So /diy/, due to watching a few videos I've decided to make a coilgun over winter break. Only problem is, I know next to nothing about electronics.
Now, I'm a decently good with my hands; I can take shit apart and put it back together, I can design bits and pieces and print them out, and I can solder. I just know next to nothing about electronics. I've only taken one class in university, and that was just pretty much "what is ohm's law and how do I resistors and capacitors". We did briefly cover inductors and transformers, but I've largely forgotten that.
I figured /diy/ would be the best place to start: how do I go about making a coilgun? Or, what do I need to learn to do so? I know I need large amounts of enameled copper wire, but don't know how to figure out how much I'd need or what I need to power it. Video below shows a prototype powered by a wall transformer (I think), but the 3D printed version uses LiPo batteries (I think). I'd be fine with using either.
https://www.youtube.com/watch?v=tTuX5pjljqk
>>913049
Buy a couple 555 timers, google up "555 timer projects" and do every single one of them. I'm sure something will start to sink in.
>>913049
Ok well yes you need a lot of wire.
Current in a wire creates a magnetic field. The strength of the field is proportional to the amount of current.
To create an electromagnet you need a large magnetic field. To increase the magnetic field in your wire you can increase the current or you can wrap the wire into a coil. Each turn will increase the magnetic field proportionally. This uses more wire which has resistance so you need to increase the voltage for the same amount of current. The problem is that a coil of wire is an inductor. What is an inductor? Well a coil of wire really. As the current increases and the magnetic field grows out induces current in the wire in an opposing direction which means the field doesn't just appear instantly it takes time to grow. So you need to factor that into your calculations later on.
For a power source you want something with low series resistance because you want the current into the coil as quickly as possible, people usually use capacitors for this, there are some fun things you can do to charge them in one configuration and then discharge in a different configuration but I can't quite remember why you would need to. Something about hvdc? It's difficult to get hvdc to charge the caps? Because you can't charge a capacitor with ac like mains electric because a capacitor will pass ac but block (and store) dc which is why they are used for filtering.
You can probably make capacitors yourself if you are insane using aluminium foil and some kind of dielectric like I don't know sandwich wrap I suppose if you want to keep it kitchen based. You will need to figure out the voltage they can handle by testing the dielectric sandwich wrap with an hv source to see what voltage pokes a hole in it. It will also junk around the sides so be aware of that.
So now you need to figure out what you want to fire. Ideally something ferrous. If you know its mass and air resistance and shit like that you can calculate the force required
>>913142
To get the projectile up to your target muzzle velocity.
If you are switching huge amounts of power then hv spark gaps might be a good idea or else some fat mosfet or something. You could use it open loop with just the calculated timing or go closed loop and put a few detectors to switch the coils at the right times.
Relays might work but remember the magnetic field will take time to dissipate when you remove power just like it took time to grow.
The other thing is that if you put too much power through the coils they will overheat and when they get hot they will increase in resistance and so the current and therefore force will decrease too. Make your coil wire oversized or only fire it once an hour or so.
Or water cool the barrel I guess.
Anything else?
>>913142
Foil and wax paper, rolled up, wrapped in thick paper and sealed in resin.
Thats how they made (admittedly shitty by todays standards, but large) caps back in the early 1900s.
Wax paper and aluminum foil are probably the best easily available materials for home made caps.
>>913268
>but is there a reason I need specifically capacitors instead of a LiPo battery?
discharge/recharge rate
you need to dump those electrons into the coils asap
>>913268
>I'd buy capacitors if I need to, but is there a reason I need specifically capacitors instead of a LiPo battery?
A "real" coilgun uses a monstrous levels of current and voltage (talking hundreds of volts and hundreds or thousands of amps at peak) for very short periods (millisecond range or less). You'd need a hugely oversized battery pack not because it takes a lot of energy, but simply to meet the peak power requirements.
Capacitors are a lot better at sourcing large bursts of current then batteries, which is why they're preferred. There's also the issue that it's not uncommon to switch the coils with some form of SCR that can't be turned off once it gets going. Thus, it's necessary for the power supply to run out on its own, which a battery back won't do until everything's turned to slag and left an open circuit.
Really, though, you shouldn't be dicking around with high-voltage battery packs if you don't already have a very good idea of what you're doing, ESPECIALLY ones with such a spectacular catastrophic failure mode as lithium polymer. And high-voltage capacitors are arguably the most dangerous thing you could ever deal with in electronics.
