Nice design, but you are creating a "Shorted Turn" on every coil with your aluminium backplate. The backplate will work like the secondary of a transformer so you will lose heaps of energy in heating them up. Put a narrow slit it the top of the plate to break the loop.
Or better get rid of the backplate entirely. They will take about 20% of the entire length that can be used for additional coils otherwise. Instead wind your coils and then use some epoxy risin to glue the wire together to make stiff modules. Also, maybe you could use the coils itself as a projectile detector, so you won't need to do blind controll. As each next coil will be the transformer's secondary to the previous one then the current flowing through one coil will induce some current in the secondary, and the induction will grow when the iron core (that the projectile will be) comes in. You'd only need to detect the growing current in the second coil (e.g. using an 1 Mohm or so parallel resistor and trigger the switch when the voltage on the resistor reaches certain level). Just my weekend thoughts :-) Anyways, nice project :-)
Use Fiber optic strands (very small diameter) to feed light from a surface mount led to a photoreceptor. Mount the strands flush with the inside of the barrel in tiny holes that match the size of the fiber optic strand. Alternatively, just use smd components on opposite sides of the barrel and just drill the appropriate holes in the barrel material. This will give you the ability to time the sequencing of the coils. If you use the signals to turn on the coils forward of the position of the projectile, you don't need an external main controller, as the timing will be automatically ideal and matched perfectly to the acceleration of the projectile.
That's what I was going to suggest.... hole through the pads for VCC & GND, larger hole behind it in the backing plate, then a thick copper wire fed through and soldered as you stack them. If you ever need to take them apart, just cut the buss wires between each module, then unbolt...to reassemble, remove the cut pieces from all the modules snd assemble with the two new buss cables. Personally, I would want an inner diameter of the barrel more like 6-6.6mm. Unless you're going big game hunting, 7.5mm is overkill (and lost efficiency)
I came to the realisation that coilguns are not good as direct fire guns. Maybe they could be used as mortars to launch ganades. They are better at launchuing bigger projectiles (with payloads) because of saturation. The con of a coil moratar would be: - it's silent - you can set it's distance by changing it's power.
Don't dismiss the time constant of charging the inductance due to reactance. The faster the projectile the less time available to impart energy to the projectile. Too many stages is a lack of time available to accelerate the projectile. Coil gun most practical for accelerating massive projectiles at relatively slow velocities. Permanent magnet projectile might be better.
The coil design with the funnel shape was actually an good design. Your flaw was that you based you data on the projectile bieng static but I came up with and process that's uses air to push the projectile into an coil that's already on that has an similar design as yours the air chamber in my blueprint design adds kinetic force while the coil acts as an amplifier p.s ( this air launch system may work with your current design as well as having an self cooling effect in the process )
you'll want to make sure you have some kind of mechanism to rotate the projectile before you begin launching it to prevent it tumbling (it'll maintain angular momentum). I saw a video of someone who did this and it made things alot better
Nice video, enjoy watching it. As I have seen in the comments you have some slight issues in the design but I didnt read all the comments, so maybe it was already mentioned. What I have experienced with high power current bursts (although it was performed by a capacity setup) is, that the winding itself will tear appart quite fast. We used quite strong tape, but still: the coil teared itself appart when shooting. Maybe you should consider a housing or sth to give the coild stability. Another thing that might happen, is that your coil is by no means the current bottleneck. If you are unlucky it will happen, that your coils burn the isolation and even reduce the resistance and inductance of the setup. Then your current will rise dramatically and then your MOSFETs might become the bottleneck once again, but thats pure speculation. Just watch out to not corrupt the isolation of the wire while winding it. One short last thing I have seen in your last video (I think it was) is your assumption of the voltage transferring energy into the projectile. Its more the current creating the magnetic field and if you have a RLC-circuit, then you have to look at the phase shift of the current. It might even happen, that the peak of your current occurs quite close to the center of your coil and thus you get a quite good performance. I haven't calculated that to the end, but I am (kinda lazy) on it. Maybe it's an idea for you in the future, cause capacities just pack quite a stronger punch... :D Well hope this comment wasn't too annyoing to read, english isn't my mothertounge, so don't judge me please. :P Anyways, i wish you good luck in the future with the project and keep it up! Greetings from Germany!
