I feel The same way, i never made a whole track before but i have made coils from wrapping wires i stripped off a transformer around a small tube like a straw and flinging small metal rods through it.
I love how tight the Maker/Engineering RUclips space is. Knowing all of my favourite channels are all fans of eachother and enjoy these videos as much as I do makes it even better.
Hey Destin! It looks like both you and Tom tend to like blowing up stuff with DIY guns... But Tom is risking it all doing it inside. Lucky it's not a gallon of mayonnaise ahah🤪
Any reason why you included the word literally when there’s nothing unclear about saying “you’ve made a rail plane.” Literally is just meaningless filler and you’re overusing and misusing the word.
@@remlya Any reason why you included the word meaningless when there’s nothing unclear about saying “Literally is just filler" Meaningless is just filler and you’re being redundant.
I would love to see more of the build and design and engineering process as opposed to the hot dog launching personally, though neither part can stand on its own. They are both important.
You know this is how particle accelerators work? Like the large Hadron collider. The particles are attracted by large electromagnets which then switch to repel the particle away - arranged in a circle this allows the particle to be accelerated to some insane speeds, before they smash it into another particle really fast because why not 🤷♂️
6:42 - Housefly bogey has entered our airspace, interceptor aircraft launched! It would be neat to see an LED on each coil (over the MOSFET output, ~5mA wouldn't matter to the coils) and/or on the Hall effect output. With the high-speed camera, you could see all the timing.
In addition to flipping the magnet polarity halfway through to increase the force applied by each magnet, you can also vary the timing between each sensor detection and the coil activation to improve the output. Since the payload speed varies dramatically along the rail, the optimum delay between sled detection and coil activation will also vary. Conveniently, you can use the time difference between the previous two detections to get an estimate of the current velocity of the sled, so you can actually make this time offset dynamic relative to the mass of your payload with no additional hardware. Some people commented to instead vary the spacing between each coil, but that's just wasting rail space and acceleration distance (not to mention the optimum spacing would vary with payload mass). You have such a good idea of the position and velocity of your sled with the sensor setup that keeping the coils closely packed and varying the timing will give better results. Also note that as suggested by some others, optical sensors might be more tolerant to the substantial magnetic field you're dealing with in case you find you're getting spurious/early triggers due to the field from the previous coils going off, which might be a bigger problem once you're flipping the polarity.
he's using hall effect to trigger the mosfet so the activation would be consistant with the position of the sled, no matter what speed. Am i misunderstanding your comment?
@@eurybaric It's the opposite; there should be LESS delay the faster the projectile is moving. This all happens really fast (to us), we're talking milliseconds or microseconds. Ages to a digital circuit or microprocessor though.
@@eurybaric This would only be true if the entire circuit from hall effect detection to coil reaching full current happens instantaneously, and if the hall effect sensor is spaced at the exact appropriate distance from the coil to maximze energy transfer. In all other scenarios, there will be some time offsets involved which means the optimum activation time will vary with sled velocity. Imagine he tests and finds he gets the most acceleration from a single coil by activating it 5 milliseconds after hall effect detection at a given initial starting velocity. This means optimum position is whatever distance the sled travels in 5ms after the sensor at initial velocity is the ideal spacing, but for the next coil that delay will be lower because the sled will travel more distance in the same time (since it's moving faster after the first push). This is the variable delay he'll need to account for to get optimum output from the whole rail; the time for the sled to travel a bit from the sensor to the optimum activation point minus whatever lag exists in his sensing/driving circuits and coil.
I'm pretty sure that despite having a rail, this is technically a Gauss-gun or Coil-gun. The title of Coil aircraft launcher is correct. If you had connected a bar across two rails that had a current through them then you'd have a true "Rail-gun."
another difference between the two is that rail guns tend to destroy itself due to friction, while coil guns (probably) could be designed to last longer if the projectile isn't in contact with the track
As a former radar technician that worked on systems with Klystrons and thyratrons, seeing you discharging those capacitors by sticking the wire inside the terminals, bare handed, gave me the heebie jeebies.
@Stephen Ferree, you are correct. I don’t think today’s railguns operate in this fashion, as the projectile case is part of the motor in them. This is also why the gun doesn’t last long.
1:40 instead of turning it off you could reverse the positive and negative charges so that the polarity of the magnetic field is reversed and pushes the magnet away
@@davidboktor4743 i was thinking that if you have an alternating timed switch that turns on one connection, say positive end to it, then once the magnet is away, it turns on negative and repeats until its no longer needed
I expect a lot of speed can be gained by modifying the timing. They Halleffect sensor together with the Mosfet and the electronic has a latency, which is the same for each module, but the speed increases along the rail, thus the later magnets tend to stay on longer than optimal
As an addition to this, I wonder at what speed does the inductance of the coils start to matter as far as delaying the full application of the magnetic field. I would imagine faster than it’s currently going but perhaps it’s a consideration for larger/more turn coils (higher inductance) or faster linear motors.
We can hear from the non-increasing frequency of the "brrrrrr" that there is no increase in speed as the object moves along. This means something must be wrong.
When you make Mk.2 of this you should make a spring stopper to catch the sled. Maybe you could even create a magnetic levitating sled for low friction.
Tom, have you considered the eddy currents generated by the moving magnets above the Al rail? Even at that distance it will cause some losses. Great Video (as always :-) CU in the next one.
Yeah I was also wondering if he spent any time tuning the configuration, so to speak. At least on video, he made it seem like he fabricated the first arrangement he thought of. But I imagine he could use experimentation to find the best location to place the Hall Effect sensor, to get as much acceleration from the electromagnet as possible. Too short a duration, and you're leaving potential acceleration on the table, too long, and the electromagnet will hold on to the sled for a moment and slow it down. There must be a sweet spot. He may have already used math or iterative design to find the sweet spot and just didn't include that in the video though.
@@DaimyoD0 If he moves the HE sensor too far back, it'll start to actuate when the PREVIOUS magnet is running, though. Which I suppose is the advantage of using Al guiderails.
I'm being simplistic. But while your coils are attracting the magnet sled, can you also detect the passing sled by each coil & reverse the DC polarity of the specific coil to REPEL the magnet sled to increase the acceleration of the sled?
Slight modification for the launcher. Put a pin across it and cut a small hook into the bottom of the paper plane so you can pull it from IN front of the C of G. The old school whitewings catapults used something like that.
increase the distance between the coils logarithmically... Also you'd think an optical sensor would be less prone to interference than a Hall effect sensor in such a magnetically noisy environment.
@@perjohansson8099 The optical sensor wouldn't be affected at all but I don't know to what extent the hall effect sensors are a problem but they are affected,
Except it isn't. His launcher isn't even close to what the navy trials for their aircraft launchers. They basically use induction to make use of the same effect as railguns do. He only uses reluctance, possibly the weakest form of coil launchers ...
@@lumi_project Yes, but in the video, he specifically talks about induction launchers (aircraft carrier and roller coaster booster). However he himselfe did not make the destinction between reluctance force (what he uses) and induced lorenz force (what the "real world" examples use).
In a rail gun the is one loop of wire, in that the rails consists of the entirety of the circuit bridged by the sled. This device can be called many things but a rail gun is not one of them.
