Final(ish) Thoughts on the Coilgun Project
Vložit
- čas přidán 2. 07. 2024
- I am unfortunately ending the coilgun project, but only for now. When I have more experience and better equipment, I will return to it. For now though, here are my final thoughts on the project. I now believe that capacitors should be used with coils of specified inductance, along with copper projectiles propelled by induced eddy current.
Gain access to CAD files and support my work - / levijanssen - Věda a technologie
Things have changed! I have been anonymously donated $1000 dollars to go towards an oscilloscope! Once I do get a nice scope, the coilgun will become feasible. I will though, be taking my time with the project and will begin by developing a test setup. I'll have to make a follow-up video showing off the scope and explaining further.
That’s awesome! You should set up a Patreon, I’m sure a handful of people would be willing to subscribe to help with stuff like this
www.patreon.com/levijanssen ;)
YAY! i was just about to write about how it saddens me to end the project, but now that goes out the window! get all the stuff you need and when you have, we will be here. :)
Dude all of this is amazing, looking forward for the next update
"anonymously" seems like darpa decided to subsidezed your proyect xd.
You really don't need a fast or digital oscilloscope for the rather slow waves of a coilgun, though a DSO could make data collection easier and more precise. An old 10 MHz analog scope will do. That is what I used with my single-stage 4000 gee Mass Driver made from junkbox parts. It heaved a 60-gram soft steel rod of 11-mm diameter and 38 mm length through an old surplus coil (also 38 mm long and 7/16-inch ID) from a pinball machine. That coil was designed to run on 24 VAC and I slammed it with 715 VDC, pushing 50 kW through the several hundred turns of 19-gauge wire for about 1 millisecond as the steel rod soaked up inductive and kinetic energy. All attempts to use a permanent magnet as projectile simply re-magnetized the magnet with the opposite polarity. The magnetic field was very intense! Power switching was via a simple relay also meant for 24 VAC, surplus from a vending machine.
For capacitors I used 2 computer-grade caps in series that I had laying around. Superficially, they seemed to be wildly mismatched, as one was rated 200 VDC and 1400 microfarads, while the other was rated 550 microfarads and 525 volts. But in series, they split the charge beautifully and each was run very close to rated limits. Total (computed) energy storage was 104 Joules. Most of that was discharged in the first millisecond when the Fire! button closed the relay and the steel rod leapt 18 mm to the center of the coil with tremendous force.
The voltage then dropped to a slightly-tilted plateau around 300 volts for about 2 ms with damped ringing at about 200 kHz for 1 millisecond. Then the voltage plummeted to about 50 volts for several more milliseconds. There was never any reverse polarity across the pair of caps, though one of the pair did get a few volts of inverse charge because they were not perfectly matched. Charging circuit was an old filament transformer from a tube-type radio with the secondary rated 440 VAC at 40 milliamps, which was half-wave rectified by a series pair of 1500-volt diodes to make 715 volts. (Clearly that transformer put out about 510 volts open-circuit on the secondary.)
Full charge was indicated by several miniature neon bulbs in series plus a series resistor to limit bulb current to near the minimum for stable glow. Combined with the filament winding, the transformer was rated about 40 watts. To prevent transformer damage during the initial phase of (re)charging the caps, the transformer primary was put in series with a 40-watt incandescent bulb, which made a near-perfect current limiter. With that recharge limit, I could fire my Mass Driver about 3 times per minute all day long, with the coil never getting more than barely noticeably warm, so about 1 or 2 degrees over ambient temperature.
No attempt was made to design or select parts for minimum losses or to electrically match up well, except for the capacitors and the diodes. Instead, the pre-firing position of the steel rod was experimentally determined for best speed, about 31 meters per second, varying slightly between shots as positioning was not perfectly repeatable for every shot. Kinetic energy actually delivered to the steel rod was thus about 29 Joules out of 104 stored, so efficiency on that basis was about 28%. Wallplug efficiency would be a bit lower, but was not measured. Total cost of the whole Mass Driver was 50 cents US, for the 2 diodes!
Came for the cycloidal gearboxes, stayed for the clever little Digi-Key ruler stand and coilgun. Great video, BTW!
It's all good man. I'm still here for everything else you'll put out
Hi buddy, forget thinking about bullet position, just use the iron rod as firing pin to push a regular lead bullet through the barrel, the iron rod will get stick in the middle but the regular lead bullet will be fired out from the cannon. Your principal goal is to get the strongest coil electromagnet ever. Good luck buddy.