Your sluminium plate forms a shorted turn co-axial with your coils so will have large circulating currents induced in it which will heat the plate, and damp the fast rise time you need in your coils - not what you want !
you can use a vector control, sensorless FOC for motor control is exist. just add a shunt resistor as current sensor at every coil, then measure the highest power draw, that'll be the 90° angle or the maximum torque
Becareful with deductions using simple kinematics, you will overestimate performance by a longshot. I suggest by deriving kinematic performance from taking the gradient of the potential energy of the coils (E = 1/2 LI^2). Which limits it down to the force being a function of the inductance gradient of your coils, which comes down to a high dependance on geometrical parameterization of the coils. Also, I am only here because someone linked me this video, but could you tell me briefly why you disregard capacitors while looking for a 500kW pulse? That's achievable with using your standard electrolytic capacitors. But I see you stated you wanted efficiency rather than power. Also, everything on your PCB is floating, especially your MOSFET gates, that can be fatal. You have a high EMI source right on top of it, if you do not destroy the gate instantly, you will fry the MOSFET by inducing a current on the gate, accidentally turning it on and destroying it. Be very careful with how you route the traces and place your polygons. They will absorb any magnetic field, inducing a current on the planes, some sharp corner traces will result in high dI/dt in the corners. Lastly, the largest ineffiencies in my opinion are in suck back and armature saturation. As the armature approaches center of coil, the magnetic field effect on it will increase significantly, however, if the armature is saturated, you will find that it's excreting at the point of highest force as heat. Now while the armature is leaving dead center, the only thing pulling it forward is either the next coil and/or it's momentum, but the coil it just passed is pulling it back with the same significant force I mentioned earlier by the (now) parasitic magnetic field that has yet to collapse. Problem can be solved using state observers and optimize coil sequencing. From all that, I love that people are still working on projects like this. If you have any more questions, feel free to come by my channel, I have a discord channel with other coilgun/railgun builders from UA-cam. Keep it up man!
I did not intend to calculate performance, not in the slightest. I just use the calculations to give myself a baseline for what would be absolutely optimal. Of course, I expect nothing near those numbers. I had not thought about leaving the MOSFET floating, that's a good point. I guess I can still pull the signal pin to ground or adjust the backing plate design to further distance the electronics (the PCBs are already ordered). We'll see how it goes. My hypothesis is that the discharge curve of a capacitor, being that it is a high current initially then drops off instantaneously, is the primary source of inefficiency. The coil can impart a much greater force on the projectile when the projectile is centered, but a capacitor bank cannot discharge when there is maximum potential for force without overshooting the center point (I struggle to explain it simply). That's the theory, at least. It is this hypothesis that leads me away from capacitors.
I'd be interested in joining the discord server. A coil gun has been on my to build list for a while along with one or two other projects, unfortunately money is a bit scarce for me right now. I'd love to bounce ideas with other people on the subject though.
@@LeviJanssen Yes, I have a video on my channel that explains the discharge curve of a capacitor somewhat. There are two components to the discharge curve, there rising curve and the falling curve, which are the inductive discharge and capacitive discharge respectively. The initial current peak is not instananeous, the time of peaking is determined by t_peak = pi*sqrt(C*sum{L_series})/2. There is a way to match the time of peaking with time it takes to reach coil center, but then you will not have utilized the full potential. I think using the first coil to get an initial velocity going is a good sacrifice, then dumping full capacity in the next coils a lot faster. Or just use a the battery for the first coil and the capacitors for the next ones.
I want to build a coilgun like this and I've built simple coilguns before but I don't have the technical or electrical knowledge to build one this complex. Is there any way you could only have 2 or 3 longer coils instead of say 10 shorter coils?
Is there a way to dump the flyback current into the next stage? Also, I think you're going to have thermal issues with this design. The heatsinks will become saturated from the MOSFETs and coils with nowhere to dump the heat. You'll end up with thermal runaway as the heat builds to failure.
Very well done. I also realized thinner coils were the way to go. I'm currently designing a coil gun of the same design but using ultra capacitors. You should check those out. Much larger energy densities than traditional capacitors.
Yes, I forgot to mention that in this video. I think the ideal power source is a small battery pack that charges a large super capacitor bank. The batteries I have will not last long under my loads, but a supercapacitor bank would not even flinch. Big ones can do 10,000 amps no problem, are rated for extremely high cycles, and have incredibly low ESRs for higher current. I would have opted for making this supercapacitor intermediary supply, but it just would have been too expensive.