A brilliant idea for the holidays. I’ll try to reverse the direction of the current repelling the magnet further. You could literally carry the whole setup in your back bag. I think it’ll be lots of fun with crazy engineers around me.
I've been working on a circuit like this for a while and i think it can be done without the hall effect sensors. If you use each coil as an analog input for a microcontroller, you will get an induced voltage on the input as the bullet/sled approaches the coil. As soon as you do, switch it to an output and apply current through a mosfet for a fixed period, then switch back to input. Repeat for each coil. This would give you the ability to launch ferromagnetic metals as well as magnets.
@@bharattanwar5047 No, I work on a lot of circuits in simulation on and off. I'd like to build it but it's not my top project. I have had a think about it though, and Ive decided that a hybrid approach might be good to start, both a microcontroller with some SPI ADCs and hall effect sensors so that I can get a timing curve for how the coils are triggered. Then the hall effect sensors could be disabled and the known timings could be used. Having that data would let you make some very fine adjustments to the mechanism.
Small correction, it's a coilgun. The difference is like between a robot and a mech. In a coilgun you have coils. In a railgun the projectile is the coil.
The first thing I thought of was using some sort of flip-flop and a second MOSFET per coil to reverse polarity as the sled passes. Also some sort of bearing for the sled (almost anything really). Glad you're going to be revising and improving! Looks cool!
VERY INTERESTING !!! You may not realize this, but high speed rail lines are just starting to use this Principle to pull/shove monorail cars to their destinations. I studied this principle in tech college in the late 1980's. The coils are only hot in one area, where the train is. Thank you for making your prototype. It was very good !
Hey Tom, did you checked how good you timed the coils? E.g. with your slow motion camera and an led per coil to show if its active? Since magnets are extremely strong close by, hitting the perfect spot to turn the spoils off seems very important.
It's worth noting that the timing might also be different depending on the speed of the projectile, so the coils at the start of the rail might need different timing from the ones near the end. If this is the case, then the mass of the projectile will also affect the timing, especially farther along the track.
One more thought, put a diode across the coils to help reduce the voltage during the collapse of the magnetic field after the voltage is turned off....you should also measure the timing between the collapse of the magnetic field and the start of the next power-on cycle of the next coil to appropriately space out the coils to maximize the velocity.
If you decelerate the sled at the end, you can recover some of the energy you used to accelerate it in the first place. Check out the energy required, and it opens up some cool ideals for reducing fuel requirements for launching aircraft and rockets.
@@benholroyd5221 AFAIK the concept itself is called a Gauss Rifle, it does not have to be a weapon, just a mass driver that uses magnetic fields. As for launching humans, I have read that methods for orbital launch of cargo and people in this way are being investigated
Linear induction motors are also used on the Skytrain (Vancouver's light rail system). So thousands of people use rails guns to get to from work everyday
I would guess this setup is quite sensitive to the synchronization of the coils. If the next coil is turning on too slow, the payload will pass by before experiencing any additional force
@Grim FPV That's right. He is hiding some complicated calculations from us. He surely knows the nominal inductance of his coil pairs and the corresponding field charge time constant. He probably gets saturation on each coil before the physical magnet passes each centerline. Designing such a system for various load weights and field strengths would be a challenge, to be sure. Luck is no small part of this sort of project if you don't have the simulation software and modelling chops during up front design.
@Grim FPV ahaha you sir are correct. I seem to know enough to be dangerous. However small bldc motors controllers precisely work by very fast switching and they reach very high rpms.Those things would have a similar inductance to these linear EMs?
The real version (EMALS) made by General Atomics is a type of linear induction motor. So instead of pulling a permanent magnet armature like yours, they create an accelerating magnetic wave which accelerates an aluminum armature. Very cool demonstration!
You might want to consider spacing out the coils progressively, and making them longer as they stretch further out! A similar thing is done with linear particle accelerators, where the accelerating electrodes become longer and longer so the particle is affected by them for the same amount of time as it accelerates: link to article: en.wikipedia.org/wiki/Linear_particle_accelerator
Great project! Maybe put a led parallel to the coil so we can see in the slow-mo when it gets activated and when it shuts of. Could be helpful to tune the timing too
I'd bet it's more likely the current/time relationship of the coils that can't turn on/ of instantly would be the limiting factor. MOSFETs can switch pretty darn fast, but the coils can't change current instantly.
This is awesome! However I feel that once the magnet reaches a certain speed, which isn't far above that created by the first coil, the other coils have a much reduced contribution due to "back EMF" limiting the current drawn by them. What you need is a constant current driver for all the coils, using a high frequency switching power supply for efficiency, and a higher voltage supply, to force that current through the coils the sled is passing quickly. Or you could set up progressive banks of the coils supplied by higher voltage than the prior ones, though you definitely should make sure the sled doesn't get stuck, or that they still have the current-limiting failsafe. I hope I explained that we'll enough to visualize, I have a different, more electronic background. Cheers! 🍻
I dunno, constant current implies higher and higher voltages for the later coils, and he's running it off a bank of supercaps, or megacaps or whatever they are. They only go up to something like 2.7 volts. I see your point though. Perhaps wiring later coils with thicker wire and fewer turns, so higher current that way? But then magnetic field is a function of the turns. There's surely some sweet spot somewhere but beyond me to think what it is. Perhaps just thicker wire, same turns, bigger spool, would be the right way.
@@greenaum yes you're right. You'd effectively add the 2.7V to whatever back EMF to get constant current, so yea, you couldn't do that with just these caps, that's the caveat. I agree with you 100% that to achieve a similar function would be to alter the coils instead of the voltage, it would just be a different E & I relationship.
Almost! A railgun has the thing being propelled as part of the electrical circuit. The slug closes the connection between the two rails at 90degrees from each, so there's three magnetic fields being made by the same electrical circuit. All three fields are generated simultaneously in series (series in the electrical sense, not the chronological one). The interaction between those three fields is what drives the electromagnetic slug down the rails. This is a gauss gun. The big advantage is being scalable without electrically welding your slug to the rails once you reach a certain amperage.
Rail guns actually work by a slightly different principle to this (it's pretty interesting and I highly suggest you look it up). What Tom's built is more similar to a coil gun I believe. Although a typical coil gun usually has the projectile travel through the middle of the coils tho so this is still pretty unique
Tom: I suspect that if you also complete the magnetic circuit, (possibly by a 'C' loop of steel under each magnet pair), your total flux density will increase. Fun project BTW.
@@jackmclane1826 that is correct. and that is what calculation and/or testing will reveal. considering that the gap is mostly open with a high reluctance - the increase in inductance may not be a problem. I have not designed such a system, so I am going on intuition. (that also can back fire) note also as a for instance: stepper motors, or servos do have relatively tight magnetic circuits with very small gaps. they seem to work. and yes this is all hand waving. a real motors dude would probably be able to say yes or no with more accuracy. but anyway i'll stand with my first comment. being wrong is not a big deal, it is a way to learn.