Hi, great series of videos. Keep at it and you will get there. Some comments. You are mixing up dumping the energy stored in a capacitor, and simply using it to make a low impedance voltage source and then switching with a mosfet. Split your project into sections and then solve each section one at a time. For example, start with a bench power supply with effectively infinite energy (for your design) and solve the coil, arrangement and switching issues first. Then move onto working out how you will make the power supply portable. Note that there was a super useful comment from a previous video about diode protection with mosfets. Remember mosfets blow short circuit! Some of my thoughts: think of each coil segment as a tuned circuit. Add a capacitor to resonate with the coil. Your idea of push-pull sounds promising. Consider using a split power supply (positive and negative) and use P-type and N-type mosfets to switch the +ve and -ve rails. Also an H bridge could be considered. All coils before the projectile are pushing. All coils after are pulling. The coil where the projectile actually is, is switched off. Your comments about efficiency are super useful. But to calculate the energy used, you need to get the area under the current curve (assuming voltage stays the same). Simple enough to do with a small micro and an ADC. As for monitoring the projectile position to time the mosfet switching, I suspect LED and LDRs type approach is going to be too slow. As you have shown, the projectile changes the current through the coils as it moves through. I would think about using that to monitor the position. I also like your 1/x ideas. You could achieve something similar by switching more current though the end coils compared with those in the beginning. I'm not a big fan of the suck and see approach. Try modelling the system in Matlab (or some such progem) first to get your system parameters. Remember with all engineering designs KISS. The full system model will be quite complicated, but a simplified model in Matlab will give you a lot of insight and allow you to start to optimisation process. Good luck - and as I say keep at it. You will get there.
Good analysis. Some thoughts, if you're still thinking about this project...
** I'm glad to see you turned the MOSFET around so the tab faces outward, for screwing onto a heat sink. Does your MOSFET have drain connected to the tab (usually the middle pin)? If so, couldn't you just connect all the FET drains to the same ground+heat sink for a double win?
** You mentioned paper as an electrical insulator between the PCB and aluminum plate anyway. Mica's a far better electrical insulator, and a better thermal conductor than most silicone rubber. It's even available precut for TO-220 packages if you do need that.
** As others suggested, wrap or glue the windings so they're held in place when firing.
** I was going to offer a reminder about flyback diodes, but you already covered that in Part 5. :-)
** If you're looking for highest possible efficiency, be aware the outer turns on the coil are not contributing as much magnetic field intensity to the projectile as the inner turns are, because of flux leakage. A "square" cross section (for example, 5 windings long and 5 windings deep) is already going to be lossy. If the coil is 3cm long, having 1cm of windings is probably close to optimal. But if you were going for velocity or small package, and packing 15 coils together, then do make the cross section square, because it's still a (smaller) net win. Then choose a voltage necessary to get the current your FET can handle.
** Optical detectors don't need to "see" their full width, so you don't need big gaps between coils for light. The LED and detector can go outside the coils, looking through a tiny gap. Even a paper-thin crack between the coils is enough to see light through and switch them, especially if the detector is a phototransistor or feeds the base of a BJT, like Mahdi's circuit. Just surround the light path with black tape or something so stray light doesn't get in. You can also buy photodetectors with lenses built in, which makes them much more sensitive in the direction you care about. Or if you want to mass produce these, as IR MadCow suggested, put detector and LED side by side but send the light to the other side of the coil using optical fiber. If it's plastic fiber, melt the end of the fiber slightly so it's perpendicular or even partly spherical to help couple light straight in or out with less scattering/reflection.
** Instead of optics, as Rene .Lem suggested, you might be able to use the voltage/current relationship in the previous coil to determine when the projectile is just entering the new one, because the inductance of the coil will be different depending on whether it's wrapped around an air core or iron core (i.e. projectile). That'll require fancy electronics, though. But on the other hand, it's a great excuse to really put your new sampling oscilloscope through its paces! :-)
** Cut the slot all the way through one end to reduce eddy currents in the plate. It will also reduce the induced electric field at the gap that the wires pass through, so there's less chance of an arc. That also gives you a path for your light and LED/detector wires.
** Don't use threaded rod. Murphy guarantees the coil or FET in the middle of your stack will fail first, and then you'll have to unscrew all the rods all that distance to swap out that one module. Just put nuts on the ends and squeeze 'em tight. Although I guess that could make alignment with optical holes a bit tricky.