I have made a coil using 32 gauge enameled copper wire but it does not work can somebody help fix this I have tried different batteries but still does not work do I need to perfectly coil the wires side by side? do I need to use a larger capacitor bank? please I need help
@@dabestname9624 it's really hard to tell, without you giving specifics, like type of batteries, their voltage, size of your coil, what your using for the projectile... But I'm going to say you probably need larger diameter wire (smaller numbered guage number). I'm guessing your batteries aren't pushing much voltage/current through that wire because either you have so much wound up that there is too much resistance and/or the small size of the wire will just burn out when you add a high enough current. You don't have to perfectly wind a coil to make it just work, but you will loose some efficiency. You might need bigger capacitors? It's hard for people to guess what you have
I have to say that I was under the impression that you had way more subscribers. Keep doing the great videos, improve audio and your channel will grow in no time
Wow! Thats ambitious. You and these commenters would make a great team. adding structure and efficiency is something I'm familiar with but using coils as detectors "Marcin Chojnacki" to me, is a huge alien project in itself and overloaded my brain with iterations on how to dial it in
Not sure if this idea has been proposed or how or if it would help, but possibly a double action hammer similar to a double action revolver could be used to energise the coils as well as push the projectile to "chamber it" and give it a small initial momentum to help it into the coil magnetic feilds. Depenting how you design the "hammer" you could have it sequentially activate each coil with a blade like switch
why not make one or more pre-accelerator series of coils with finer treads and lower current and a seperate controller(s)? each fine tuned to the projectile velocity that the previous series of coils achieved? its basically your first idea of a 1/x cros section coil but sliced into many pieces.
You might want to add an air gap in your aluminium as you will get inductive losses in the effective single closed coil of Al, Magnetic shielding between the coils could help too IMHO.
Alum backing plates? OK for thermal, but what about magnetic field blocking. I would use alum in the back of the mosfet, then sandwich it to a delrin plate to which the coil is mounted. I would also NOT use carbon tube for the barrel due to electrical conductivity of carbon fiber = increased losses. Love watching the project, thanks for doing it.
Yes, or course, I don’t know why I didn’t think of that. The aluminum may need to change. I don’t think the carbon fiber should be too much of a hinderance, though. The conductivity is quite a bit lower than any metal and the internal structure is largely chaotic, which could prevent eddy currents similarly to how motor stators use laminated steel sheets. I’m not sure, it will require testing.
You have to consider that eddy currents won't be created in the aluminum when the coil reaches saturation, therefore if the time the coil takes to reach saturation is much shorter than the time you intend for the coil to be energized I think the aluminum next to the coil shouldn't be a problem.
Literaly i had the same idea for my coil gun. I never knew about those batteries and with them i can make my coilgun full power. And to top it all my calculated speed for projectile is 1239km/h (2mm projectile) and if you are reading this. Could you in next coilgun video mention how you could make the projectile spin for accuracy of shooting because that is my biggest challange now. Edit: I forgot to tell you hov mouch i love your videos. Thank you for them and stay safe.
Really cool to see your progress so far, before you spend your money though there's a few things I think you need to reconsider or look at in more detail to confirm you're making the right move. Out of curiosity, what makes you think capacitors are inefficient, since you (I think?) stated a few times that's why you're using a battery? Also, if you really want efficiency, have a look at brook's coil ratios. It'll give you the maximum inductance (i.e. magnetic force) for a given length of wire (i.e. for a given copper/resistance loss). Your large diameter pancake coils probably aren't super efficient at the outer edge of the coil. Super flat coils also suffer more from weakening due to end effect. You might also want to look at magnetic saturation of your projectile since you're interested in efficiency. You want to just barely saturate the projectile, but not more, as after that point you get diminishing returns from the strength of the magnetic field of your coil. Also, wouldn't the large aluminium plates capping the front and back of each coil which is experiencing very high change in magnetic flux just lead to lots of eddy currents, weakening the magnetic field of the coil and leading to parasitic losses from resistance and eddy currents in the aluminium plates? Wouldn't a non conductive material be better? Clear acrylic or poly carbonate would probably look pretty cool, acrylic would probably also save you a decent chunk of change over the aluminium. Also, you're planning to use a ~33V battery and hoping to discharge 300A through a module, you'd need a total series impedance of around .1 ohm which I don't know is realistic. At the end of the day, coil guns are linear electric motors. Typical electric motors are very efficient, but I believe a big chunk of what makes coil guns inefficient is the demand for high acceleration in a short span of time and a short barrel. This forces you to work with very high current at short pules through an inductor and working past the magnetic saturation point of the projectile.
My hypothesis, and effectively what I'm trying to test with all this, is that the instantaneous current pulses from a capacitor discharge is the primary inefficiency. The idea is that you cannot discharge a capacitor at the point where there is greatest potential for force without over shooting the center point of the projectile and pulling it back. It is difficult to explain it simply, but I went more into it in a previous video. You're right about the aluminum plates, it was brought up before and is something I hadn't thought about. It may very well need to change. The coils I've designed should have a resistance aound 96 milliohms, and it's not uncommon for BLDC motor coils to have very low resistances like this. A coilgun certainly is a linear motor. Motors are very efficient, coilguns are dreadfully inefficient. It seems to me that the primary difference between these systems is how they are driven. A BLDC motor is driven by smooth currents approximating sine waves, no spikes or pulses. Capacitor based coilguns are strictly driven by current spikes, nothing is smooth or uniform. My goal is to power the coilgun as smoothly and consistently as possible, to power it more similarly to how you would power a motor. Hopefully we'll see some improvement.