@@mr1enrollment I'd also try it... the idea is good, but I don't have a clear gut feeling for it right now. It leans towards that switching speed in the end is crucial when you want to really launch something. Not when you want to launch a paper plane at a couple of meters per second, of course... ;) It could even come to the point that an iron core is detrimental due to HF losses. (Thought of it: When you want to launch something at 100m/s that means switching speed in the kHz range is necessary. Don't get close to it with any iron. Maybe a stack of very thin sheets may work. But with massive iron you just make an induction heater and no coil gun... ;) ) My first change would be to invert polarity of the coil when the magnet has passed to push it away. This can also backfire into magic smoke when done wrong... ;) This would effectively half the time that is available for current rise. Very interesting device. I think I put one of these setups on my own experiment list. ;) My first guess would be to move to a lower impedance setting. Thicker wire, less windings but more current meaning lower inductance as no. of windings go squared into inductance.
Yup, deffo, but might be worth looking at how bidirectional ESCs for quadcopters handle that same problem? There must be some good ways of mitigating it
ESC circuits often both switch a little late when there is some back EMF to keep the current down, and they can shunt the excess current to the next set of coils too. But they're mostly designed to dissipate the heat, and they can put out a lot of heat for big ones.
with 4 mosfets per coil (pointing in and out at both ends of the coil) you can control the current direction. and using the internal diodes of the mosfets and leaving one mosfet still ON, you can ensure that current can continue flowing in a circle through the coil.
Some thoughts: You might be able to get more acceleration with different timings, also having 1 or 2 ahead of the sled on may have some boost as well. If you put some optical sensors across rail at the end and a U shaped thing for the sensors a bit out from the end that the opto sensors can give a real time speed indication.
why aren't the magnets spaced out further as you go towards the end? and have you thought about somehow taking the back emf generated from each coil as the field collapses whem you de-energise and using that to boost the power to the next coil?
The back emf would probably be handled by some flyback diodes (or snubber)(maybe with resistors to reduce the current running through the diode) (maybe the diodes just weren't mentioned in the video) or just mosfets that can handle a huge voltage spike. It may also be possible that the coils do not produce much bemf for reasons like eddy currents (but it does seem that they produce a lot of bemf because of the large, fat arcs that occurred when Tom disconnected the coil from the supercaps).
His technology is limited so he can only get the mosfet to turn on so quickly of the projectile gets faster he doesn't have the equipment to match that faster speed therefore spacing it out keeps it in sync in theory
Because that reduces the efficiency in terms of acceleration per rail length. Rather than increasing spacing, you simply decrease timing between electromagnets as the cart progresses.
It is always something special to make a plane fly! This video made me wonder if the paper plane folding experts among your viewers might not have ideas for models that are able to remain longer in the sled, i.e get a greater acceleration.
@@seskalarafey9285 Even if the controller is infinitely fast you still have less and less time, so the impulse might stay the same but the total energy is still only fractions of a second, and the faster you accelerate it the less time you have to accelerate it meaning the "barrel length" or coil count starts diminishing. So you definitely want to do that to get the most out of the system, but even with theoretically instant controller you're not going to get double the speed but you still get an improvement without increasing size or coil count. I think most mosfets are in the millionth of a second on/off time, but if ;you're using an H-bridge to drive it then you'll have more latency, still though, a proper controller for this should be insanely fast and let you get 95% of the theoretical limit out of it. Biggest improvement would be how he detects the object as you'll want to play with those positions in case of latency in the system.
@@jmikronis7376 it's more laws of physics. They can make the controllers switch near infinitely fast, but we also are mostly limited by the strength of whatever the coils can handle and the time they can impart force to the object. I think it becomes more of a mechanical design challenge than anything, because now we need a better design than just a linear rail: something like a looping track in the rear so it can pre-accelerate it making several loops before entering the straight track. Or we may just end up using a spinning sling arm etc. Also I didn't list other unfortunate physics effects such as Eddy currents and effects of fast moving magnets next to coils which is really more of a materials engineer design problem.
except this is not nearly powerful enough to be dangerous, not for Colin at least. [0.015kg*(11m/s)^] /2 = 0.9J muzzle energy. Some permitted airsoft guns go over 5J.
With your high-speed camera you could also try tuning the on-times for the coils - as the payload travels down the track and accelerates it's likely that the coils should be triggered a little earlier for maximum acceleration (because it takes a finite time for the current to build in each coil).
This. Also, another very cool improvement would be printing a holding mechanism in the 'sled' to hold the plane. By the end of the rail the tail is pushed down and the angle of attack is way too high. The holding mechanism could be something like a moving piece inside the sled that holds the plane like a thumb and hits another 3d printed part at the end of the rail, moving it to the release position.
Wow a beautiful implementation.. The big difficulty with railguns is that as you get faster and longer the timing of the pulses gets shorter and more critical. This also means the power loading each coil unit needs gets higher too..(since it gets divided by time) Above a certain speed things start to get exotic very quickly. (over about 80 - 300 m/s) My own work was only crude and theoretical - with the idea of using a railgun like a rocket engine. Not really a good idea for about a dozen reasons, the killer being that it just took too much energy to ever work. Another being that the rocket 'exhaust' was rather dangerous.
If you fire off in a vacuum, you essentially have the makings of a magnetic space launcher, either on Earth or in space, even space craft launcher , or a hyper loop transporter system! Or in this case a aircraft launcher, or a coil or rail gun!
5:43 Interesting strobing effect. Three questions Tom: What is the frequency of your mains electricity? What was the frame rate of your high speed video footage? What was the shutter speed?
Ещё желательно мосфеты расположить максимально близко от датчиков холла до катушек, чтоб снизить временное влияние. А так же попробовать с разными расстояниями этих датчиков холла до управляемой катушки, чтобы добиться максимального эффекта.
U.S. Navy: "We build railguns to shoot bad guys." Tom Stanton: "We build railguns to launch household items. For our amusement and all the giggles." 😁😁😁
Its not quite a railgun though, its a coil gun. Coilguns magnetically attract the projectile, and railguns repel the projectile via the Lorentz force (which requires current running though the projectile)
i've always wondered if you could build one of these in a loop where one portion of the loop had a falling gate that allowed you to fire it once up to the speed you want. and if that would allow you to get much faster speeds in a much smaller package?
Hey Mr Stanton, how about instead of turning the coils off you switch the polarity. You should get an increase of acceleration of at least 60% :D (But my math is very basic and probably wont cut it). Please try, would love to see another video about this :D
Nice 👌🏻 You know what Tom, try magnetically levitating that 15gm weight so friction is even less and also make it aerodynamic like bullet shape or something to achieve even more velocity and efficiency!
that first coaster in the video is called the skyrocket! its in a park called Kennywood outside of Pittsburgh. great ride and kinda neat to see in a video!
Great project, Tom! I was also thinking; You can make it 2 times faster by not only switching on/off but also reversing the current direction. To do that, replace each MOSFET with an H-Bridge.
Could you make a sled that holds the airplane in place until it reaches the end of the slot, and then releases it? It would get more speed from the sled, and then shoot up after release. I'm trying to imagine how you do that without introducing too much friction or weight to the sled, though...
Imagine a sled Mechanically designed that it can only release the plane if the planes goes forward. Put a stopper at the end of the rail, as soon as the sled touches it, the plane is launched since it's now free to go past the sled. It could be a pinching mechanism in tension by a spring, the pinching part on top and the releasing mechanism on the bottom. At the end of the rail, there's two reducing rails that apply abruptly pressure on releasing levers at the same time and the plane is released. 😅
Very low pressure pen refill spring with a stopper to release at dead-end. It may launch earlier incase lift exceeds spring hold force without hindrance to either of 3 mechanism.