** You mentioned replacing the coils with linear induction drive. But depending on how you implement this, you might end up with projectiles that hug and drag along one wall or the other, sapping kinetic energy. Coilguns are inefficient, but one of the reasons they're simple is because they tend to self-center the projectile. The magnetic field lines as the projectile enters the coil are concentrated toward the middle, which tends to pull the projectile to the center of the tube, away from the walls. With linear induction, you would have separate north-south poles along the walls, which is inherently unstable if the projectile isn't held on a rail.
Great comments from many others here as well: Azagro, Rene .Lem, many others...
Good luck, and looking forward to your next update!
I have spent some time over the holidays on this topic. I am an electronics engineer with a university degree and just liked the idea of building a BB coilgun. There are several problems with that, for ones the BB projectile is very small, but I like the idea of using cheap standard projectiles. From what I see in most other projects, they are just building it by trial and fail. Not even bother to do some calculations. But if you do and understand physics then you see also the advantage of a small BB projectile. There is a limit of the magnetic force you can apply to it so if you design it properly it should become a very energy efficient coil gun.
Some design considerations I have made so far.
- The length of the coil should not exceed the diameter of the BB (I plan to use 6mm BB) much.This is why you latest design was an eye catcher for me.
- No optics to trigger the current in the stages. All will be timed by a micro controller. And yes you will need an oscilloscope to fine tune the timing. As there are a lot of variables like friction, air resistance etc. that would be hard to calculate.
- I plan to use capacitors or better capacitor banks. But the issue I see in most design is that they use the same coil and capacitor for each stage. But as the projectile accelerates with each stage, the time it spends in the coils magnetic field is shorter. So the current must be applied for less time than the previous stage or it will slow down the projectile. The simplest way to do that is to size the capacity and thereby the time the current can flow in the coil. The current should be 0 or close to if the projectile is in the middle of the coil, otherwise it will drag it back. You could also try to shut the current somehow in the right moment, but I guess that would be a very difficult problem to solve.
I assume if done properly it should be possible to build a BB gun that can shot one BB after the other without having to wait seconds to load the capacitors between the shoots.
The only thing I am currently struggling with is how to load the bb into the barrel. I have see a design with a servo(too slow). Maybe I will use a solenoid like push system.
The other problem is gravity, how to prevent the bb rolling of out the barrel, but not having a lot of friction.
Some thoughts on the wave idea
This might work if you have a magnetic projectile that has polarity, but I doubt it will work on a ferromagnetic projectile. Maybe best to explain is why a AC solenoid works. I guess your thought is that if you apply the negative portion of the wave in the right time, you would accelerate the projectile further. But the opposite will happen. It will drag the projectile back in the coil and you loose velocity.
Simplified if the current will change direction, the magnetic field does as well and change polarity. But for the projectile the polarity does not matter unless it is a magnet by itself.
BTW once you have your oscilloscope you will see that there is a negative part of the wave that is caused by the coil that wants to continue the flow.
The only thing I am currently struggling with is how to load the bb into the barrel. I have see a design with a servo(too slow). Maybe I will use a solenoid like push system.
The other problem is gravity, how to prevent the bb rolling of out the barrel, but not having a lot of friction.
how about a revovering like drum feed, the drum only lines up when when in use and the first coil is on
But the issue I see in most design is that they use the same coil and capacitor for each stage. But as the projectile accelerates with each stage, the time it spends in the coils magnetic field is shorter. So the current must be applied for less time than the previous stage or it will slow down the projectile.
how about changing the length of each coil so the time it spends in the coils magnetic field is the same
To help better time the stages of the various coils, why not dope or use part of a Thorium Welding Rod (very safe, low cost and easy to purchase) for the projectile. The low cost and safe radioactivity could be used to trigger on/off the coils.
As for loading the BBs, check out the old West German G-11 caseless assault rifle rotating bolt design. This weapon was designed in the 1980s but never made it to full production.
To make the later stages give a faster pulse isn't it ideal to use less inductance rather than capacitance, so you have the same amount of energy but delivered quicker
You could use the magnetic saturation in the ferromagnetic projectile to push it out of the 1st coil when it passes the middle point making use of the negative cycle of the sinewave. Using the negative cycle will increase efficiency and decrease the magnetic saturation due to the change in polarity, thus preparing the projectile to be even more attracted in the next coil stage. With fine tuning it may even saturate the other way around when leaving the 1st coil so it will have more pulling force from the 2nd stage. Like wipping a rope moving the hand in both directions. But I think this level of synchronism would be hard to pull and very suceptible to temperature changes.