@@LeviJanssen The suck back issue you're worried about with capacitors also exists with batteries if you don't switch off the coil in time, it's not an issue unique to capacitors. Also, in a coil gun you're unlikely to fully discharge a capacitor due to the significant current rise time in a coil (seems fast, but with a fast moving projectile in a short coil, there isn't much time), so you're not going to see the full capacitor discharge curve you're thinking about, you'll see a fairly small section of it near the beginning. Another thing is, the reason coil guns tend to operate at such high currents for such shorts amount of time (i.e. current spikes) regardless of being capacitor or battery driven is to achieve reasonable acceleration. In a short barrel like you plan is maybe 1-3 times the circumference of a hobby BLDC motor. To get one of those from 0 to max speed in 1-3 revolutions would require huge current spikes, and wouldn't work with a smooth, gentle sine wave. Even if you did generate your current in a sine wave (which wouldn't be efficient, since you want a rapid rise, to be able to make the most of the coil pulling the projectile to center, and a rapid fall, to avoid suck back) it would look like a current spike due to the high frequency it would be at in order to match the high speed of the projectile.
I have made a coil using 32 gauge enameled copper wire but it does not work can somebody help fix this I have tried different batteries but still does not work do I need to perfectly coil the wires side by side? do I need to use a larger capacitor bank? please I need help , sorry i have commented this a few times
Aluminium is paramagnetic so you might have some problems with it. Screwing the mosfet into an aluminium lump would be good for cooling considering the current you're switching, but the plates separating your coils might be better in fibreglass, abs, or carbon fibre.
While we're at it, the carbon fibre barrel isn't a very good material either, consider a rigid HDPE tube instead: low friction, non-conductive, insulating, and doesn't fail catastrophically and spray harmful particulates if it breaks.
Thanks for sharing what you have learned, I was very interested in what coil parameters are optimal in this arrangement. Still in the early stages of research before I start building some test models myself. From what I have seen so far, a many staged design is the way to go for another reason. A key limiting factor seems to be how fast a coil can be "spun up" and "down" with timing of turn off being key. I expect the field collapse and kickback is also a large source of inefficiency and hope that by having at least a few coils fully energised at any time that rather than collapse, the field will be coupled or migrate to neighboring coils and build on itself. Still leaning towards a capacitor/SCR/coil repeating module to keep costs down and allow for an allotted amount of energy to be dumped into each coil in a wave like pattern. Hopefully the time taken to discharge and build field strength allows for this overlap and the many coils lessens the need to turn off so quickly. Mechanically, I want to eliminate any spacing between coils to maximise coupling, maybe even have coils of a single wire thickness each and common ground (perhaps a conductive barrel with lengthwise cut) or have flattened "wires" exit each coil to save any long high-current runs. This will also necessitate using all the surrounding space by winding the supporting components around in a helix or radial pattern. It will get quite crowded but at least there would be no need for bus bars. Then there are only low current charge and control lines to manage.
I'm also wondering how to close the control loop, or at least get some data back for analysis. Since a supersonic projectile is still a way off, perhaps an ultrasonic rangefinder could provide some feedback as to the whereabouts of the projectile at low cost. Even if you manage to break the sound barrier, feedback during the initial acceleration might be enough to extrapolate the trajectory.
Fantastic work dude. I will say, I have had this thought about a battery as well and decided against it. The reason being is I don’t think you’re taking the current slowing property of an inductor into play as much? Albeit those solenoids do look lower turn numbers than most coil guns. When you pulse into a inductor the current rises, now it will rise dependent on the inductance, number of turns ect. But my point is the pulse won’t be front loaded with current as I might be thinking you think it does? I am design a coil gun as well but focusing on a very high capacitance bank with a very low ESR and switched via IGBTs and some careful circuit timing so that the center of mass of projectile does not pass the center of coil while it’s on ect. I find this approach quite fascinating though, especially having multiple coils on at one time, it’s quite novel and interesting. Also, I do not if you’ve even read this far.... but what are your thoughts on a hollow cylinder as a projectile? This way you could keep mass low (ie high velocity) and the ferromagnetic material would only be in the areas where the B field is largest. Just a thought, I will per sally be experimenting with this idea. Best of luck, I will follow this project closely your doing a great job.
I am assuming that the inductance of the coil will be low enough that current rise won't be too too big of a concern. If it is quite considerable, though, I think the timing of the coils can be adjusted to compensate just by leading the projectile a little. My goal is to accelerate the projectile as consistently and smoothly as possible, not pulses of energy like the discharge of a capacitor bank, so the current rise could actually help. maybe. we'll see. On the matter of hollow projectiles, I wouldn't be inclined to think is was really necessary. Coilguns are already so inefficient, I don't imagine the hollow core would do too much to help. I also think of coilguns as specializing in slow, massive projectiles. A more massive projectile means the steel is less likely to reach magnetic saturation and slower velocity means more time in the barrel which is more time to impart energy onto the projectile.