I think you could see a performance increase by getting rid of the control board and instead using open-drain sensors. That way, the sensor can turn on the electromagnet by itself. It might speed things up and let you tighten the timing margins. Also, the idea to reverse the current through the coils is interesting. Are the coils low-inductance enough to make that feasible?
@@shreyasshenoy7215 you are going to need the extreme amounts of current that the super capacitors can output, Batteries cant match a capacitors ability to discharge and charge quickly.
Optimally you reverse the electric polarity to repulse the sled magnate so you push as well as pull. Edit, I suggested it before he mentioned it at the end.
0:15 HECK YEAH! KENNYWOOD! Also the MOSFET is also know as a transistor and is opened a small amount of amperage, not voltage. Although, voltage is still needed to push amperage through a wire.
Tom, that was so far the most interesting video I have watched on you channel. I would love to see a real Concorde (of course a model) starting from your rails!
Hey Tom, love the channel and all the crazy stuff you make. I have a couple of ideas for you to improve your device. 1st, make sure you have freewheeling diodes across your coils so that you don't overvoltage your FETs during turn off. You may already have them in there and I just didn't see them. 2nd, you should devise a way to check the timing of your coils relative to the position of the sled. As the sled accelerates, the turn on of subsequent coils need to be advanced in time compared to the upstream end. I would not be surprised if the sled is being slowed down near the end of the track. You could use LEDs on each coil and high speed photography to visualize the timing. I suspect you may need to move the hall probes upstream as you go down the track to give the same timing advance as you had near the beginning. This brings me to #3, you may want to explore spacing the coils farther apart as you go down the track. This will also alter the timing of the system. Again, cool builds and I look forward to seeing more.
"Aircraft Launcher" , thats a freaking gun Tom, a cool gun though
You should be making vids not watching them 😉😉👌
Maybe a launch system for your future rockets?
@Integza It would probably destroy tomatos!
Was about to say same
Rail gun it is then!
I love your channel so much. When I was a kid I tried to make a maglev track in my room. This makes me feel the same childlike wonder.
I feel The same way, i never made a whole track before but i have made coils from wrapping wires i stripped off a transformer around a small tube like a straw and flinging small metal rods through it.
I love how tight the Maker/Engineering RUclips space is. Knowing all of my favourite channels are all fans of eachother and enjoy these videos as much as I do makes it even better.
Hey Destin! It looks like both you and Tom tend to like blowing up stuff with DIY guns... But Tom is risking it all doing it inside. Lucky it's not a gallon of mayonnaise ahah🤪
@@mateosegura1520 Seriously! :D
You're still a kid Destin!
5:13 That look of betrayal, staring directly into the camera. That cheese stick trusted you.
Nice work on those slomo shots in this one
who you calling slomo
The shot where there was a beetle in slow motion really captured it! Couldn't have planned it better.
Yo im early
when you and your sibling are supposed to be asleep ruclips.net/video/RXqP7l3TXHA/видео.html
He should put a lamp on the subject to reduce the noise
Those jet sounds are fire, Tom. You’ve literally made a rail plane. Love it.
Yes
Any reason why you included the word literally when there’s nothing unclear about saying “you’ve made a rail plane.” Literally is just meaningless filler and you’re overusing and misusing the word.
@@remlya you are literally angry aren't you?
@@remlya Any reason why you included the word meaningless when there’s nothing unclear about saying “Literally is just filler" Meaningless is just filler and you’re being redundant.
when you and your sibling are supposed to be asleep ruclips.net/video/RXqP7l3TXHA/видео.html
So many comments, but no one mentions :
The bee captured at 00:06:41flying as fast as the paper plane.
Tom: "Your turn little one"
Coil: "Are you winding me up?"
Hmm
That's it! I'm calling the coppers!
Tom: "That depends, are you attracted or repelled?"
yes, cause i m attracted to you
Joel?!
I'm glad I stayed untill the end. Seeing those planes being launched was awesome!
the video really took flight at the end
I would love to see more of the build and design and engineering process as opposed to the hot dog launching personally, though neither part can stand on its own. They are both important.
Always excellent videos. Bonus material......Slow motion bee captured at 6:44 in the background, racing the plane at take off.
I thought it was a fly
@@wayfa13 Because it is a fly
@@markzaikov456 Yip
ok
Looked like a Cicada with such big flappy wings.
Tom: "Aircraft Launcher"
Everyone else: "Railgun"
its a coil gun not rail gun
Just need thicker wire and Mo Powa 👶.
Slap a barrel on with a covering for looks.
Can call it an ElectroRifle.
Surprised noone is calling it a gauss rifle, im used to this term more than railgun, from playing a lot of fallouts and stalkers!
@@mrcourgette6275 - It's a rail gun - look up the difference.
Correct
You know this is how particle accelerators work? Like the large Hadron collider. The particles are attracted by large electromagnets which then switch to repel the particle away - arranged in a circle this allows the particle to be accelerated to some insane speeds, before they smash it into another particle really fast because why not 🤷♂️
@@alicenulla5264 just yeets it across the universe at nearly the speed of light 😆
Just when I thought your videos couldn't get much better, you go ahead and weaponise a sausage
It caught me off gaurd
@@甘いお茶漬け (edited) wanted to add a typo, nice, what were you guarding ?
@@jyvben1520 Dude are you okay? It’s just a typo.
@@甘いお茶漬け youtube is also educational, anyone can make a mistake, correcting it is not a crime.
@@jyvben1520 No dude, I made one small typo and you are getting all worked up about it
6:42 - Housefly bogey has entered our airspace, interceptor aircraft launched! It would be neat to see an LED on each coil (over the MOSFET output, ~5mA wouldn't matter to the coils) and/or on the Hall effect output. With the high-speed camera, you could see all the timing.
that was a very fast housefly bogey
Lol i wondered who else saw it
Was coming here to see if anyone else noticed the fly!
@@mr2ben Yup me too!!
6:42 That fly was so glad that she could make it.
In addition to flipping the magnet polarity halfway through to increase the force applied by each magnet, you can also vary the timing between each sensor detection and the coil activation to improve the output.
Since the payload speed varies dramatically along the rail, the optimum delay between sled detection and coil activation will also vary. Conveniently, you can use the time difference between the previous two detections to get an estimate of the current velocity of the sled, so you can actually make this time offset dynamic relative to the mass of your payload with no additional hardware.
Some people commented to instead vary the spacing between each coil, but that's just wasting rail space and acceleration distance (not to mention the optimum spacing would vary with payload mass). You have such a good idea of the position and velocity of your sled with the sensor setup that keeping the coils closely packed and varying the timing will give better results.
Also note that as suggested by some others, optical sensors might be more tolerant to the substantial magnetic field you're dealing with in case you find you're getting spurious/early triggers due to the field from the previous coils going off, which might be a bigger problem once you're flipping the polarity.
he's using hall effect to trigger the mosfet so the activation would be consistant with the position of the sled, no matter what speed. Am i misunderstanding your comment?
Wait i think my question now is why is there a delay the higher the velocity?
@@eurybaric It's the opposite; there should be LESS delay the faster the projectile is moving. This all happens really fast (to us), we're talking milliseconds or microseconds. Ages to a digital circuit or microprocessor though.