Hey man love the work you’ve done! Seriously it’s great work and I’m happy to say I’ve learned a lot from watching your videos. I’m also glad that you’re sticking to your goal of designing the most efficient coil gun as well as knowing where to take a break from a project. I do hope you return to this project and accomplish your goal. Also I do hope that you do eventually build you’re previous design since you did get the parts and it would really be cool to see one of your designs come to the real world even if it may not the most efficient at this time. I still think it would be awesome! Any way I can’t wait to see what you come up with next! See ya!
Thank you for this series even if you are abandoning it for now. It's definitely given me some ideas. Focusing on efficiency is the best way to make it more powerful aside from strapping more capacitors to it.
2 years later, still waiting for the next coilgun episode...
This last video brought a lot of valid questions to mind. In consideration of the copper/aluminum/silver projectiles; they probably only accelerate until they reach saturation, and I'm not sure how much magnetic push they get until that point arrives and their greatest acceleration would be when they go from being fully saturated with the field going one direction, then the field switching and going the other direction. Skin effect also plays into this and using a solid copper slug would mean it is unlikely to become fully saturated in any individual coil so the extra mass is dead weight. I think it would probably be impractical to do it with a copper projectile, however reversing the field of every other coil would most likely still give it acceleration through the entire length of coils.
Another thought about copper/aluminum/silver projectiles. Unlike iron projectiles which are pulled to the center of the field, the copper/aluminum/silver projectiles are pushed from the center of the field. That means timings will be wholly different and you will have to give the projectile an initial push or start the projectile from slightly off center exiting from the first coil.
Hey Levi! I went through learning exactly what you’re talking about in this awesome video, last year (trying my hand at a “better” coil gun) I had a slightly different approach but drew a lot of the same conclusions. I was more focused on the power source; I was trying to harness the eddy currents created by permanent magnets as they pass through a copper tube, or in my case, a series of copper rings.
I’d love to share some of my designs that I’ve drawn, 3D printed and fabricated with you. Let me know if you’re interested.
The projectile could be powered like railgun and the electrolyte capacitors could be used in tandem with H bridge that switches it's leads when current goes through 0.
One idea that I have had about the number of coils, is that adding more turns takes more length of wire due to the larger diameter, reducing the efficiency of adding more coils. At lower coil counts, it takes more current to get the same magnetic field strength, which increases the power usage. I used an online calculator to find the magnetic field strength, and I believe the optimal coil count is around 7 layers when using a 10mm barrel. There is a small improvement at 14 coils, but it’s only about 15% for a much higher cost.
A 4-channel 250MSPS oscilloscope from Rigol etc. is quite affordable, FYI (~300 monies). You don't need GSPS sample-rates for a coil-gun, or most other electronics applications.
I actually feel engaged watching you talk.
Could you make a coilgun with overlapping coils? Like once the first coil reaches the middle, theres another coil on top, to provide constant power with no breaks, and underneath the overlapping coil, theres another coil underneath it and so on. and the coils wouldnt be winded one by one, because that would make the overlapping coil stick out way from the barrel, instead you wind one coil on the bottom coil, then wind one layer on the top coil, and repeat so that all the coil layers are sandwiched together.
Noooooooo. Just discovered this series and I was sooooo sad to end up here. We need part 7!
Good job learning!
Oh snap, did I just find myself a new favorite CZcamsr? Yes, yes I did.
I'm wondering if eddy currents generated in the projectile will be a significant factor in the forces acting on the projectile. Specifically, if currents induced in the projectile (or even other parts of the coil gun assembly such as aluminum backing plates) will resist the forward acceleration of the projectile. Based on my admittedly vague understanding of electromagnetism I think there must be some sort of parasitic losses along these lines, but I don't know how significant they are in comparison to the force accelerating the projectile. Good luck on the project! Great to see someone putting thought into it, rather than just trying to dump as much power as possible. Subscribed to see your progress.
I have some experience with pulsed power systems, so let me give you a comment on why coil guns actually use capacitors.
High voltage capacitors have (reasonably low) capacitance and low ESR. I'm saying something like 1uF, 3-10kV (you can buy something like that for microwaves). You should be able to discharge a pulse with powers of the order of 1-100MW if you properly design it. Combine this with a low inductance inductor and you'll have a system that can provide peak forces of order 100-1000N with relative ease. If you don't use capacitors, the parasitic capacitances of the system (mostly battery) start limiting your time constants. That's why research-grade coil guns still use capacitors, they're the way to go.