I've seen people say that the reason people have such a low efficiency is because the steel projectiles reach their maximum magnetic flux... I think that is something you have considered, but might want to look into that. Also 154 joules is insane power if it succeeds, for certain the most powerful coilgun I've seen. More power than most crossbows, absolutely deadly. Considering you are powering directly from the battery you could probably go semi-auto or full auto, at witch point the CIA shows up at your doorstep. But good luck getting it to work tho).
It is possible that max saturation is reached, but typical systems have so many other inefficiencies that I doubt thats the main cause. This can also be overcome by simply using a more massive projectile. A system like this is most efficient with a large projectile at low speeds, especially since that gives the barrel more time to impart its energy.
You're right, so why not use an iron barrel of say 2mm wall thickness as to increase the magnetic flux directly over the projectile alike to the steel plates in electric motors which increase its strength. I've also had it in my head that you Have to use a non conductive barrel for some reason, just as long as your enameled wire doesn't arc to it. Also very important, is to find pure iron rods to make the projectiles from.
have you considered capturing the flyback through the diodes to recharge the battery or a storage capacitor for reuse later? a possible way to improve overall efficiency i would have thought...
I'm thinking that the centre Carbon fibre tube should be replaced with a steel tube so the magnetic field will be stronger to push the projectile forward harder, not sure if it's the right thought though
Usually to-220 transistors have their center pin(pin 2) connected to the metal package.Connecting them to a common plate shorts them all.Also the fact that the coils are "air coils" probably decreases their efficiency alot.Maybe add some ferrite to concentrate the field.
Hi Levi, very interesting, but how will you manage the activation timing of each module during projectile acceleration ? And furthermore, how do you manage the duration of the current in each coil ? I suppose that to optimize the projectile speed the current in each coil should start flowing when the head of the projectile is aproaching che coil center, and the current should sudently stop when the projectile's mass center is aproaching the coil center.
His idea is to pre suppose the behavior and run the necessary activation based on the physical model. That said in my opinion it might still be possible to add a light sensor on the pcb he just has to figure out how to make it with smd parts.
I am definitely not knowledgeable on this topic at all but have you thought about solid-state lithium-ion batteries? I have no idea if they are being sold but its supossedly as close as it gets to have all the advantages of standard lithium-ion and super capacitor at once
you should fine tune coil diameter and lenght and power and prohectile lenght. i mean the partion of your coil which is radially farther away from the center of the barrel is used less and wastes power. . and theres the other thing which is that the outer half of the magnetic field is not used. isnt there a way to bend magnetic lines towards the center? i think that is not possible but who knows.
Why can't you just split one long multi-layer coil into smaller section-coils one after another? You could weld all of the inner coil ends to the barrel (as a common input) and all of the outer ones to separate bars, this way you could avoid having any gaps between coils thus achieving maximum efficiency. Here are some illustrations I made to show what I mean i.imgur.com/E9vgNAV.png i.imgur.com/FdmOZYz.png i.imgur.com/vintT6H.png
I use a small capacitor Bank that can output 140kW over a 2ms and this is more than enaugh for a coil. It is much saver than shorting out Lithium batteries. However i can se your point because you do not have to reload a capacitor But for testing you rarely fire that often in a row Not only because the heat buildup in the coils that can messen up your measurements. I would still suggest to up the voltage because it keeps the whole system a bit smaller and gets rid of the heating Problems... at lest for a moment. Your Design still has same flaws but if you get it working you will find out by your self. If you want help, try to contact me.
You have to insert a resistor for every diode when you plan to build parallel circuit with them because they could kill them self threw the different U/I-graph (tolerances) they could have.
@@gregthomas7328 but the problem is the ESR. They can be different and so the current could be also different. So some diodes could get more current and so they could destroy them self. If you wont use a resistor then you should use a diode that withstands much much higher currents.
Making a Coilgun - Part 5: Barrel Module Add thermal paste between electronics and heat sinks in the barrel. Regarding using projectile sensors in the barrel to trigger the next coil in sequence, would either of the below solve the sensor/firing electronics speed problem? -use the collapsing magnetic field of the coil last used to trigger the next required coil. The circuit could be connected direct between coils with a one time setting of a pot for initial calibration. -use the Hall effect of the beginning of the needle entering into the next coil to trigger it. And the same as above the circuit could be connected direct between coils with a one time setting of a pot for initial calibration. -Can the system be scaled down to accommodate 1mm sewing needles? The eye of the needle in the tail may provide aerodynamic stability.
