Hey guys,
I'm making a coil gun. I plan to start with two stages powered with capacitors. Any tips for me? Where should i look for wire? (eurofag here)
>>1001129
>eurofag here
DELET THRED
>two stages powered with capacitors
timing will be a real pain.
>>1001151
My plan is to use arduino and photoresistors or phototransistors placed in the middle of each coil. I think it should work.
>>1001129
Nice
>>1001129
Nice
>>1001174
why do you need an arduino? is it realtime?
connect your interrupt to an adjustable timing circuit and tune it manually.
if you want to you could use the arduino to check the final speed for optimizing.
>>1001178
I have a few arduinos and nothing to do with them.I also thought there will be more shit like switches, diodes and charging circuit, so it will be easier to hook it all up to one arduino.
Did you check TME eu? Prices aren't too horrible even for small orders.
>>1001129
The main thing about a coilgun is that you need to cut the coil current when the projectile reaches the centre of the coil, otherwise it will retard the projectile (pulling it back into the coil).
>>1001233
I just bought 300 meters of 0,3mm wire. I think it will be okay and it was cheap as fuck.
>>1001129
You gonna die.
>>1001129
>Where should i look for wire?
If you don't know this, you don't have enough experience do finish this project.
>>1001129
Why not use conductance in the bullet to generate a field an send it flying without any coils and shit?
>>1001853
Huh? You mean a railgun?
Because OP's thingy is much cheaper and easier to build.
>>1001860
Why is it cheaper? You need copper coils, that's expensive! The concept of a railgun needs less materials?
>>1001870
Because powering a coil gun is cheaper than powering a railgun. The cost of the copper wire is trivial in comparison to the other costs.
>>1001129
1st off, have fun anon!
2nd I think you'll find it a bit more difficult than it appears. Coils and electromagnets are nothing new and getting a strong field is easy. What isn't easy is both strong and quick and you need quick because as >>1001245
points out (and is almost correct)
>you need to cut the coil current when the projectile reaches the centre of the coil, otherwise it will retard the projectile (pulling it back into the coil).
it's not the current that needs to stop it's the magnetic field that needs to be GONE and for this to occur before the projectile passes center point the current must be stopped even earlier because the magnetic field will take a brief period to die out completely. If you have any solid state devices in the circuit with the coil the counter EMF of the collapsing magnetic field can cause problems. If you have a capacitor in the circuit you will have resonant oscillations to worry about as well.
Magnetic field strength is all about Ampere - Turns. How many Amps thru how many turns of wire. To get the Amps you need Volts to push it and many coil guns charge their capacitors to 400v and more to get the needed short rise time pulse to drive the coil.
I've always thought that a simple method would be to use small contacts in the barrel instead of active timing circuitry. The small contacts would be placed at the start end of the coi such that the projectile would complete the circuit to discharge the capacitor thru the coil only momentarily as the resulting magnetic field would cause the projectile to move down the barrel and off the small contact area. This would create the needed small pulse and insure the current was off well before the projectile reached the center of the coil.
This should work as the barrel could be made of plastic or other non-conductive material. The projectile will be accelerated towards the center of the coil / center of barrel bore by solenoid action anyway
Op here sorry i abandoned the thread for a while but now I'm back.
>>1001853
As far as i know building a railgun is way harder and more expensive.
>>1001440
Thanks for your info on photoresistors.
What do you mean by saying i fucked up? Is the wire too thin? It's not a real problem tho because it was really incredibly cheap.
>>1001965
Thanks so much for your comment. I have 4 nichicon 4700uF 400V capacitors and it's probably the best i can get for now. I'm gonna research the magnetic field dying time, but it would be great if you could elaborate some more on it. Your idea with contacts is nice but wouldn't it be less reliable than phototransistors? Also could you tell me some more about resonant oscillations, because I can't really find anything on this topic in my application.
>>1003119
>railgun is way harder and more expensive
If done poorly, best case it will weld together, worst case it will explode (flash). Not to mention coilguns theoretically don't wear down. So yeah, coilgun is better in that department.
>Is the wire too thin?
Thinner wire means higher resistance per meter, so lower amperes for the same voltage power source.
As solenoid's magnetic field strength is directly proportional to ampere-turns, you'll need more turns of thinner wire to achieve the same field strength.
Having more turns increases inductance which in turn makes the current curve rise and fall slower which is the exact opposite of what you want in a coilgun.
