, and a mosfet. the microcontroller is overkill.
These are used inside hobby models. Such as Gundam or cars. When you set them on the stand. They glow.
It is interesting to try making a large diameter reel under the table to surprise your friends.
That’s nice , using smd components, the leds can easly be made smaller. I see some comments of people wanting to implement this in a chess board… hmmm in that case tuned receivers could be a way to selectively lit up some leds while others don’t
Would these LEDs work through a wooden door of approx. 4 cm thickness? It would be interesting to have the LEDs wirelessly controlled from inside of e.g. a garage to lit the key hole.
I think so - they worked through some quite thick books. You would need to try it though. You can push more power through the driver coil - power it from 12 or 24 volts.
If you can reduce the driver circuit to key chain size, then you can put LEDs behind key hole. So when you try to enter key it’ll lite the keyhole.
Just turn on a light?🤷
@@mrtechie6810 from outside using bluetooth?
In lord of the rings hobbits, they couldn't find the keyhole in broad day light!! 🗝 🚪 If the keyhole itself shined then they could've save some time and suffering.
@@GhostsOfSparta use a motion-detector floodlight. Or a Bluetooth switch that turns the light on when your phone gets in range.
You could also put a tiny light behind the keyhole, on the inside.
If you use an LED, the power is so low that you could have it on all the time.
Very interesting , thank you for sharing this knowledge
nice. I purchased the wireless led kit from Adafruit, and being inspired and added some components to a mouser order I was placing: 2.2mH inductors ($0.43 each), 240 pF capacitor, and red LEDs. I thought using 2 LEDs was a brilliant move, but it turns out not so brilliant. That is, 1 LED is much brilliant (~2X the brightness of an LED when using 2).
Where did you find 2.2mh inductors at cd43 size?
@@smoll.miniatures It was either Mouser or Newark.
Thanks for doing such a great work.
Nice explanation 👍
you should contact Mr. Tesla for more info! have fun
Bingo LOL He would be an expert & appreciate this demo of induction if he were here today! :)
You shouldMaybe you could give it a go with the bigger coil version too since those have much better range. Great work. Will replicate for my projects.
Very well explained
OK, you reproduced it, now let's see some improvements!
For example, can you get more range with a larger coil size, or more power?
Does tweaking the frequency improve the range?
Are you sure the shrink wrap on the inductors shield anything? This is not the case on the shrink wrapped inductors we use. Works actually slightly better with the wraps since the permeability is 1.00002. The resonance frequency changes quite a bit with the load you put on. Can you test out what it is by varying the oscillator for max power?
How about a single npn transistor with 1.5V battery. This is the simplest and cheapest wireless led solution and it works.
Construct a project that we can use the driver coil under a coffee table or desk and anything placed on them wil light up...
I love to learning 😍
What’s he best way to increase the range of the coil?
I had it working with shielded inductor but on top of a QI charger which pumps out awful lot of power tho...
Now I want to make this-ish but overkill...add a supercapacitor (there are some good cheap tiny options), and as someone else suggested, cast into resin...possibly with coils at right angles - the ali video showed the power drop off significantly when the pixels were rotated against the magnetic field) - add a bridge rectifier, and the cap, and you've got yourself some fun glowy marbles that ...do things.
Nice video. I am trying to think of a nice Escape Room prop to wrap around this tech as it somehow seems a perfect match.
That's an interesting idea - some kind of puzzle where you need to find the right place to put something so it lights up. Or when the lights are arranged so they light up they point to some clue.
@@atomic14 you need to do a collaboration with playful technology channel me thinks :D
What if I just use a 9 volt battery and a coil, skipping all the chips and capacitors and resistors and gates? Just current directly to the coil. Would that over power the LEDs?
It needs the oscillation to transmit the power unfortunately.
Would a small metal dish underneath the coil increase the effective range, or would that interfere with the signal by muddling up the frequency?
I imagine a small metal dish (assuming good conductivity) would make the LEDs not light up or reduce the range at least. The time-varying magnetic field from the base coil would induce currents in the metal dish that aim to reduce the magnetic field from the coil. So it wouldn't interfere by muddling up the frequency, but by reducing the power transferred from the big coil to the LED coils.