Maybe too late to comment but did you think on providing a low-reluctance path for the magnetic circuit? probably a case around the coil to concentrate the magnetic flux on 3 sides of the coil. the inside being the projectile?
The longer the coil the more the peak field stretches out towards the end of the coil versus being concentrated in the center. These coils will be able to take the heat though no problem.
1) Use Aluminum PCB's, instead of FR4 and Aluminum heatsink. 2) Integrate the coil in the multilayer PCB itself instead of using a separate hand-wound wire coil.
Interesting video, nice to be able to follow your process! 154J is a lot of energy, like a 22 long rifle from what I can gather. Definitely enough to kill you. Better make sure to always stand behind it! :o You need a mike, dude, it sounds like recording in a box. A cheap lapel mike is like $25, would increase the production quality by heaps and bounds almost for free! The close up parts sound waay better.
@@LeviJanssen I'm sure they're not as good as the one you've got, but there's no way they're worse than the mic on your camera/phone due to proximity. I often find that I prefer a stable but low quality of audio over a changing but sometimes really high quality. Such as someone filming themselves with a lapel mic and then doing "b-roll" on a high end mic. It's jarring. Remember this; a hiccup in video will go mostly unnoticed - but a hiccup in audio is waaay noticeable. If video cuts out for three seconds - no one really cares. If that happens in audio, people click off. I hear you about forgetting - one task as you move forward is to have a process - a process where review your stuff before you commit to the effort. Hit record, tap your mic, check that audacity is picking you up and then GO. I know I'm harping on, but I really think this would make a difference - notice how Tech Ingredients ( ua-cam.com/users/TechIngredients ) often has an audio capture visibly running in the background? It's because they've realised that people come for what they have to say, as much as the visuals. And if the audio fails, they have to start over. Anyway, not trying to put you down, but instead increase your production value - the video was very interesting! Looking forward to give a like to the next one!
Sandwiching MOSFETs isn't a good idea, the aluminium plates has to have some area to give away the heat. Also the circuit boards should be definitely isolated from the plate IMO.
Does "pulsed current limit" for a MOSFET (example 800A) means that in a short duration/short pulse like a capacitor discharging in a coil can attain 800A? And does it mean that its continuous DC current limit would be the case only if the battery is passing current for 10 seconds straight without interruption ? I don't totally understand if that's only a question of duration, for the mosfet to be able to take much current.
Hey man awesome new Video was really interesting and I'm curious how the coilgun will turn out. A bit constructive criticism for your video I like that you go in detail and explain everything as we sit right next to you it`s awesome but I would maybe consider telling your plan and what you did at the beginning in a short version like "here is how my system works for dummies" then people like me that don't have that much theoretical knowledge have it easier to follow up in the rest of the video. Keep it up!
Levi: I'm interested in your conclusion about power source? I always thought that high amp/voltage caps would give more instant power? coupled with the power supply in parallel, as caps emit their power almost instantly, is this not correct? As it happens I would in my project, prefer to use a constant battery source it would save money and bulk. Would you be so kind as to give me a hint of what kind of coil and power I would need to drive a ball bearing of 30mm weighing 110 grams at a very fast speed. I'm not great at electronics but more suck it and see lol. I see that you are going for large gauge and short coil that would take a very large current, is this better than a many turn coil with lower current? In my project I was going to use 100 volts at 2 Amps with a thousand turn coil. Sorry loads of questions here. Thanks for your informative uploads
Its cool to see another update. About your choice of silicone. Im not am expert on the matter, but you mention that silicones have a high thermal conductivity. A quick search of wikipedia says that silicones in general have low thermal conductivity. Do you have a link to a particular silicone pad that you intend to use? What about a thin layer of thermal paste/grease or thermal epoxy instead? Although before settling on anything you will probably just test the specific mosfet that you ordered. No point solving problema that dont exist. Looking forward to the next update video!
Silicone pads are commonly used to thermally couple mosfets to heat sinks. There are also used as a clean replacement to thermal paste when connecting pc components to heat sinks.
@@LeviJanssen and coming from the pc building aspect of it, where i have the most experience, they are beaten consistently by even low grade thermal paste. Im sure it doesn't matter too much tho.
@@mblenczewski But CPU heatsinks and packaging aren't electrically active, MOSFET tabs are connected to the drain. They conduct during operation. Silicone pads are an industry standard for providing isolation and thermal conductivity.