There's also the issue of heating and peak current duration on your switch of choice (semiconductors).
tl;dr yes
>4 nichicon 4700uF 400V capacitors
Why does everyone have access to those 4.7mF 400V caps? Am I missing something?
>magnetic field dying time
Only need to worry about the current actually. Since the collapsing magnetic field induces a current in the coil, the current passing through a given coil at any point in time is directly proportional to the field.
There are simulators for that too. Saves a headache plotting the thing yourself.
>contacts
Not a fan of the idea. Only complicates the design, wear down, less reliable and might cause your switching stage to overheat if the contact isn't ideal.
>resonant oscillations
https://en.wikipedia.org/wiki/RLC_circuit
A coilgun is basically an RLC oscillator with a switch. Except you want less than half of the first sine.
And don't forget your snubber diodes or you're guaranteed to blow your semiconductor switches.
>>1001290
These could work provided you use high voltage, somewhere in the 300-600v.
Try using IRF MOSFETs to switch it on and off. If not then try SSRs.
>>1003196
Just jumping in here, but I totally agree with using mosfets instead of capacitors. They are cheap enough and it's a lot easier to change the timing of a mosfet than it is to drill new holes in your barrel to reposition photo interrupters. Not to mention working with big caps is tough because the tolerance on them is so wide (+/-20%), that you will never get two caps that are exactly the same and you will have to modify your design to fit the caps that you have.
I built a coil gun back in '05 or so using mosfet timing, and I even shared the old schematics here on /diy/ A few months ago hoping that people could use it as a starting point to create a diy coil gun and improve on my design (which could definitely use some improvement), but it turns out that the people who start these coil gun threads never follow through. They always lose interest after they wind some coils but before they build the circuits. Hope that doesn't happen to you OP.
>>1003203
>using mosfets instead of capacitors
That makes no sense. You probably meant using active switching instead of relying on the rlc damped response for timing the shutoff?
Sorry, I'm pretending to know stuff by using big words.
>old schematics
Care to share them again? I'm planning on building a smart multistage one myself (just some primitive self-calibration that will probably get messed up by projectile inconsistencies). Not this quarter though (swamped with work), but I'd like to see your schematics.
Currently trying to work out optimal coil parameters. And what to use for switching that wouldn't bankrupt me. MOSFETs are expensive at the high power end, even if I gamble with peak current.
Also thinking of one with a lot (32) of narrow stages firing bearing balls.
>>1003212
You are kind of right, mosfets themselves don't provide power, they only switch it on and off. My design did not include any caps for providing power because caps suck at providing quick stable power. Instead, I designed a high voltage dc power supply for the mosfets to use. Basically, I took 120vac from a wall outlet and rectified it to 170vdc, then added a couple of filters to smooth it out. It worked great, even on a 15A outlet; since the power it provided was in quick bursts, it never threw the breaker. It does require that you plug the gun in to the wall though, which is a limitation.
Mosfets are still kind of expensive, but really only a few dollars each. The price for power mosfets has come down significantly over the last 10 years. Also, sometimes you can contact the manufacturer and get some free samples if you tell them that you're working for a company that is considering including those mosfets in to a product.
Protip: the hardest type of ammunition to fire is a sphere, so if you will use ball bearings, be sure to put them in a press first to flatten them out somewhat. Bolts are much easier to fire because they are already roughly cylinders.
I'll see if I can go back and find those schematics again. Nothing personal, but I don't have a high level of confidence that you will actually use them. If anyone is interested, the schematics include the circuits, the pcb layouts, and the bill of materials. I may even have an extra copy of a couple of the boards left over. I'd just prefer to give it to someone that will actually use it and potentially improve the design. It was a great hobby but I simply don't have the time to work on it myself. All I can do at this point is provide the design and some pointers to someone else who wants to run with it.
I'll see if I can dig them up when I get back to my desktop.
>>1003224
>caps suck at providing quick stable power
That...that's the entire point the point of putting a filter capacitor in a circuit, and why you will never see a railgun (which demands much higher power than a coilgun) of appreciable size run off anything but a large bank of pulse-rated capacitors.
You can get away with using other power sources in a coilgun specifically _because_ the minimum power requirements can be lower. But then you also have to account for actually switching off the coils, which, in higher-powered guns operating at hundreds or thousands of amps, is not necessarily trivial.
For projectiles, just buy some steel rod and cut/grind it to size and shape. It's cheap and should be plenty straight.