Exelent video document, tks 🎩
Thanks for the video. Is very good. Can you say how the construction of LEDs changes so that they light up with the frequency of a mobile phone?
Awesome work! I might have to try this project out. You can get some mcu's for pennies, or try using an oscillator chip or circuit. You can also get those SMD LED's and caps pretty cheap if you buy in bulk, so that will greatly shrink the size, but they're harder to solder. Now I just need to find a use for them.... I'm wondering if there's enough power for a wireless micro bugbot thing.
I think a low lower mcu could definitely be powered. You could also use the energy to charge up a capacitor for a more lower hungry device and then wake it for a brief time to do some work.
I really need to buy some of those please don't wanna say why cause don't want no you stealing my ideal lol this is what I've been looking for
really interesting video - thanks!
Would it be possible to use a simple 555 timer as oscillator? So you would not need a MCU and also no gate driver because you can run the 555 at 5v without problems.
One 555 timer can output only >=50% square wave, also don't know if the 200mA it can output would be enough for such a high frequency
Yes, no problem, and you dont need a FET driver, and if you keep the current under 200mA avg (check your 555 datasheet, not all 555 are built the same), you dont even need a FET.
@@Ozzy3333333 how do you do less than 50% duty cycle with 555?
@@alexstone691 In a standard 555 astable circuit the capacitor is charged through R1 and R2 and discharged through R2 alone. Putting a fast diode in parallel with R2 to short the charge path through R2 allows the duty cycle to approximate R1/R2 and that can be adjusted to give any desired duty cycle. Note that with a 5V supply, the diode drop has a significant effect on the rough calculations, so some experimentation or trimming of values may be necessary. Using a Schottky diode would bring the observed times nearer to the ideal ones. There's a exposition at en.wikipedia.org/wiki/555_timer_IC#Shorter_duty_cycle
@@Ozzy3333333 A standard 555 won't source or sink more than about 100mA from a 5V supply, and you'll find the high output voltage is going to be reduced to around 3V, which is very unlikely to turn on/off a power mosfet properly, even if it is "logic-level".
Very nice weekend project, thank you. Btw when you use a driver you dont need a logic level mosfet, any would do.
The SMD LEDs are not hard to create, just use SMD coils, they are easy to get hold of and definitely can work without the capacitors although they don't perform as well they work well enough, and have to be REALLY out of tune to not work actually, so it's hard to make them not shine at all.
So DIY LEDs + an induction heating driver can work as a starting point for your wireless powered LEDs.
put them on flying saucer that levitate ( you can do it from Floating Pot )
I wonder if this can be scaled up to be effective in say 1 meter X 2 meter applications.
Been looking for wireless uv leds for florescent 3D prints, but couldn't find any, time to put my dropout electrical engineering degree to use
Is there a way to increase the magnetic field? The lights work only a few inches away & I wanted to increase that
Will zvs driver or tesla coil work with the LEDs?
Thank you very much.This is absoluteley stunning.
I was wondering how changing the coil Turns and diameter would affect this built and which physical laws could help me figure it out.
I quite suck at understanding induction, but would like to experiment with this for different output voltages.
Good infomation
In addition to adding a LED in reverse to the main one, try adding 3, 4, 5 or more led's in series (instead of one), the inductive kick voltage can power many led's in series, and no need for the cap across the led imo (actually its not helping at all at that freq), and the same with the drive coil cap, replace it with a string of led's + resistor, its free energy in the since your cap is just snubbing the drive coil kickback voltage, why not use it to power some leds. Try it and report back.
How realistic would it be to tune these on 2.4GHz and let it light up next to the WiFi router / Phone ?
Hmm, tricky, aside from the power from the device and the manner in which it's transmitted, to achieve the resonant frequency of 2.4GHz, even with the smallest value capacitor of 0.1pF, you'd need to find an inductor value of 0.04uH which would be challenging...
@@Umski use PCB traces instead of discrete components.
@@mrtechie6810 maybe, but realistically even a perfect 2.4GHz patch antenna may only achieve at best a -30dBm sensitivity which is 0.001mW of receive power crudely speaking - even the small LEDs shown need more than that to light up :/
Hang a load of these in a large room and make some antenna's
Could be a volumetric representation of the fields?