The reason researchers rather use high voltages and high currents can be explained in the equations (I recommend you do it on paper so you can see): Coil force F is proportional to I^2. Say you have a pulse (assume a square pulse for simplicity) of "dt" long. The velocity you add to the projectile is dv=a*dt; which through Newton's second law is dv=F*dt/m. Since F~I^2, you get that dv~I^2*dt/m. But the pulse energy is proportional to I*dt (E=I*V*dt), meaning the shorter the pulse (the smaller dt), the higher I is for the same energy. So the shorter the dt, the higher I, but you get the benefit of I^2 for velocity increase; which works in your favor, efficiency-wise. In summary, you always want to minimize dt and maximize I. The easiest way to accomplish that in real life is to have the lowest capacitance possible since time constants (of both RC and RLC circuits) are proportional to C. The high voltage is a by-product of that effect to keep the energy (1/2CV^2) the same. My mental picture of this is like if each coil is like a hammer that "dings" the projectile, imparting its energy on it to accelerate the slug.
Keep studying. You're really young and will accomplish great things if you just keep at it! I recommend you look into the academic literature if you're interested in seeing what serious people do with coil guns (use Scihub if you can't access the papers). Independent thought is important, but as you mentioned in one of your videos, you're not the first one to come up with these ideas. Reading brings you to the "shoulders of giants", so when you think and look ahead you can see things that no one has seen before!
I don't think eddy current go deep in copper with the constant changing of the mag force. So copper pipe with a steel core would be cheep , and by heating the copper and freezing the steel you'd get a tight enough shrink fit. Oh, and the manning up about recognizing your limits, welcome to Adulthood , bud.
I’d be worried that the inductor used as a projectile would encounter the same saturation limits due to its own resistance and inductance, could be significantly higher than magnetic saturation, worth testing! Lovely to see you got a scope :)
You do it at your own pace.
thank you for your work with this.
Don't use a copper projectile. Use aluminum. It's got a bit higher resistance, but similar properties and it's quite a bit cheaper.
A suggestion for the projectile:
You could use a glas fibre arrow with copper rings on it.
If you want a steel arowhead, you could place it outside the gun.
You're getting a lot closer to the truth of things. One thing to ask yourself is why have multiple capacitors versus one for all coils. Then you simply won't have to worry about RLC circuits anymore (except last stages) but only inductor design. I can tell you that efficiency is attainable but only at very low power levels which makes it not worth designing for because you end up with less muzzle energy and you were never constrained by efficiency anyways. Induction is in theory a great idea but that's a whole other bag of questions which I'm currently working on. I do currently believe you may get higher efficiency (without power ceiling sacrifice) but it comes down to the size of eddy currents you get at sort of coil timescales and coupling. It's totally up in the air as of now as far as I can tell
God damnit somebody must have an oscilloscope laying around, get this man his Oscilloscope
Consider a oscilloscope module that uses a PC for viewing the data collected. Purchasing half the system.
I like how you talked yourself into using Lorentz force… I mean it took you a while, but that’s ok. 😂😂😂 Anyway, keep it up kid- you’re going places!
A few points, if you want to get a sinusoidal/forward and reverse current through the inductor, without charging/discharging the capacitor in reverse, you could use a full bridge driver. In this case the waveform would be "artificial" rather than the natural oscillation, because you'd be controlling it with the full bridge. I think it would make tuning it much easier if you can change the waveform and period etc. from software rather than having to rebuild the hardware for any changes. I fully agree that a single period sine wave, with the period being exactly how long the projectile is within the influence of the coil would be perfect, but this would need different capacitor/coil arrangements for each stage since the projectile speeds up with each stage and spends less and less time in each subsequent stage. Fine tuning this would require building and rebuilding each coil every time you test in order to fine tune the time constant.
Secondly, you can push a ferromagnetic (non permanent magnet) projectile in a coilgun. Once the projectile is magnetized by a coil, and then leaves the end of the coil, you can flip the polarity of the coil and it will give the projectile a little push. This is because the projectile is magnetized, and doesn't immediately demagnetize when the coil is switched off, or changes polarity, so for a brief moment it acts like a permanent magnet projectile would. Of course if you keep the coil on in the reverse polarity for too long the projectile will demagnetize, remagnetize in the new orientation and get sucked back. Changing the polarity of the coil, and demagnetizing the projectile as it leaves the coil with the little reverse polarity push also boosts your efficiency because you're making use of the magnetism left over in the projectile, and desaturating the projectile for the next coil stage, which is good because you lose a lot of efficiency when dealing with over saturated projectiles.