I've just finished a test with one layer of the coil. Powered from one of my old projects ( basicly a 250v 1000uF capacitor) it is able to push a 7mm steel bearing ball through the barrel so i think it's not that bad.
My main problem for now is a way to charge 400v capacitors. I will probably have to buy a 230v/400v transformer.
>>1003135
Thanks for your post. And about the capacitors I bought them on my country's auction site for about 6 dollars each.
>>1003224
It would be great if you could find your schematics. Also why is a sphere the hardest type of ammunition to fire?
I also plan to hook it all up (detectors and mosfets) to an arduino or some other microcomputer to make the timing and overall control easier.
>>1003224
>wall outlet
Was actually considering that for a few moments. Mains is 230 here and it sounded like a very bad idea safety-wise though.
>Mosfets are still kind of expensive, but really only a few dollars each.
Only if you rely on peak current specs which might or might not destroy them. The beefy ones get pretty expensive.
My plan was to parallel two of them. Double the performance at just double the price. Since they are positive temperature coefficient, they would always be balanced.
>hardest type of ammunition to fire is a sphere
Thus the large oddly shaped coil number.
Round objects don't wobble during flight though. It's as much kinetic energy as it is delivering it to end target. No point building a coilgun that shoots in random direction and has no range.
>don't have a high level of confidence that you will actually use them
Not taking offense. And you're probably right in that I won't use them. Certainly not as is, but maybe I'll take a detail or two from them. Mostly I'm curious.
Did you make any calculations for coil optimization or just made a guess?
>>1003232
>7mm steel bearing
Try a nail roughly the length of the coil.
Hope you didn't forget the snubber or you're destroying your capacitor.
>charge 400v capacitors
Boost converters. Might be horribly inefficient, but they are a lot more portable than a transformer. And more practical.
>why is a sphere the hardest type of ammunition to fire
Once the entire projectile is inside of the coil, the magnetic field is uniform and there's no more force being applied to it.
>arduino or some other microcomputer to make the timing and overall control easier
Got the same idea here too. Saves screwing around with hand tuning the timings. Just if you use arduino, use proper C and not that arduino whatever-it's-called crap. You want it to run fast, not to be artist-friendly.
>>1003227
True, caps can provide a large amount of power, and they can deliver it quickly, but stable power comes from the combination of an actual power supply and a filter (tank) capacitor. If you just discharge a cap in to a circuit, the amount of power that it delivers directly depends on the RC time constant, and is smaller even one dt later than the instant that you start discharging it.
Rail guns use caps because of the high amount of instantaneous power needed and the fact that you can get that instantaneous power by charging the caps up with a low cost power supply. It's really just a trade off in that you can get really high power in exchange for a slower rate of fire. It's possible to use a regular power supply to power a rail gun, it's just not worth it financially.
>>1003237
I considered doubling up the mosfets too, but when I was experimenting I started with just one for each coil and didn't have any problems.
I never did make any specific calculations on the coils. I started this back in 2004-2005 and was a lot dumber back then. It looks like I updated the schematics back in 2012, but I was still working with the 3 coils that I started with so I still never made any additional models or simulations. I guess I was just hoping that the flexibility of the timing would be enough to optimize it.
>>1003240
If you are going to use caps for powering the coils, a boost converter is a really good idea. It may be a while before the caps are fully recharged, but a boost converter would be small and cheap. You might consider putting in a small volt meter and led indicator on each cap to indicate when it is fully charged.
I found my old files here, and I'll send a Dropbox link in a few minutes, but first let me preface it with the fact that I was a lot less experienced with design when I made this coil gun. I'm sure that there is plenty of room for improvement, including using an arduino for timing instead of a series of 555 chips. 4chan can rip on me all they want because I should have done this or that, or this is inefficient or that isn't safe. Go ahead and rip on me, but this design did work which was enough for me at the time. Moreover, actually bringing a design to completion is probably more than 4chan trolls have ever done, even if it isn't a perfect design. With all that being said, I did design a couple of safety features in, but this is still an 'attempt at your own risk' project.
>>1003247
The schematic files and pcb files are all in Express PCB format, since that is the company I used to order the boards. It's a free download from the Express PCB site. The BOMs are in excel files.
The power supply includes a key switch to prevent just anyone from turning it on. It also includes an led to indicate HV is present and a button to rapidly discharge HV after turning the supply off.