Let’s say I wanted to make some liquid core resin dice, and I wanted to cast a a wireless led in a bit of shield of resin (maybe just 5-6mm in diameter), put it in the liquid core vessel with a weight on the bottom so that it always sinks to the bottom of the liquid core and shines upward, illuminating the top 70% of the dice.
Now let’s say I wanted the LED to slowly flicker so that it looks like candlelight is radiating from the dice when rolled in a dice tray with the coil in the bottom of it.
Would something like that be possible? I’ve seen people on Etsy put flickering leds in miniature terrains, like wall torches, fireplaces and campfires, burning debris ect, but it’s wired to something that I assume has the flickering element “programmed” into it and then connected to a battery obviously.
What would you have to do with the coil and its power supply (if anything) to cause that flame flicker effect I’m interested in? Thanks in advance.
Yes, but this tech isn't really that bright. I got it to work, but you would need to super tune the transmitting coil with what is used in this video, run at 24v high power. There are tons of dead spots around the coil. The coil varies output power (brightness) a Lot rotating in 3 axis. I was gonna use this to power minis with light up lanterns, torches, etc. It works, but a coin battery (2032) is just better. Better go with a Big D20, and the coil under a THIN dice mat at the bottom of a dice tower. If you really can cast dice perfectly, let me know, I can send you what makes the flickering circuit work.
Terima kasih banyak, ilmu yang sangat bermanfaat
Given that component prices for LEDs and similar are always in freefalll, the main factor missing here was the time spent prototyping, making mistakes, hitting walls, lessons learned and then the actual production of consistently 100% functional pieces in a mass-production mindset for whatever one's given project is: whether one LED is needed or fifty per project.
Factor in whatever your time is worth financially (If you earn $15 at your day job, your down-time is worth the same or more) and then subtract those hours lost in other life-pursuits and I expect that mass-market LED is far cheaper than suggested here.
The real questions are always "How much is your time worth?" and "What are you sacrificing by spending that time needed on 'this' instead of 'that'?" In these terms those LEDs could cost a dollar each on the mass-market and still be worth the value brought home when factored vs. your personal time and peace of mind.
I know right, instead of building DIY versions, they could buy the commercial versions and then be spending thier time watching, and commenting on others doing it DIY.
@@notahotshot Or spend that free time commenting on commenting or even commenting on comments commenting the original comments. One wonders where all the time comes from.
I know this video is 2 years old, but I am wondering if these micro wireless LEDs will power with just a normal wireless charging pad.
They will light up, but tent to flash on and off. Most modern charging pads will detect if there is a phone on the other side and will power off if they detect a foreign object. So you get periodic flashes.
Where did you find the resonant frequency calculator?
Can you public a full schematic?
I don’t get it. Why use ESP32 if the original didn’t have it? What does it replace here? Why is a different microcircuit used here than in the original?
I really want to repeat this project, but I have very little money and I want to do it as cheaply as possible, so i don't want to buy nothing extra.
It's interesting .. forgetting the resonance .. it looks like an inductor that receives the power .. and a capacitor that rectifies the output with the diode
Big Clive said it was a 330pf cap? Which is correct?
you can learn a lot doing that yourself. I would prefer DIY.
It's a surprisingly simple circuit - a good experiment to try and get working.
Nice video, keep it up, thank you :)
Thanks for watching!
please do a video on how to track human using ESP 32 cam with servo motors, I try to make a robot like a vector robot so I use your DIY Alexa video for robot talking system now I want how to detect objects and how to track a human.I waiting for your video, please reply
What about a joule thief to extend distance?
Now that is a great idea!
@@atomic14 now I'm intrigued (only vaguely heard of them)
Big Clive did a good video on them - ruclips.net/video/K53beWYdIpc/видео.html
Is it possible to build a modular (linear) charging platform so that the wireless LEDs would be individually addressable via their position in the sequence?
(E.g. a strip of 4 "slots", configured so that each slot would light up individually in increasing sequence. In such an arrangement, a person could insert/remove different LEDs in those slots but they would always light in the same sequence)
Source: I have many RGBIC led strips at home, but am curious if a more modular platform could be made for repurposable needs.