Also, glad you're looking at induction coilguns, I've heard they can be much more efficient due to the lack of saturation, but I've never seen any DIY implementations.
Well said, but do realize how expensive the components to fullbridge hardswitch the massive currents at very high speeds to create an arbitrary waveform would be. For reference, the (varying topology) circuits used in QCWDRSSTC’s bus mod perform this, and are lower voltages and currents than a coil gun would be (not to mention their size), plus you’d need one of these for ever stage.
@@Ferraday absolutely fair point, I just mentioned these as he's been stressing that he's aiming for efficiency above all else.
I would say that there is no real point in switching to the opposite polarity (making it a waveform) because for timing purposes, you want the projectile to physically be closely approaching the coil by the time you reach peak field intensity. Assuming that you actually succeeded in accelerating the projectile, even if you had significant induced magnetism (which you could do before ever firing with for ferrous projectiles, if you wanted to go that way), by the time you got down to zero on the magnetic field the projectile (which is moving away faster than it approached) would be past the physical point that you could be having a significant effect on the now magnetized projectile, AND THEN you have to start forming a new opposite field. The inductance of the coil will likely not allow you to EFFIECIENTLY switch rapidly enough the polarity to be worth it. Not to mention, depending on your coil spacing, you could be influencing your next coil's magnetic field that you would have to take into account in your simulations. It is possible that, assuming you are still worried about the direction of the induced magnetic field in the projectile, that you could actually be inducing a current against the desired power flow in the next coil.
Late to the party but was thinking,
- What if you were to use a NEMA 'C' or 'D' type motor in a linear configuration and provide power with a VFD that has a high locked-rotor output as your power supply.
- You can modify the 'zero point' of a sinusoidal power supply by connecting an additional DC power supply.
You really should finish the project. Yes, it is not optimal, but 1) you could still learn a lot from the build. 2) you are falling into analysis paralysis, which is not a good place to be. Make this and understand it is a iterative process. You don't need perfection the first time up. 3) you should deliver on the project, unless a major issue develops. Not getting optimal results is not a major issue in a iterative process. Think SpaceX.
Also, dont give up on a concept before prototyping it, you cant know for certain how efficient or not it will be without testing it.
i'm not done yet, but don't you need MOVEMENT through the magnetic field with the copper projectile to induce the eddy currents? that would imply you'd need to inject the copper projectile into the system much like a railgun. just a thought, curious if i'm missing something.
There's some smartphone-sized portable oscilloscopes on ebay and amazon for around $100. They're not perfect, but definitely affordable compared to the $1.5k professional scopes.
Hi, I liked the idea on your 3rd video about coilgun, personally I would have reversed the current and direction of the projectile so that the maximum force would be at the beginning and accelerated could escape the high magnetic filed, but still having acceleration till the end.
at the same time I do not see why you do not use a ferromagnetic pipe which could increase the magnetism and at the same time increasing efficiency of the overall (motors that you mention use a ferromagnetic center to increase efficiency).
lets say you power your coilgun with batteries only, you somehow manage to pull constant current from your battery (hard to do, but maximum power draw) and you have 100% energy transfer to the projectile (basically impossible). with these assumptions, you can calculate acceleration/final velocity for a given barrel length. at reasonable barrel lengths, with reasonable projectiles, getting anything faster than a paintball gun requires many kW of power. you NEED capacitors, unless you want a really big (EV sized) battery. Your next limiting factor with capacitors is the mosfets themselves. they too have a power rating (VxI) and again, you need some that can do many kW.
Don't give up.
If you were using ferromagnetic materials as projectiles you would only need that positive part of the sinewave, a magnet could do the same and I think be twice as efficient, but if you used a full period then you could double the efficiency again.