The switch array board has 3 mosfets, q1-q3 and a mosfet driver for each of them. The mosfets can be mounted directly to the board, or you can use a bit of wire to extend them to a bigger heatsink if you want.
The timing board has three 555 chips, each configured as a monostable oscillator, and each one triggered by the falling edge of the previous one. Use the potentiometer to adjust the timing of each trigger individually. It also includes a plug so you can daisy chain a second timing board to it to control up to 6 coils.
I'll be happy to go in to further detail if you guys have questions or if anything isn't clear. If you want to build this design, you can order the boards directly from Express PCB inside the pcb software. If you know how to etch your own boards, you can print the files from the pcb program and they will come out at the right scale. I think most of the parts in the BOMs have digikey part numbers.
Here's hoping that this will be useful to someone!
https://dl.dropboxusercontent.com/u/40563809/Coil_Gun.zip
>>1003257
Also, please let me know if you downloaded the files, since I manually typed in the Dropbox link on my phone.
Also, to replace the timing with an arduino, just get rid of the timing board all together and feed your arduino outputs to pins 1, 2, and 3 on J2 of the switching board. The output straight from the arduino should be enough to trigger the mosfet drivers directly.
Things that I already know I would change about this design:
* I would probably select a different type of connector than the ones in the BOM because those IDC connectors ended up being fairly hard to work with.
* I would still use an aluminum barrel, but I would cut it in to three sections and mount a coil at the end of each section, then couple them together with a plastic fitting like in pic related, this would help reduce eddy currents from slowing down the projectile because the 3 sections would be electrically isolated from each other, and when you turn a coil off the projectile only has a short distance to go before it gets to the next isolated section.
>>1003266
Well, I hate to dominate this conversation with 3 replies in a row now, but digging up those old files got me interested in this again.
Anyway, turns out that I do still have a few unpopulated boards and some unused parts still laying around. When I order from Express PCB, I always do the cheapest option, which includes 3 copies of the board.
Is anyone actually interested in building this thing?
>>1003245
Power supply with caps is mostly the same as just caps for this application. If the actual power source can't provide enough current, the voltage drops to where the current draw becomes the max it can provide.
>>1003247
>didn't have any problems
IXTH110? Pulse seems to be limited at 300A. So lots of firing and no blown MOSFETs?
>>1003266
>please let me know if you downloaded the files
Link's good. Got them but don't have the time to check the schematics right now.
>tube
I'm considering using carbon fiber from a hobby shop. It's ridiculous how hard it is to find any rigid tube in not-pipe diameter here.
>>1003283
Yah, IXTH110. 300A sounds about right. The mosfets need heatsinks, don't get me wrong, but the pulse is short enough that they can withstand it. I blew some pretty expensive mosfets back in '05, but I don't recall blowing any from this revision. If I worked it harder, I'm sure I could have eventually blown one, but the 18AWG wire in the coils started heating up pretty quickly too, so I never really pushed the thing to the max.
>>1003283
>>1003286
Also, don't forget there's a limit to how much current you can get through one of these coils anyway. For example, my coils were all wound with 18AWG enameled wire, they had an inductance of around 15 mH, and they had a DC resistance of around 3 ohms.
Well, with a DC resistance of 3 ohms, and an applied voltage of 170 VDC, that results in a DC current of only around 57 Amps. The coil itself is a current limiting device. As a result, 300A seemed like a pretty reasonable choice at the time. You still need heatsinks though.
>>1003301
>3 ohms
Are you sure? That's ~140m of 18AWG copper wire.
>>1003385
Pretty sure. Each coil was about 3 lbs. I would estimate each one comprised around 500', which is pretty close to the 140 meters you quoted. I found some notes that I made that said 3 ohms. I have another note that says the whole rifle weighed around 10 lbs, and that was all 3 coils plus the aluminum rod.
Why? How much wire are you guys using to wind a coil? What gauge? Even at 18awg, the coils started to warm up after only a few shots, so next time I was thinking of using 16awg if I could.
>>1003419
Again, I didn't do any calculations on these coils before I wound them back in 2004, I just figured the bigger the better. So I went out and bought a really big spool of wire and wound 3 coils. I'm kind of interested in what the 'optimal coil calculations' look like now. Do those calculations take heat disapation in to account? What is the optimal length/gauge/# of layers/# of turns?
>>1003232
use the left side of this circuit, it might take a quite a few minutes to charge your caps. you can also use a disposable Kodak camera's Flash circuit, it's cheap.