If you stick with a lower-voltage, lower-power gun (which, again, you should until you're more familiar with electronics), a high-current lipo pack would work well, with the added benefit that you don't have to wait for it to recharge between shots (like you do with a capacitor bank).
>>913241
also leyden jars. Tesla builders love leyden jars.
>>913296
>low-voltage, low-power
Exactly what I'm wanting to do; start small. I don't want to electrocute myself or something because I didn't know enough about this, that would be dumb.
I was thinking of using small ball bearing balls as the projectile at first, would that be decent?
>>913328
>would that be decent?
It'd work. "Decent" is relative, though.
Actually getting an efficient transformation of the magnetic field produced by the coils into motion of the projectile is by no means a trivial feat. That's not something to concern yourself with right now, though. Really won't be up until the point that "just dump more energy into it" stops working well enough to be worth it, actually.
>>913346
So for now: get large-gauge enameled copper wire, learn to use 555s, get capacitors, and apply Google liberally? Or is there something I'm missing?
>>913358
Don't bother with capacitors for now. Again, they're only necessary if you NEED that 100kW+ peak power capability, which a small coilgun simply doesn't.
Reading through all the info you can on the subject is a good idea no matter what you're doing, of course.
And don't buy anything until you've got a solid plan and design laid out. I'm guilty of this all the time and I have such a ridiculous amount of crap and miscellaneous parts lying around from projects that didn't pan out or are still in the works I don't even...
>>913361
So I'm unsure what to use for power. Batteries, or wall socket (maybe an old USB charger or something)? Or will this become obvious as I learn more?
>>913362
It should become apparent as you learn more, but, quickly:
A wall wart or other small transformer/SMPS is unlikely to have enough power except for very small projectiles. It will certainly work, it just won't have much grunt.
An old PC power supply might work alright. They can usually source a fair bit of current on the 12V rails.
A battery (particularly the ones used for RC stuff) is probably what you'd want to go for. They can source quite a bit of current and are portable. Bit of a downside is that you'll need a special charger for them. And lipos do not take kindly to damage, as some YouTube searching will illustrate.
If I were you, I'd probably lean toward the PC supply. Between the availability, reasonable output, 5V and 3.3V outputs for any logic or lower-voltage parts, and various protections built into it to hopefully keep you from starting a fire or electrocuting yourself, it's not a bad good option for babby's first coilgun. Just don't expect too much. 12V is kinda low for this sort of thing.
>>913367
Splendid, thanks; I have a bunch of old PSUs and I can take apart / brick with no consequence.
I've got experience with lipos, but would probably make a metal enclosure for them in case of accident. I've got a few puffed ones laying around, which I should really get rid of, but might as well try to use them until they die.
I'll be happy if I can hit a sheet of paper from a few feet away, will build on that once I get to it.
Dumb question: Is there a reason people use specifically enameled copper wire instead of insulated for their windings?
Since the enamel prevents conduction through the wire anyway, why would plastic/whatever insulated copper wire not work? It's a lot cheaper and I already have a ton of it.
>>913583
Flux density. You can fit more conductor in a given volume of space (in this case your coil) with enameled wire, simply because the insulation itself takes up a significant amount of room on its own. So you can either fit the same amount of magnetic flux in a smaller volume, or more flux in the same volume.
Regular PVC jacket shouldn't be significantly (if any) cheaper, though...
The thing about a coilgun is that you need to increase the coil current quickly, and reduce it just as quickly (or even more so).
An energised solenoid (coil) pulls ferromagnetic objects towards its centre. If you want to accelerate an object in a particular direction, you need to energise the coil to pull the object toward the centre of the coil, then (this is the important part) de-energise (quench) the coil as the object reaches the centre. If you don't, once the object passes the centre, you'll end up pulling it back toward the centre, i.e. decelerating it. If you just energise the coil and leave it energised, the object will just oscillate before reaching equilibrium at the centre of the coil.
A fundamental property of a coil is that the current through it cannot change instantaneously. The rate of change of current is dictated by the applied voltage and the inductance of the coil (which is determined by its dimensions and the number of turns). Higher voltage = faster rate of change of current, larger inductance = slower rate of change. dI/dT=V/L.
So, the "performance" of a coilgun depends upon:
- How much current you can supply. The more current, the better.
- How quickly you can increase and decrease the current. The more voltage, the better.
As for the coil itself, fewer turns means lower inductance, which means that you can increase and decrease current more quickly. It also means less force for a given current, so you need higher current to compensate.