There are severals ways: have separate coils for each position in which the color data is fed through each coil is one example.
if instead of single LED i use a LED module, I need to change the value of the inductance and the capacitor?
So how can I creat a transmitter that would cover a shelf area on 11 inches X 23 inches or 28cm X 58cm?
why not use 555 + mosfet ?
It's quite hard to get high frequency on the 555 with a breadboard circuit. The mark to space ratio on the standard circuit is also always >50%.
Using an MCU gives a nice easy way of generating the required frequency with an easily adjust mark to space ration.
@@atomic14 @atomic14 FYI hmm.. ruclips.net/video/OXsu29K_Ap4/видео.html
@@KeyCe as a relative beginner to electronics, I've found getting my head round a 555 to be one of the hardest parts so far. Compared to programming microcontrollers to kick out a certain duty cycle, theres a lot to learn for a 555 even if you are just following some configuration instructions. Understanding how they work involves a lot more. Sorta ironic given they were what was churned out when I was school... but then again I didn't go to school in the days of cheap microcontrollers.
Because driving power mosfets at switching speeds of 200kHz from a 5V supply is hard work. They need their gate capacitance to be well charged or discharged to avoid excessive dissipation. That means you need to pick a mosfet that has a total gate charge low enough to be efficiently charged or discharged by the 100mA or so that you can extract from a 555.
Because even 100mA will pull the high output of the 555 down to around 3V, you'll also need your mosfet to have a gate threshold voltage well below that. Something with a total gate charge of no more than 10nC and a Vt of around 1.5V would probably be okay at 200kHz, but the only mosfet I can think of with that sort of spec is the AO3400A. You won't get away with just any old "logic-level" mosfet.
@@RexxSchneider interesting! Thanks for that info
Try doing that with a pancake coil.
What Oscilloscope do you use? you said 220 KHz or so, The ranges I'm seeing for purchase are in MHz (I could be dumb).
What if you wanted somthing that would light up when on the wireless charger, but could be connected to a battery pack/on off switch so you could use it away from the coil as well?
@atomic14 Is it possible to make the LED addressable, to control each LED, especially for RGB. If yes, how?
great
There is a misscalcution in the overall cost. The cost for the MCU is not added. This is OK for testing builds, but if you like to build this fix into a project, the cost of the MCUs had to be added. Then the Chinese version is cheaper. If you need 10 or more kits, then the transportation cost is only 1 or 2$ each. Then the really nice homebrew version is much more expensive. The "drama" of the most DIY projects :-) :-)
By the way: Looking for a cool project? If you build houses with LEGO the wiring is often terrible for lightning. So adding the big coil underneath the 1st floor allows a wireless lighting. I'll give it a try. It seems that the small LEDs may be fit into a manipulated clear 2x2 brick.
Indeed. It’s very easy to look at the price and think - that’s just a few components that only cost a few cents. But that ignores things like labour costs - even if you’re doing it yourself you could be spending you time on something else. The DIY option is definitely just for the fun of it.
Nice idea for a practical use!
Could someone please tell me what mosfet was used in the video? I will thank you very much
Great vid! Are you able to make the tx coil an oblong or pill shape, rather than a full circle?? Thanks
I believe so - I’ve certainly seen coils on aliexpress that are oval shapes.
Is there any value in attempting to maximise efficiency by varying drive freq while looking for peaks or troughs given all the component variations
Sure. I think more turns on the drive coil will go a long ways in increasing EFF and distance. Every turn multiplies the magnetic flux, this is why relays/solenoids/etc have hundreds/thousands of turns, to make them more EFF.
Any way to make a selectable version of this? I.e. to be able to turn on or off a particular LED from the group?
Could a circuit be designed to drive LED strip or tape lights so that they could be attached with cable ties or clear tubing to 40vac extension leads so they light up when the cable has mains power applied?
That would be useful!
I love it good
Надо было раскачивать катушку с помощью генератора на ГУ-78Б, я думаю светодиоды светились бы даже у соседей на три-четыре этажа выше.
"You too can waste power and keep the oligarchs in charge!"
I need these where can I buy them...Thank you