@@Azagro fair enough. In the video levi mentions this, but then says that the datasheet doesnt mention this explicitly. Thats why i said that the mosfets should probably be measured first in my original post. Im a complete novice in electronics, so you are going to have to excuse my lack of knowledge :)
Impressive design. You are clearly a talented electrical and mechanical engineer with good theoretical background. Your design is also very neat and clean. But I see a lot of practical problems with your design. Firstly - those coils are going to be a PAIN to wind. Hand-wound coils won't sit neatly like that. Timing will be very difficult with all those modules especially with no optical gates. Also you should do a simulation of pulse decay time though those coils, especially the later ones. As Azagro mentioned - with your diode and mosfet configuration you will begin to encounter HEAVY suckback in your later stages. I think you'd be lucky to get 40 m/s with this system.
just found your channel and like what your doing. I'm putting a supply list together for doing some testing myself, I'm living outside the states at the moment so can only do small scale for the moment. I do this as a hobby and ignorant to the math/ formulas involved but wanted to get your thoughts on a few ideas before i waste my time with something i haven't built yet. Can reversing the polarity of the coil (at some point while the slug is in the coil) increase the speed of the projectile. second is is there any benefit in accelerating the slug prior to entering to coil? appreciate your time.
The aluminium plate will both generate heat and dempen the magnetic field, why not use glass and glye it to that, the threaded rods also needs to be plastic as not as to again dampen the magnetic field, else i really like your design and your overall good conciderations that is a good deal better than all the copy paste endeavours here on youtube, thanks for your good efforts and please make at least one of these guns, you might be able to sell it if it becomes better than all the lackluster chinese kits, it is very cheap to get pcbs made, that and material components collection and a bit of glass work for the cooling plates instead of aluminium and you would have a killer kit you could sell.....why not? I would order a bit here for little denmark for sure!!
I love watch this guys progress its honestly spectacular
That potato gun project I was planning is starting to feel a bit silly. 😋
Hey, any project is a good project
Nice design, but you are creating a "Shorted Turn" on every coil with your aluminium backplate. The backplate will work like the secondary of a transformer so you will lose heaps of energy in heating them up. Put a narrow slit it the top of the plate to break the loop.
Or better get rid of the backplate entirely. They will take about 20% of the entire length that can be used for additional coils otherwise. Instead wind your coils and then use some epoxy risin to glue the wire together to make stiff modules.
Also, maybe you could use the coils itself as a projectile detector, so you won't need to do blind controll. As each next coil will be the transformer's secondary to the previous one then the current flowing through one coil will induce some current in the secondary, and the induction will grow when the iron core (that the projectile will be) comes in. You'd only need to detect the growing current in the second coil (e.g. using an 1 Mohm or so parallel resistor and trigger the switch when the voltage on the resistor reaches certain level).
Just my weekend thoughts :-)
Anyways, nice project :-)
Use Fiber optic strands (very small diameter) to feed light from a surface mount led to a photoreceptor. Mount the strands flush with the inside of the barrel in tiny holes that match the size of the fiber optic strand. Alternatively, just use smd components on opposite sides of the barrel and just drill the appropriate holes in the barrel material. This will give you the ability to time the sequencing of the coils. If you use the signals to turn on the coils forward of the position of the projectile, you don't need an external main controller, as the timing will be automatically ideal and matched perfectly to the acceleration of the projectile.
bro we need new part
I've started (resumed) my coilgun project because your videos have me ideas. I use the HV capacitor setup instead though.
Jansson style coil gun, you will go down in history 🤙
For the bus-bars, I'd recommend using plated vias so that you can just thread a long thick cable through all of them at once.
Great video!
That's what I was going to suggest.... hole through the pads for VCC & GND, larger hole behind it in the backing plate, then a thick copper wire fed through and soldered as you stack them. If you ever need to take them apart, just cut the buss wires between each module, then unbolt...to reassemble, remove the cut pieces from all the modules snd assemble with the two new buss cables.
Personally, I would want an inner diameter of the barrel more like 6-6.6mm. Unless you're going big game hunting, 7.5mm is overkill (and lost efficiency)
I came to the realisation that coilguns are not good as direct fire guns. Maybe they could be used as mortars to launch ganades.
They are better at launchuing bigger projectiles (with payloads) because of saturation.
The con of a coil moratar would be:
- it's silent
- you can set it's distance by changing it's power.
Don't dismiss the time constant of charging the inductance due to reactance. The faster the projectile the less time available to impart energy to the projectile. Too many stages is a lack of time available to accelerate the projectile. Coil gun most practical for accelerating massive projectiles at relatively slow velocities. Permanent magnet projectile might be better.
I have always wanted to build a flywheel energy storage device to power a small rail gun and this is making me want to get back in to that project.
Very excited to see how this project progresses. Great work!
the heat sink of n-channel MOSFETs is usually (maybe always) connected to the Drain input
Imagine having a stage in the middle fail and having to undo all of those nuts just to get to it.