But wire is cheap. The main limiting factor is how much you can afford to spend on transistors. More voltage and more current both mean more expensive transistors. The rest is just getting the timing right.
I reccomend starting small OP. Learn the basics of a coilgun by following some diy stuff on instructables.
Pic related, simple coilgun I built for a school project.
So, just realized something; I have a few power supplies like >>914625 sitting around; using these would be way easier (for a prototype) than a wall socket or computer PSU, right? Or would there be something else I'm overlooking?
>>913049
A spring and mechanical build could give the same results as this for much cheaper cost per shot. I don't know much about this setup, but I'm not so sure that this could ever be a viable replacement for chemically projected firearms, unless there is some sort of breakthrough I'm unaware of. What is the benefit of this, besides cost of projectiles?
>>914901
(Military grade)Railguns can get heavy projectiles to a velocity of several mach very easily (I'm sure this can also be achieved with a multistage coilgun, but I bet it's less practicable).
In the end, from the point of view of a hobbyist this kind of projects are just fun/interesting to make.
>>914901
>>914916
Ek = 1/2mV^2.
Chemical projectiles are limited to the detonation speed of the propellant; after that you can only increase the energy linearly by increasing the mass.
EM accelerators are not limited to any speed, so you can keep the projectile mass resonable, yet increase the impact energy as much as you like.
>>914901
At the kind of scale the OP's talking about, they're not a practical replacement for a firearm, crossbow, speargun, etc.
At larger scales, railguns or coilguns have a number of advantages. The muzzle speed can be much higher than is achievable with chemical propulsion. You can use a ship/vehicle/aircraft's existing fuel supply to power the generator, so rather than needing X amount of fuel and Y amount of explosive, you just need fuel which can be used for either. And it avoids the need to store explosives.
Current military research is focusing on railguns, as the projectile only needs to be conductive (for a coilgun, it needs to be paramagnetic). Coilguns have the advantage of not requiring contact and thus having less wear; large-scale railguns tend to turn a significant portion of the rails and/or projectile into plasma.
>>913049
not exactly a coil gun but still kick ass
https://www.youtube.com/watch?v=vAs9EHtKfVc
>>915386
>not a practical replacement for a crossbow
Isn't it? If I made it slightly longer than forearm length, had enough coils, and had a small enough projectile, wouldn't it be fairly effective at closer ranges?
>for a coilgun, it needs to be paramagnetic
I've always wondered about this; why not have a magnetic insert on rails that acts as a catapult for the actual projectile?
>>915792
How d'you retrieve it at the end of the barrel?
If you're not planning on retrieving it, why not make it the projectile?
>>915792
>wouldn't it be fairly effective at closer ranges?
No, not really.
For example, if you got a 20g crossbow bolt travelling 100m/s, you're looking at 200j of energy in the projectile (for reference, a .22lr is only 160j).
Your average coilgun will struggle to see even 3% total efficiency. You'd need nearly 7000j of capacitor (that is a portable, but still hefty bank) just to match a typical crossbow.
For a point of reference, some cursory YouTube searching pulls this up:
https://www.youtube.com/watch?v=7LjnhhtHojM
That's only 1.25kJ in the capacitors. You're looking at a bank that will likely be at LEAST five times that size. For a single shot (how fast you can shoot is going to be determined how fat your power supply is).
Coilguns, as it stands, are simply aren't really practical for anything. Other than being cool, at least.
>>916125
>Your average coilgun will struggle to see even 3% total efficiency.
Lurker here. Why is this?
>>916161
Basically, everything you're trying to get to work for you in a coilgun is actively attempting to sabotage the whole shit.
The magnetic field goes all over the damn place instead of acting on the projectile, the projectile saturates above a certain flux density so adding more/directing it better doesn't help much anyway, inductors resist the fast changes in current that you need, resistance is still a thing, etc.
Basically a whole bunch of crap adds up to virtually guarantee that 95% or more of the energy in the capacitors doesn't do anything useful.
I recall reading some random site about some guy experimenting with how to get the efficiency up, and, if I remember right, he only managed to just top like 6% with a reasonable projectile velocity.
You could actually easily get up to ten times that, IF you were direct the flux better (like you do in a motor with the iron laminations) and use a very heavy, large, and slow projectile. But at that point it's not much of an improvement over just chucking some steel balls by hand.
>>913049
There OP, i made a shitty gif
>>916132
or just using the same magnets in reverse at lower power to pull it back... I thought this was an electronics project.