Use many toroidal coils which increase with size and are shielded between each other, but all have same size middle part.
while watching i got an intresting thought. Same as induction motors have rather weak torque on start the bullet will have low speed at start. so you would need extremely long barrel to get more efficiency and it defeats the purpose of a gun. it would need to be more like a looped particle accelerator
"Is there another option"
Why not just have the magnet be a permanent part of the coilgun, and have it travel the length of the "barrel", pushing the projectile down the bore, and then having a spring or something reset it to it's starting point once the force is removed? I know absolutely nothing about electronics, but aside from the innate lowering of efficiency in that the gun now has to overcome the weight of the magnet and the push-back from the spring in addition to the load from the projectile, is there any major downside with this I'm somehow missing? (Pardon me if you address any ideas like this later, I'm typing this mid-video.)
if i am not wrong the trick is not remove the capacitors, it is just design the coils ( in theory ). Then i don't know what are your calculations and only have an idea of what you want to get try put diode on the circuit to avoid the negative effect of the discharge of the capacitor remember it is not a rail gun that can be another good project.
i have messed around with some coils on a much smaller scale and i found striking the projectile so it was already moving before the coil got energised let the 6mm steel bb i was using pick up allot more speed than just using the coil. my theory as to why it worked better than the coil alone is that in the time it had before the capacitor was discharged the bb was already moving so the distance travelled in a active field was greater and allowing it to pick up more momentum.
Come to think of it if you had a set of spinning rollers before the first stage to get it moving you could have them twisted slightly to give the projectile spin as well as forwards momentum. Not sure if the spin would improve efficiency but it would make it more accurate.
have you considered a coil gun for the acceleration force as well as a rail circuit in the projectile in order to induce a magnetic field? could work instead of flipping the field with the RCL transfer function (or could add more efficiency)
Also balance the energy storage of the capacitor and the inductor.
Knock on the door from the men in suites ?
I load copper bullets for my rifle, maybe I can use those as the projectile. Thank you for your insight.
How close are you to restarting the project? I’m super excited to see it continue, and I was also wondering how having a pre acceleration phase (ie some gas pneumatic system) would affect the coils and their pulses. Basically I’m trying to say that if it has an initial velocity could it become more efficient by removing the need for those beginning driver coils and allow for cheaper parts because of a smaller pulse window. I’m just rambling though lol.
But with eddy currents you will have a ball of molten copper shooting out the end. Also it is not constant the heat generation will be significant.
My problem with coilgun is it's acceleration. It's low compared to firearms. To compensated you would need a very long barrel.
I wonder if a ball shaped projectile can be sped up in a loop then released (like a particle accelerator loop) . Technicly it would mean infinit lenght for acceleration.
It would work with DC current from a battery.
Was thinking the same thing
The problem to solve with this concept is how to switch from traveling very fast through a toroidal barrel into a straight one. Also, you can't really use constant DC for this purpose, it has to be pulsed on and off somehow or else the force experienced as the projectile enters the coil will cancel out with the force experienced as the projectile leaves the coil.
I've already come up with a design idea a month ago that could hypothetically pump up the efficiency over +30% after I did my calculations and the simulations for it but unfortunately I need a 3d printer to print the parts that I need since they require some accurately calculated math to manufacture correctly which is something that's very expensive for me in my 3rd world country. Fml.
That sounds promising, id love to see your design
Also look at energy recovery as projectile passes.
How to avoid negative voltage? cant you play with offset? set your -v to gnd of the battery, and your 0v to half(?) the battery voltage
if im not mistaken I believe the navy uses aluminum projectiles
A coil gun that shoots coils. Woah!
You can get a cheap Hantek scope on amazon for 150-200$ it'll do the job they're quite good especially since a coilgun pulse is in the kilohertz range. They might have come up in price a little for now because of the supply shortages from covid but it's not like they're unattainably expensive. Christmas is coming up perfect time lol
Someone get this man an oscilloscope
Correct me if I am wrong but I think the idea to use copper as a projectile isnt going to work as the magnetic field is parallel with the shooting Direction and by my understanding of the lorentz force the force applied to the projectile should be 90 degree to the wanted moving direction, at least with the "standart" coil layout.
Not so sure about my comment.
I will try to figure it out when I have got time.
will just metal projectile (not a magnet)
really repel from coil ?
(when you reverse-think that should be full period, using magnet)
sooo maybe you need just a pulling half ?
ok. you already know)
You can build very large variable capacitors.
Doesn't the negative portion of the Amp vs Time get clipped since you are using a flyback diode?
How did you come to learn all of these concepts? Was it through school, your projects, or perhaps, general self study? Nevertheless, this is quite an interesting video. Can't wait for your next upload!
I learned this stuff in college level physics. I would expect it to be expanded upon in electrical engineering courses
@@dadsfriendlyrobotcompany That makes sense. Thanks.
I am curious about your findings with an eddy-current projectile. Personally I do not believe in it, because eddy-current can be used to brake, but not to accelerate.