Optical sensor would make sense.
@@eurybaric This would only be true if the entire circuit from hall effect detection to coil reaching full current happens instantaneously, and if the hall effect sensor is spaced at the exact appropriate distance from the coil to maximze energy transfer. In all other scenarios, there will be some time offsets involved which means the optimum activation time will vary with sled velocity.
Imagine he tests and finds he gets the most acceleration from a single coil by activating it 5 milliseconds after hall effect detection at a given initial starting velocity. This means optimum position is whatever distance the sled travels in 5ms after the sensor at initial velocity is the ideal spacing, but for the next coil that delay will be lower because the sled will travel more distance in the same time (since it's moving faster after the first push). This is the variable delay he'll need to account for to get optimum output from the whole rail; the time for the sled to travel a bit from the sensor to the optimum activation point minus whatever lag exists in his sensing/driving circuits and coil.
I'm pretty sure that despite having a rail, this is technically a Gauss-gun or Coil-gun. The title of Coil aircraft launcher is correct. If you had connected a bar across two rails that had a current through them then you'd have a true "Rail-gun."
S.E.M.L
(Supersonic electro magnet launcher)
another difference between the two is that rail guns tend to destroy itself due to friction, while coil guns (probably) could be designed to last longer if the projectile isn't in contact with the track
This is a Gauss
The difference between gauss and coil is that coil is a single magnet pair, gauss uses a series for additional acceleration and force
YT does not like the 'G' word.
@@oompalumpus699 Zimmerman St. Charles is dead
As a former radar technician that worked on systems with Klystrons and thyratrons, seeing you discharging those capacitors by sticking the wire inside the terminals, bare handed, gave me the heebie jeebies.
Tom, put an Led next to (electrically accross) each coil. It would be cool to see them switch on in slow motion.
Yep. For all the geeks, so says Captain Obvious.
This is a great idea! It’ll help with making sure they’re all timed properly for maximum acceleration too!
Linear induction motors are insane!!
@Stephen Ferree, you are correct.
I don’t think today’s railguns operate in this fashion, as the projectile case is part of the motor in them. This is also why the gun doesn’t last long.
1:40 instead of turning it off you could reverse the positive and negative charges so that the polarity of the magnetic field is reversed and pushes the magnet away
Thats a really good point. How would one be able to reverse the positive and negative charges?
@@davidboktor4743 i was thinking that if you have an alternating timed switch that turns on one connection, say positive end to it, then once the magnet is away, it turns on negative and repeats until its no longer needed
I expect a lot of speed can be gained by modifying the timing. They Halleffect sensor together with the Mosfet and the electronic has a latency, which is the same for each module, but the speed increases along the rail, thus the later magnets tend to stay on longer than optimal
As an addition to this, I wonder at what speed does the inductance of the coils start to matter as far as delaying the full application of the magnetic field. I would imagine faster than it’s currently going but perhaps it’s a consideration for larger/more turn coils (higher inductance) or faster linear motors.
We can hear from the non-increasing frequency of the "brrrrrr" that there is no increase in speed as the object moves along. This means something must be wrong.
6:41 - It flies as fast as a fly in the middle of the screen. How did you do that? How did you manage to teach a fly to fly along your paper plane?
that fly must be going 11m/s
heavy air traffic that day
With the help of Microcenter!
magnetic fly i guess
We missed the start where the fly is like "You wanna go, bro? Ya, you wanna?! C'mon let's go!!!"
When you make Mk.2 of this you should make a spring stopper to catch the sled. Maybe you could even create a magnetic levitating sled for low friction.
If you make a make a magnetic levitated sled, wouldn't it interfere with the magnetic field of the coil?
The sound effects at 1:45 got you a like and a sub!
Tom, have you considered the eddy currents generated by the moving magnets above the Al rail? Even at that distance it will cause some losses. Great Video (as always :-) CU in the next one.
Yeah I was also wondering if he spent any time tuning the configuration, so to speak. At least on video, he made it seem like he fabricated the first arrangement he thought of. But I imagine he could use experimentation to find the best location to place the Hall Effect sensor, to get as much acceleration from the electromagnet as possible. Too short a duration, and you're leaving potential acceleration on the table, too long, and the electromagnet will hold on to the sled for a moment and slow it down. There must be a sweet spot. He may have already used math or iterative design to find the sweet spot and just didn't include that in the video though.
@@DaimyoD0 If he moves the HE sensor too far back, it'll start to actuate when the PREVIOUS magnet is running, though. Which I suppose is the advantage of using Al guiderails.
I'm being simplistic. But while your coils are attracting the magnet sled, can you also detect the passing sled by each coil & reverse the DC polarity of the specific coil to REPEL the magnet sled to increase the acceleration of the sled?
@@XYZY64 Did you not watch to the end? That's his next design iteration.
The hardest part is switching that match Power quick enough. Also reversing the polarization of the coil has a pretty slow characteristic curve
You can shorten the rail by 2 times or leave it long (increase speed). By adding the function of changing polarity when the projectile leaves the coil
Thats what came across my mind too
6:47 he quite literally says that in the video...
Slight modification for the launcher. Put a pin across it and cut a small hook into the bottom of the paper plane so you can pull it from IN front of the C of G. The old school whitewings catapults used something like that.
Not wholly dissimilar to the connecting bars used by real aircraft catapults. Now Tom just needs to devise an arrestor system . . .
I just experienced a ratatouille style flashback when I read "whitewings"
@@kingsizedmidget7294 I just experienced a ratatouille style flashback when I read "ratatouille"
used to love those sets.
increase the distance between the coils logarithmically... Also you'd think an optical sensor would be less prone to interference than a Hall effect sensor in such a magnetically noisy environment.
why?
@@perjohansson8099 WHY NOT?!
@@perjohansson8099 Complete guess, but as it moves faster, it would spend less time near each coil in the linear arangement
@@perjohansson8099 - to perhaps match the acceleration curve a little better
@@perjohansson8099 The optical sensor wouldn't be affected at all but I don't know to what extent the hall effect sensors are a problem but they are affected,
4:36
Tell me you're at a spar, without telling me you're at a spar
I love how everyone knows it's a railgun without him having to say the word even once lol
but it isn't though
Except it isn't.
His launcher isn't even close to what the navy trials for their aircraft launchers.
They basically use induction to make use of the same effect as railguns do.
He only uses reluctance, possibly the weakest form of coil launchers ...
@@lumi_project a chicken is a bird. a duck is a bird. a duck is not a chicken.
@@lumi_project
Yes, but in the video, he specifically talks about induction launchers (aircraft carrier and roller coaster booster).
However he himselfe did not make the destinction between reluctance force (what he uses) and induced lorenz force (what the "real world" examples use).
In a rail gun the is one loop of wire, in that the rails consists of the entirety of the circuit bridged by the sled. This device can be called many things but a rail gun is not one of them.
A brilliant idea for the holidays. I’ll try to reverse the direction of the current repelling the magnet further.
You could literally carry the whole setup in your back bag. I think it’ll be lots of fun with crazy engineers around me.