I Have a question, would a coil with much less windings and much lower gauge be better for this? Would it decrease inductance so it could charge and discharge quicker? Or would the decrease in inductance require such a higher current to get the same strength it doesnt really matter.
>>916270
So... bigger, more complex gun with lower efficiency and lower rate of fire in order to be able to use bullets which have no steel parts in them?
Why would this be a reasonable compromise?
>>916272
If you have circuitry controlling the magnets making them run backwards isnt really costing you anything.
>>916271
Fewer windings results in lower inductance, but then you need more current to get the same force.
>>913142
>You can probably make capacitors yourself if you are insane using aluminium foil and some kind of dielectric like I don't know sandwich wrap I suppose if you want to keep it kitchen based. You will need to figure out the voltage they can handle by testing the dielectric sandwich wrap with an hv source to see what voltage pokes a hole in it. It will also junk around the sides so be aware of that.
Everything about your story is close to true except for this, you are NEVER going to be able to roll caps used for supplying a coil gun yourself. (and if you oppose: please show me anyone who has managed to make a cap of several dozens of µf themselves, let alone the thousands needed)
>>916274
You need the same amount of energy to stop the sled than you used to accelerate it. This matters, unless we are talking about the wimpiest toys possible. If you want to use only the existing magnets without reducing the bullet energy, you'll also need a mechanical arrangement to bounce the sled back, with the associated problems.
Again, why would this make sense in a gun?
Can someone do a shitty /k/-tier infographic on how to make one of these?
>>914290
>something like pic related
>>916695
>probably
>if you are insane
i dont know man
i haven't tried it
but it might work.
why not?
>>916695
I couldn't help but comment.
Granted, the cap in my hand was professionally rolled, I have a 500MFD (old school before the whole µ thing) aluminum dialectic capacitor. Similar enough to what would be made at home.
Over all, I concur, that while HV caps can be hand made, due to the potential for failure alone, they should not be used for anything more than a novelty factor.
With that being said, theoretically one could make a comparable cap.
>>917095
And I dropped the pic
>>917095
>aluminum dialectic
>500µF/150V
>polarity marked
That's an electrolytic capacitor and they're built differently.
Not that you can't buy professionally made plastic or ceramic capacitors with huge capacitances, but we were talking about homemade ones, made from aluminum foil and plastic sheet.
>>917101
That plastic sheet IS a dialectic. So lets see, rolled aluminum dialectic vs rolled aluminum foil and a dialectic.
Anyways homemade would have a high failure rate.
>>916743
>need the same amount of energy to stop the sled than you used to accelerate it
Not quite. A hydraulic / spring stop at the end would work. Zero (active) energy required.
>>913049
>know next to nothing about electronics
>want to build coilgun
>want to use bigass lipos
What could possibly go wrong...
>>917253
It's not me who wanted to use a magnetic brake. But yeah, you can stop the sled by mechanical means. The question is still the same: why would this make sense in a gun?
>>917219
Oh, that "dialectic" wasn't a random typo.
The dielectric of an electrolytic capacitor is aluminum oxide, not plastic. It's very thin and that's why electrolytic capacitors can have very high capacitances in comparison to their size.
>>919025
If some ten-year-old can do it using a 9V and a glue gun, there's no reason I can't try the same but safer.
rail guns are easier to make and have greater energy potential.
>>913131
Not OP, but I actually built a coil gun using 555 timers. The timers trigger mosfet driver chips that in turn power the coils. As a result, it's totally 'programmable' in that you can adjust the timing with three slider pots. Plugged it in to 170vdc and it worked like a charm.
>>Tfw no bullshit capacitor networks to mess everything up.
>>tfw no optical components drilled in to the barrel for timing
I should really publish my notes on this.
>>919364
Yeah, I'd appreciate some kind of documentation. So you had no power storage, just a direct line?
>>919374
I took 120vac from the wall, rectified it to
170vdc, filtered it, and provided it directly to the power mosfets.
Just got off work for Christmas, so I may have some time to go looking for those old files. Pretty sure I have a pdf that's half way done, as well as express pcb files and a BOM. It was many years ago when I built this, and I was a lot dumber back then, but if I can dig up the files I'll post a link here, even knowing that /diy/ will likely rip in to me and tell me every little thing I did wrong.
As a preface to posting the files, why don't I just say that we can all use my design as a starting point, and refine it to make the official /diy/ coil gun, Rev. A.
Commencing namefagging.