>>1003496
How does this work anyway? Is this like a voltage tripler?
>>1003710
https://en.wikipedia.org/wiki/Boost_converter
>>1003423
>I'm kind of interested in what the 'optimal coil calculations' look like now.
I've always wondered about the wisdom of all the coils with the same geometry.
that is, each coil has 500 turns, should all three have 10 layers with 50 turns per layer and thus each coil have identical length?
or should the 1st coil be wound with 25 layers of 20 coils and the 2nd be wound with 10 layers of 50 turns and the 3rd coil wound with 5 layers 100 turns.
thus each coil getting longer, corresponding with travel time in each barrel for the magnetic field to impart acceleration to the projectile.
>>1004619
I've been wondering the same thing. When I wound my coils, they were all roughly the same and with the same spacing between them. I believe you may get a stronger field by having a shorter length of coil with more layers, but you would probably also get better results by increasing the distance between subsequent coils since the projectile will be moving faster after the second coil than after the first.
I've been getting interested in this again after I found those schematics and boards. I've got an extra arduino laying around too, and it would be pretty easy to replace the 555 timers with it. My only problem is that I scrapped the old rifle when we moved recently, so I would have to dump another $100 in to wire to make a new one, which I wouldn't want to do unless I can figure out how to make a well designed rifle first. I've got more resources avaliable this time, like the ability to 3d print spools to wind the coils on, but I'd love to hear from these people who are actually 'optimizing' the coil design first.
>>1003419
>Each coil was about 3 lbs
Damn. My current plan is only 40x40mm with 20mm air core.
>>1004619
>for the magnetic field to impart acceleration
Once the projectile is completely within the solenoid cavity, there's no acceleration since the magnetic field is uniform within it.
Might be worth doing the opposite. With the first coils being on longer, thus being affected by inductance a bit less.
>>1004630
>3d print spools
You just solved my spool issue. Now I have a 3d printer issue. And my frame issue was solved by making it a CNC issue... It turned from a simple project into a fully modular computerized smart gauss rifle (with spin included). Help.
>>1001129
are these the weapons of the future?
>>1004644
>Once the projectile is completely within the solenoid cavity, there's no acceleration since the magnetic field is uniform within it.
A magnetic field's strength falls off in proportion to the distance from the center of the field.
The acceleration of the projectile continues as long as the projectile is under the influence of the magnetic field.
The magnetic field strength inside a solenoid is NOT uniform. It is weaker at the ends and strongest at the very center.
The problem is that you need the magnetic field to cease before the projectile reaches center or the forces acting on the projectile will slow it down after it passes the center.
you need to go back and re-learn what solenoid action is all about
>are these the weapons of the future?
absolutely! you can put a dent in a tin can from 5 feet away with a 30-pound weapon, and you can do that a dozen times before you even need to stand down for 2 hours to recharge your batteries. what army can afford to forgo that kind of capability?
>>1004776
>The magnetic field strength inside a solenoid is NOT uniform.
"Nearly" uniform. Which means it is in any practical sense.
>https://en.wikipedia.org/wiki/Solenoid
>http://spiff.rit.edu/classes/phys313/lectures/sol/sol_f01_long.html
>http://farside.ph.utexas.edu/teaching/302l/lectures/node76.html
>my physics notebook
>every other source you can find
>>1005079
please go back and read that stuff carefully anon, you'll find statements like
>the term refers to a coil whose length is substantially greater than its diameter
get that
>substantially greater than it's diameter
that usually means a ratio of 10:1
for a truly uniform field they talk about
>a solenoid with infinite length but finite diameter
when you get to coils like the one's we are discussing except for
>>1004619
>3rd coil wound with 5 layers 100 turns.
most of these coils will NOT have a uniform field inside, it will be close but the shorter the more distortion.
My whole point is that a solenoid pulls the armature / projectile to the CENTER of the coil, in our case you only get acceleration of the projectile for the first half of the coil's length. The current pulse needs to cut off before the projectile reaches the midpoint of the coil.
I would turn on all three coils at the start and then turn off the 1st coil, 2nd coil etc... before the projectile reaches the midpoint of each coil in turn. Such that the pulse width for coil 1 was the smallest, coil two a bit wider etc...
Coil #3's contribution to the acceleration of the projectile while the projectile is still inside coil #1 would be small but it would be a contribution nonetheless.
Any coil that is energized while the projectile is past that coil's midpoint will serve only to slow the projectile.