I've been working on a circuit like this for a while and i think it can be done without the hall effect sensors. If you use each coil as an analog input for a microcontroller, you will get an induced voltage on the input as the bullet/sled approaches the coil. As soon as you do, switch it to an output and apply current through a mosfet for a fixed period, then switch back to input. Repeat for each coil. This would give you the ability to launch ferromagnetic metals as well as magnets.
So, bro did u complete yours.
@@bharattanwar5047 No, I work on a lot of circuits in simulation on and off. I'd like to build it but it's not my top project. I have had a think about it though, and Ive decided that a hybrid approach might be good to start, both a microcontroller with some SPI ADCs and hall effect sensors so that I can get a timing curve for how the coils are triggered. Then the hall effect sensors could be disabled and the known timings could be used. Having that data would let you make some very fine adjustments to the mechanism.
@@silverywingsagain i am not able to understand electrical circuit because Tom didn't explain it. That's why i asked u 😅.
Military: Noooo, You can‘t just have a railgun at home.
Tom: Haha, Paperplane launcher goes brrrrrr
Small correction, it's a coilgun.
The difference is like between a robot and a mech.
In a coilgun you have coils.
In a railgun the projectile is the coil.
The first thing I thought of was using some sort of flip-flop and a second MOSFET per coil to reverse polarity as the sled passes. Also some sort of bearing for the sled (almost anything really). Glad you're going to be revising and improving! Looks cool!
H-bridge is the standard ticket.
VERY INTERESTING !!!
You may not realize this, but high speed rail lines are just starting to use this Principle to pull/shove monorail cars to their destinations. I studied this principle in tech college in the late 1980's. The coils are only hot in one area, where the train is.
Thank you for making your prototype. It was very good !
Hey Tom,
did you checked how good you timed the coils? E.g. with your slow motion camera and an led per coil to show if its active? Since magnets are extremely strong close by, hitting the perfect spot to turn the spoils off seems very important.
It's worth noting that the timing might also be different depending on the speed of the projectile, so the coils at the start of the rail might need different timing from the ones near the end. If this is the case, then the mass of the projectile will also affect the timing, especially farther along the track.
@@arantala Perhaps optical sensors.
One more thought, put a diode across the coils to help reduce the voltage during the collapse of the magnetic field after the voltage is turned off....you should also measure the timing between the collapse of the magnetic field and the start of the next power-on cycle of the next coil to appropriately space out the coils to maximize the velocity.
thats actually really smart
work on your grammar man
Great work. 👍
This is the only video that makes electronic magnets make sense to me
If you decelerate the sled at the end, you can recover some of the energy you used to accelerate it in the first place. Check out the energy required, and it opens up some cool ideals for reducing fuel requirements for launching aircraft and rockets.
So now he has a Gauss Rifle. Sweet :)
When are you launching your own spaceship? :D
It isn't a rifle. I'm not sure you'd want that if you're launching humans anyway.
@@benholroyd5221
AFAIK the concept itself is called a Gauss Rifle, it does not have to be a weapon, just a mass driver that uses magnetic fields.
As for launching humans, I have read that methods for orbital launch of cargo and people in this way are being investigated
@@ursa0607 I rifle imparts spin. This would just be a Gauss gun.
Linear induction motors are also used on the Skytrain (Vancouver's light rail system). So thousands of people use rails guns to get to from work everyday
I would guess this setup is quite sensitive to the synchronization of the coils. If the next coil is turning on too slow, the payload will pass by before experiencing any additional force
@Grim FPV And pull the hall sensors back further too to trigger the coil earlier.
This is a thing for design engineers to solve.
@Grim FPVmosfets can switch on and off at speeds above 50kHz. This thing is way to slow for you to even consider increasing the spacing.
@Grim FPV That's right. He is hiding some complicated calculations from us. He surely knows the nominal inductance of his coil pairs and the corresponding field charge time constant. He probably gets saturation on each coil before the physical magnet passes each centerline. Designing such a system for various load weights and field strengths would be a challenge, to be sure. Luck is no small part of this sort of project if you don't have the simulation software and modelling chops during up front design.
@Grim FPV ahaha you sir are correct. I seem to know enough to be dangerous. However small bldc motors controllers precisely work by very fast switching and they reach very high rpms.Those things would have a similar inductance to these linear EMs?
The real version (EMALS) made by General Atomics is a type of linear induction motor. So instead of pulling a permanent magnet armature like yours, they create an accelerating magnetic wave which accelerates an aluminum armature.
Very cool demonstration!
Like an squirrel cage induction motor, nice.
That hotdog slow-mo was exactly what I needed today! keep up the content!
You might want to consider spacing out the coils progressively, and making them longer as they stretch further out!
A similar thing is done with linear particle accelerators, where the accelerating electrodes become longer and longer so the particle is affected by them for the same amount of time as it accelerates:
link to article: en.wikipedia.org/wiki/Linear_particle_accelerator
I love that the slow mo capture also caught a bee in the background flying parallel to the fence at 6:42.
I mean it's a fly but ok
alternative title : "I made an paper plane launcher" (6:00)
(it was really cool btw)
Great project! Maybe put a led parallel to the coil so we can see in the slow-mo when it gets activated and when it shuts of. Could be helpful to tune the timing too
It would be interesting to put that in a loop, I wonder what the max speed would be
i woud guess ,if you neglect the friction only thing is mosfet switching speed that is the limit
@@Shreyam_io Mmm, I love the smell of toasty mosfets!
that's just a regular motor with extra steps
I'd bet it's more likely the current/time relationship of the coils that can't turn on/ of instantly would be the limiting factor. MOSFETs can switch pretty darn fast, but the coils can't change current instantly.
you’ve invented a motor
2:29 that solder job is amazing
This is awesome! However I feel that once the magnet reaches a certain speed, which isn't far above that created by the first coil, the other coils have a much reduced contribution due to "back EMF" limiting the current drawn by them. What you need is a constant current driver for all the coils, using a high frequency switching power supply for efficiency, and a higher voltage supply, to force that current through the coils the sled is passing quickly. Or you could set up progressive banks of the coils supplied by higher voltage than the prior ones, though you definitely should make sure the sled doesn't get stuck, or that they still have the current-limiting failsafe.
I hope I explained that we'll enough to visualize, I have a different, more electronic background.
Cheers! 🍻
I dunno, constant current implies higher and higher voltages for the later coils, and he's running it off a bank of supercaps, or megacaps or whatever they are. They only go up to something like 2.7 volts. I see your point though. Perhaps wiring later coils with thicker wire and fewer turns, so higher current that way? But then magnetic field is a function of the turns. There's surely some sweet spot somewhere but beyond me to think what it is. Perhaps just thicker wire, same turns, bigger spool, would be the right way.
@@greenaum yes you're right. You'd effectively add the 2.7V to whatever back EMF to get constant current, so yea, you couldn't do that with just these caps, that's the caveat. I agree with you 100% that to achieve a similar function would be to alter the coils instead of the voltage, it would just be a different E & I relationship.
In this episode of "Tom Stanton invents stuff" He makes a railgun!
Almost! A railgun has the thing being propelled as part of the electrical circuit. The slug closes the connection between the two rails at 90degrees from each, so there's three magnetic fields being made by the same electrical circuit. All three fields are generated simultaneously in series (series in the electrical sense, not the chronological one). The interaction between those three fields is what drives the electromagnetic slug down the rails.