>>919407
Go for it. I'd be very interested in seeing whatever notes you publish, if only to use as a starting point that (I assume) works. If I can improve it, I gladly would and would publicly (or to you) publish improvements / modifications.
>>919421
Well, I found schematics, pcb drawings, and the BOMs. Still looking for that pdf and some actual pictures. If nothing else I can go down to the basement and dig out the box and take some pictures of the assembled boards and rifle.
I'll be back with the files in an hour or so. Can I just post a Dropbox link, or is 4chan gonna hack my Dropbox?
>>919600
If you're concerned, use FileDropper. Served me well in the past and is pretty simple (though non-dynamic like dropbox). DB should be a "view-only" sort of thing, I've never seen a name attached unless it's on purpose. You could go in incog/private browsing and see what a shared link looks like from there.
>>919620
Thanks. Instead of spending a bunch of time figuring it out, I guess I'll just take my chances.
https://dl.dropboxusercontent.com/u/40563809/Coil_Gun.zip
It turns out that the pdf is from 2005 when I first tried this. The schematics are from 2012 when I updated them and ordered boards from Express PCB. You should be able to order the boards directly from Express PCB as well, using the same files. If anyone here shows particular enthusiasm, I may still have a blank board laying around that I could mail out. The BOMs have digikey part numbers that should still be good, so you can look up the datasheets. Please let me reiterate, this is a prototype and a starting point for the official /diy/ coil gun, Rev. A. We can talk about all of the problems and shortcomings, but personal attacks will be ignored.
As a final note, please please be careful. The voltage in this circuit is extremely deadly. You will literally an hero if you are not extremely careful. I tried to build in some protection, like an HV warning light, an HV discharge button, and a key lock to turn the power supply on, but none of that will matter if you're sticking your fingers on the circuit board while it's energized or even when it's off but the filter caps are still charged.
With all that in mind, here's the files, as promised. Let's finalize a design for a coil gun folks. I should have a few more pictures to contribute later. I'll be around until the thread dies I guess.
>>919634
Also, I'm doing this from my phone, so would someone please verify that the link works and you got the files?
>>919642
Looks like it does. Will look at it in-depth in a bit.
>>919642
Also, in case it's not clear, there are physical wires that connect the PS to the timers and mosfet boards. There is also an integrated fan. I'm not sure if I actually mounted the mosfets to the board, as indicated on the pcb, or if I wired them to the board and mounted them on a heatsink.
Looking at these files is getting me interested in working on it again, so I probably will head down to the basement tonight to snap a couple of pictures.
>>919670
I guess the first modification I would make is to remove the 9vdc power rail and run the 555s off of 12vdc. That would get rid of the lm7809 and clear up some space to make traces wider on the PS board.
>>919736
Anyone take a look at these schematics yet?
>>920966
Haven't had time, I'm afraid, been busy over the weekend; will take a look around Tuesday and hopefully start making my own around then as well. Will update thread with my progress.
While we're playing with magnets, has anyone built a maglev model train?
So, coilguns are super inefficient, but relatively easy to build if you don't mind having something that's basically just an office toy.
What about a rail gun? Are those more efficient? Why hasn't there been as big a hobbyist surge with rails as there has been with coils?
>>922088
> What about a rail gun?
A railgun doesn't strictly require any electronics (there's no timing involved). It just requires getting hold of some pulse-discharge capacitors (a railgun only has a single turn so it needs far more current) and some mechanical engineering.
At the low end, a popular choice for capacitors was from the flash circuit in disposable cameras. But those are largely extinct now. Larger pulse-discharge capacitors are fairly expensive unless you're lucky enough to get hold of some used ones which are a) still functional and b) significantly cheaper than new.
If you try to use regular capacitors, you'll either get something that's not even a toy (not enough current) or you'll fry the capacitors on the first shot.
tl;dr: expensive and not particularly interesting.
>>913268
The thing is, you will need to be able to load the capacitors with a specific amount of energy, without damaging the batteries. When the discharge occurs, the capacitors will be able to deliver the required energy in a blink.
>>922088
You have it backwards. Coilguns can be very efficient, railguns much less so. But coilguns are considerably less adept to achieving extremely high accelerations and velocities that are possible with a railgun due to inductive and switching limitations, and are more adept to moving heavier projectiles at lower velocities (whereas railguns are more effective with very small projectiles at high velocities). Most homemade coilguns barely resemble "guns" at all, due to the challenges involved with achieving respectable velocity with them.