But maybe the dynamic behaviour of an accelerated projectile together with a certain time-constant of the eddy-current, could change my mind...?
essentially what you are talking about is no longer a coil gun but like the thinking anyways ,to go back to the earlier attempt with multiple coils and looking for a way to time them a simpler solution is to make the barrel as a linear set of commutation rings which are triggered by contact wings on the projectile these wouldn't need to carry the full load just act to switch relays between individual capacitor and coil circuits directly ahead of the projectile (a poor mans hall sensor essentially)
in order to gain speed along the path each coil has more windings because it is momentary only you could do away with the relays and capacitors and just use the battery
Talk about energy storage in a capacitor and energy storage in an inductor.
How would an aluminum projectile our work
I wonder if you can use an inductor to power a coil stage rather than a capacitor... you might get a stronger current pulse.
Effectively, you would discharge a capacitor into a large inductor (this inductor doesn't need to be optimized for field, just inductance). Then you switch the capacitor out of the circuit and connect the inductor to the coil and a shunt resistor to dissipate the energy. The collapse of the high inductance magnetic field will create a large derivative in the current (i.e. the inductor flyback).
I think you have magnetic saturation incorrect. The magnetic moment generated by the projectile will saturate, you can only align the atoms until they are all aligned. However, you can still increase the external magnetic field. The force comes from the interaction of the external magnetic field and the magnetic moment of your projectile. F~B*m. If m saturates, you can still increase F when you increase B.
Look into the mass driver circuit designed by Dr Gerrard K O'Neil and the Space Studies Institute. They got an acceleration of 100,000 gravities.
14:22 You just reinvented mag-lev :)
One thought, as your projectile moves it creates back emf. In short as it moves past the coil it will burn up almost all the “reverse” current.
This is a good thing. Think about a BLDC motor rolling. If it’s designed well and sized for its load there would be zero back emf. You’re coil gun is the same thing. If you tune the coil and projectile and cap correctly there will be zero inductive energy left in them after the projectile leaves the coil.
However if your projectile is to big, you will burn all your energy and generate back EMF before it has passed the coil slowing it down. Conversely if your projectile is to small you will not overcome your reluctance in time. By the time you do you will have no moving projectile to absorb the magnetic field and you will be left with energy still in your inductor after your capacitor has passed. Now that energy goes back into your cap backwards.
Point is if it’s perfectly tuned there is 0 EMF and 0 inductive energy in the coil at the end of the cycle. The ferrous projectile is nearly optimal for this job.
One caveat is that as the velocity changes all of this changes. You need a larger inductive coil at first or lower voltage with more amps. Each coil and capacitor should be tuned for the speed of the projectile at that point.
and now you can cannibalize real bullets
AAAA battery projectile, capacitor projectile, like the battery through the loop of coil train
I hope you find the equipment you need and get back to this project
Instead of copper tubes how about copper-jacketed lead like the firearm industry currently uses?
Αγόρι μου είσαι αυνανας.
same
I love how he abandoned his first design. I think his original design still has promise. Coil guns will never be effecient, that's not their job, they are supposed to be cheap. If you wanted effecient, build a railgun or true gauss gun with electromagnets instead of coils.
Water >steam is a 1600x expansion. And less energy required than all the electromagnets....
8:44 damping :)
Yup, sorry about that :)
So basically use a brushless ESC somehow.
As weird as you think the coil projectile is... you just invented the AC induction motor. Set it down, and pick it back up once your brain chews on it some more. And get yourself a scope :) Cheers!
That's mare like rail gun :D seem less effective
Rail gun
Adolph Quitler..
look at this coilgun czcams.com/video/izW1X2555Wg/video.html, amazing the power and speed that it has in "shotgun/a10 mode". Are you considering on going back to the coilgun proyect? it's amazing what the people who build the coilgun that is tested on that video achieved
Sound like you want to make a railgun instead of a coilgun 😏. DO IT!
You can get an oscilloscope for 20$ that actually work fine. To know where the projectile is in the coil gun you can use lasers and a photo resistor. But please just build the coil gun.
A simple diode would stop the negative
Going back
I just know i can get 3x the field strength in a coil wound my way
You spent the time and money on the current design already ... so just build it.
Make short entertaining build videos with sliiightly clickbaity titles, grow the channel and set up a patreon.
You are too focused on the end result and controlling all the technical variables.
Take a more easy going approach, the experience and funding gained from it should get you where you want to be.