This is a gauss gun. The big advantage is being scalable without electrically welding your slug to the rails once you reach a certain amperage.
Rail guns actually work by a slightly different principle to this (it's pretty interesting and I highly suggest you look it up). What Tom's built is more similar to a coil gun I believe. Although a typical coil gun usually has the projectile travel through the middle of the coils tho so this is still pretty unique
it's a coilgun.
No no, it's part 1 of his model of Cern.
:)
AYYYYYY THE ROLLERCOASTER AT THE BEGINNING IS THE SKYROCKET, MY FAVORITE RIDE AT MY HOME PARK
I'm a little surprised you didn't make a self-releasing hook to tow the plane from the nose for better takeoff/flight performance.
Tom: I suspect that if you also complete the magnetic circuit, (possibly by a 'C' loop of steel under each magnet pair), your total flux density will increase. Fun project BTW.
True. But it will also increase the inductance of the coil and will reduce the speed the current rises... this can backfire!
@@jackmclane1826 that is correct. and that is what calculation and/or testing will reveal.
considering that the gap is mostly open with a high reluctance - the increase in inductance may not be a problem. I have not designed such a system, so I am going on intuition. (that also can back fire) note also as a for instance: stepper motors, or servos do have relatively tight magnetic circuits with very small gaps. they seem to work. and yes this is all hand waving. a real motors dude would probably be able to say yes or no with more accuracy. but anyway i'll stand with my first comment. being wrong is not a big deal, it is a way to learn.
@@mr1enrollment I'd also try it... the idea is good, but I don't have a clear gut feeling for it right now. It leans towards that switching speed in the end is crucial when you want to really launch something. Not when you want to launch a paper plane at a couple of meters per second, of course... ;) It could even come to the point that an iron core is detrimental due to HF losses. (Thought of it: When you want to launch something at 100m/s that means switching speed in the kHz range is necessary. Don't get close to it with any iron. Maybe a stack of very thin sheets may work. But with massive iron you just make an induction heater and no coil gun... ;) )
My first change would be to invert polarity of the coil when the magnet has passed to push it away. This can also backfire into magic smoke when done wrong... ;) This would effectively half the time that is available for current rise.
Very interesting device. I think I put one of these setups on my own experiment list. ;) My first guess would be to move to a lower impedance setting. Thicker wire, less windings but more current meaning lower inductance as no. of windings go squared into inductance.
@@jackmclane1826 let me know when your video is up
@@mr1enrollment I'm not making videos. I'm not allowed to film where I do that kind of nonsense... (and I'm not too keen on publicity).
nice shot with the fly in the background at 6:42 badass
Trying to get the coil current to zero instantly is gonna probably fry the mosfets in the long run
Yup, deffo, but might be worth looking at how bidirectional ESCs for quadcopters handle that same problem? There must be some good ways of mitigating it
ESC circuits often both switch a little late when there is some back EMF to keep the current down, and they can shunt the excess current to the next set of coils too. But they're mostly designed to dissipate the heat, and they can put out a lot of heat for big ones.
with 4 mosfets per coil (pointing in and out at both ends of the coil) you can control the current direction. and using the internal diodes of the mosfets and leaving one mosfet still ON, you can ensure that current can continue flowing in a circle through the coil.
Eh. Mosfets are cheap. Let em fry!
Some thoughts: You might be able to get more acceleration with different timings, also having 1 or 2 ahead of the sled on may have some boost as well. If you put some optical sensors across rail at the end and a U shaped thing for the sensors a bit out from the end that the opto sensors can give a real time speed indication.
Nice job. And as an EE, excellent description for the layman.
And then the inner child comes to the surface, which was fun to watch.
why aren't the magnets spaced out further as you go towards the end?
and have you thought about somehow taking the back emf generated from each coil as the field collapses whem you de-energise and using that to boost the power to the next coil?
The back emf would probably be handled by some flyback diodes (or snubber)(maybe with resistors to reduce the current running through the diode) (maybe the diodes just weren't mentioned in the video) or just mosfets that can handle a huge voltage spike. It may also be possible that the coils do not produce much bemf for reasons like eddy currents (but it does seem that they produce a lot of bemf because of the large, fat arcs that occurred when Tom disconnected the coil from the supercaps).
His technology is limited so he can only get the mosfet to turn on so quickly of the projectile gets faster he doesn't have the equipment to match that faster speed therefore spacing it out keeps it in sync in theory
Because that reduces the efficiency in terms of acceleration per rail length. Rather than increasing spacing, you simply decrease timing between electromagnets as the cart progresses.
well done😁 you made a electro magnetic rail gun 4:00
3:52 makes it look like the barrel of a railgun
It is always something special to make a plane fly! This video made me wonder if the paper plane folding experts among your viewers might not have ideas for models that are able to remain longer in the sled, i.e get a greater acceleration.
I've watched this 3 times today and every time I'm still impressed by your smooth segue into the sponsored content
I imagined a model roller coaster using this design and you did it. AWESOME 3000!
I also thought reversing the polarity of the coils the moment the sled passes them would double the speed😅 But apparently you were faster👌
But the time to Shut on/Off the coils is getting shorter because of the increased speed. Is there a Limit on the On/Off function on the controllers?
@@seskalarafey9285 Even if the controller is infinitely fast you still have less and less time, so the impulse might stay the same but the total energy is still only fractions of a second, and the faster you accelerate it the less time you have to accelerate it meaning the "barrel length" or coil count starts diminishing. So you definitely want to do that to get the most out of the system, but even with theoretically instant controller you're not going to get double the speed but you still get an improvement without increasing size or coil count.
I think most mosfets are in the millionth of a second on/off time, but if ;you're using an H-bridge to drive it then you'll have more latency, still though, a proper controller for this should be insanely fast and let you get 95% of the theoretical limit out of it. Biggest improvement would be how he detects the object as you'll want to play with those positions in case of latency in the system.
@@jakegarrett8109, a solution to be solved by electrical engineers.
@@jmikronis7376 it's more laws of physics. They can make the controllers switch near infinitely fast, but we also are mostly limited by the strength of whatever the coils can handle and the time they can impart force to the object. I think it becomes more of a mechanical design challenge than anything, because now we need a better design than just a linear rail: something like a looping track in the rear so it can pre-accelerate it making several loops before entering the straight track. Or we may just end up using a spinning sling arm etc.
Also I didn't list other unfortunate physics effects such as Eddy currents and effects of fast moving magnets next to coils which is really more of a materials engineer design problem.
You would post this perfectly in time to finish up the Colin furze solar-mowar 😅 cheers for the share!
ikr
except this is not nearly powerful enough to be dangerous, not for Colin at least. [0.015kg*(11m/s)^] /2 = 0.9J muzzle energy. Some permitted airsoft guns go over 5J.
i'm already a regular microcenter customer. they're like 1.8 miles away from my house. lifesaver.
With your high-speed camera you could also try tuning the on-times for the coils - as the payload travels down the track and accelerates it's likely that the coils should be triggered a little earlier for maximum acceleration (because it takes a finite time for the current to build in each coil).
This. Also, another very cool improvement would be printing a holding mechanism in the 'sled' to hold the plane. By the end of the rail the tail is pushed down and the angle of attack is way too high.
The holding mechanism could be something like a moving piece inside the sled that holds the plane like a thumb and hits another 3d printed part at the end of the rail, moving it to the release position.
The IR drop would also increase, changing pulse timing the further along the gun the coils get, assuming all from common end source.
Wow a beautiful implementation.. The big difficulty with railguns is that as you get faster and longer the timing of the pulses gets shorter and more critical. This also means the power loading each coil unit needs gets higher too..(since it gets divided by time) Above a certain speed things start to get exotic very quickly. (over about 80 - 300 m/s)
My own work was only crude and theoretical - with the idea of using a railgun like a rocket engine. Not really a good idea for about a dozen reasons, the killer being that it just took too much energy to ever work. Another being that the rocket 'exhaust' was rather dangerous.
Thanks for the information. I'm going to go make a railgun now.
If you fire off in a vacuum, you essentially have the makings of a magnetic space launcher, either on Earth or in space, even space craft launcher , or a hyper loop transporter system!
Or in this case a aircraft launcher, or a coil or rail gun!
5:43 Interesting strobing effect. Three questions Tom: What is the frequency of your mains electricity? What was the frame rate of your high speed video footage? What was the shutter speed?
Mains frequency here is 50hz, I think the camera frame rate was 1519fps in that shot. Not sure on the shutter speed, but probably 1/3038.
I love how you caught the paper airplane going twice as fast as a fly in your slow mo
Ещё желательно мосфеты расположить максимально близко от датчиков холла до катушек, чтоб снизить временное влияние. А так же попробовать с разными расстояниями этих датчиков холла до управляемой катушки, чтобы добиться максимального эффекта.
U.S. Navy: "We build railguns to shoot bad guys."
Tom Stanton: "We build railguns to launch household items. For our amusement and all the giggles." 😁😁😁
Its not quite a railgun though, its a coil gun. Coilguns magnetically attract the projectile, and railguns repel the projectile via the Lorentz force (which requires current running though the projectile)
If Tom's next version can launch a 3kg slug to Mach 10 then it's going to be an amazing video.
Really cool! I could hang out at Micro Center all day.
i've always wondered if you could build one of these in a loop where one portion of the loop had a falling gate that allowed you to fire it once up to the speed you want. and if that would allow you to get much faster speeds in a much smaller package?
What you're describing is exactly how early particle accelerators worked. It's a great idea.
Could get stuff stupid fast. But a nightmare to build
Love the videos, Tom. You've reignited my interest for engineering!
Machine learning pid power full iron stationruclips.net/video/KXNRlJPkWIkc/видео.htmlsed24nmb
absolutely beautiful slow motion shots
finally, my day is made.
Bro🥲
Machine learning pid power full iron stationruclips.net/video/KXNRlJPkWIkc/видео.htmlsed24
Hey Mr Stanton,
how about instead of turning the coils off you switch the polarity. You should get an increase of acceleration of at least 60% :D (But my math is very basic and probably wont cut it). Please try, would love to see another video about this :D
He said near the end of the video that he was planning to do this. Should be good, looking forward to it :D
@@alomega01 must have overheard it :D thanks for the hint
Me 25 years ago taking things apart and finding magnetic actuator thinking that I could build a rail launcher out of it 😂
Now build a smaller version on a rc plane carrier, that can launch a mini rc plane ✈️. That would be amazing I think, but great video loved it!
Nice 👌🏻 You know what Tom, try magnetically levitating that 15gm weight so friction is even less and also make it aerodynamic like bullet shape or something to achieve even more velocity and efficiency!
Aluminum will repel a high velocity magnetic field.
no need to levitate magnetically, a simple air cushion rail will do just fine, ie pneumatic linear bearing.
@@ikocheratcr yup that'd work too nice idea !
that first coaster in the video is called the skyrocket! its in a park called Kennywood outside of Pittsburgh. great ride and kinda neat to see in a video!
Great project, Tom!
I was also thinking; You can make it 2 times faster by not only switching on/off but also reversing the current direction. To do that, replace each MOSFET with an H-Bridge.
5:01 Tom.exe is not responding.
you explaining it makes it seem so easy
Awesome! Theoretically, could you also add opposing magnets on the sled, and guide rail to reduce friction (basically a floating sled)?
Maglev sled!!!
That already exits in japan;D
Would that not interfere with the magnetic field of the coil???
Could you make a sled that holds the airplane in place until it reaches the end of the slot, and then releases it? It would get more speed from the sled, and then shoot up after release. I'm trying to imagine how you do that without introducing too much friction or weight to the sled, though...
Imagine a sled Mechanically designed that it can only release the plane if the planes goes forward. Put a stopper at the end of the rail, as soon as the sled touches it, the plane is launched since it's now free to go past the sled. It could be a pinching mechanism in tension by a spring, the pinching part on top and the releasing mechanism on the bottom. At the end of the rail, there's two reducing rails that apply abruptly pressure on releasing levers at the same time and the plane is released. 😅
The wing would crumble if the speed is too high.
Very low pressure pen refill spring with a stopper to release at dead-end.
It may launch earlier incase lift exceeds spring hold force without hindrance to either of 3 mechanism.
Instantly one of my favorite posts on yt. Thanks Tom.
I think you could see a performance increase by getting rid of the control board and instead using open-drain sensors. That way, the sensor can turn on the electromagnet by itself. It might speed things up and let you tighten the timing margins.
Also, the idea to reverse the current through the coils is interesting. Are the coils low-inductance enough to make that feasible?
hi buddy i just wanted to ask something is it fine to use a battery source rather than supercapacitors
@@shreyasshenoy7215 you are going to need the extreme amounts of current that the super capacitors can output, Batteries cant match a capacitors ability to discharge and charge quickly.
Optimally you reverse the electric polarity to repulse the sled magnate so you push as well as pull.
Edit, I suggested it before he mentioned it at the end.
u smug rn?
FUCK YEAH! BASIC SCIENCE!
Jumped the rail gun.
0:15 HECK YEAH! KENNYWOOD! Also the MOSFET is also know as a transistor and is opened a small amount of amperage, not voltage. Although, voltage is still needed to push amperage through a wire.
The part that begins at 5:52 is the best!
Tom, that was so far the most interesting video I have watched on you channel. I would love to see a real Concorde (of course a model) starting from your rails!
THANK YOU FOR USING MY HOME PARK LAUNCH COASTER AS THE EXAMPLE. FINALLY KENNYWOOD IS REALIZED!
Hey Tom, love the channel and all the crazy stuff you make. I have a couple of ideas for you to improve your device. 1st, make sure you have freewheeling diodes across your coils so that you don't overvoltage your FETs during turn off. You may already have them in there and I just didn't see them. 2nd, you should devise a way to check the timing of your coils relative to the position of the sled. As the sled accelerates, the turn on of subsequent coils need to be advanced in time compared to the upstream end. I would not be surprised if the sled is being slowed down near the end of the track. You could use LEDs on each coil and high speed photography to visualize the timing. I suspect you may need to move the hall probes upstream as you go down the track to give the same timing advance as you had near the beginning. This brings me to #3, you may want to explore spacing the coils farther apart as you go down the track. This will also alter the timing of the system. Again, cool builds and I look forward to seeing more.
good pointers. 2nd could be tested fast with turning of like 10% of the downstream coils and measure the speed.