The coil design with the funnel shape was actually an good design. Your flaw was that you based you data on the projectile bieng static but I came up with and process that's uses air to push the projectile into an coil that's already on that has an similar design as yours the air chamber in my blueprint design adds kinetic force while the coil acts as an amplifier p.s ( this air launch system may work with your current design as well as having an self cooling effect in the process )
you'll want to make sure you have some kind of mechanism to rotate the projectile before you begin launching it to prevent it tumbling (it'll maintain angular momentum). I saw a video of someone who did this and it made things alot better
Like rifling in a barrel does?
Nice video, enjoy watching it. As I have seen in the comments you have some slight issues in the design but I didnt read all the comments, so maybe it was already mentioned. What I have experienced with high power current bursts (although it was performed by a capacity setup) is, that the winding itself will tear appart quite fast. We used quite strong tape, but still: the coil teared itself appart when shooting. Maybe you should consider a housing or sth to give the coild stability. Another thing that might happen, is that your coil is by no means the current bottleneck. If you are unlucky it will happen, that your coils burn the isolation and even reduce the resistance and inductance of the setup. Then your current will rise dramatically and then your MOSFETs might become the bottleneck once again, but thats pure speculation. Just watch out to not corrupt the isolation of the wire while winding it.
One short last thing I have seen in your last video (I think it was) is your assumption of the voltage transferring energy into the projectile. Its more the current creating the magnetic field and if you have a RLC-circuit, then you have to look at the phase shift of the current. It might even happen, that the peak of your current occurs quite close to the center of your coil and thus you get a quite good performance. I haven't calculated that to the end, but I am (kinda lazy) on it. Maybe it's an idea for you in the future, cause capacities just pack quite a stronger punch... :D
Well hope this comment wasn't too annyoing to read, english isn't my mothertounge, so don't judge me please. :P Anyways, i wish you good luck in the future with the project and keep it up!
Greetings from Germany!
Your sluminium plate forms a shorted turn co-axial with your coils so will have large circulating currents induced in it which will heat the plate, and damp the fast rise time you need in your coils - not what you want !
you can use a vector control, sensorless FOC for motor control is exist. just add a shunt resistor as current sensor at every coil, then measure the highest power draw, that'll be the 90° angle or the maximum torque
You should design a threaded barrel for projectile stability.
this is seriously impressive
Becareful with deductions using simple kinematics, you will overestimate performance by a longshot. I suggest by deriving kinematic performance from taking the gradient of the potential energy of the coils (E = 1/2 LI^2). Which limits it down to the force being a function of the inductance gradient of your coils, which comes down to a high dependance on geometrical parameterization of the coils. Also, I am only here because someone linked me this video, but could you tell me briefly why you disregard capacitors while looking for a 500kW pulse? That's achievable with using your standard electrolytic capacitors. But I see you stated you wanted efficiency rather than power.
Also, everything on your PCB is floating, especially your MOSFET gates, that can be fatal. You have a high EMI source right on top of it, if you do not destroy the gate instantly, you will fry the MOSFET by inducing a current on the gate, accidentally turning it on and destroying it. Be very careful with how you route the traces and place your polygons. They will absorb any magnetic field, inducing a current on the planes, some sharp corner traces will result in high dI/dt in the corners.
Lastly, the largest ineffiencies in my opinion are in suck back and armature saturation. As the armature approaches center of coil, the magnetic field effect on it will increase significantly, however, if the armature is saturated, you will find that it's excreting at the point of highest force as heat. Now while the armature is leaving dead center, the only thing pulling it forward is either the next coil and/or it's momentum, but the coil it just passed is pulling it back with the same significant force I mentioned earlier by the (now) parasitic magnetic field that has yet to collapse. Problem can be solved using state observers and optimize coil sequencing.
From all that, I love that people are still working on projects like this. If you have any more questions, feel free to come by my channel, I have a discord channel with other coilgun/railgun builders from UA-cam. Keep it up man!
I did not intend to calculate performance, not in the slightest. I just use the calculations to give myself a baseline for what would be absolutely optimal. Of course, I expect nothing near those numbers. I had not thought about leaving the MOSFET floating, that's a good point. I guess I can still pull the signal pin to ground or adjust the backing plate design to further distance the electronics (the PCBs are already ordered). We'll see how it goes. My hypothesis is that the discharge curve of a capacitor, being that it is a high current initially then drops off instantaneously, is the primary source of inefficiency. The coil can impart a much greater force on the projectile when the projectile is centered, but a capacitor bank cannot discharge when there is maximum potential for force without overshooting the center point (I struggle to explain it simply). That's the theory, at least. It is this hypothesis that leads me away from capacitors.
I'd be interested in joining the discord server. A coil gun has been on my to build list for a while along with one or two other projects, unfortunately money is a bit scarce for me right now. I'd love to bounce ideas with other people on the subject though.
@@LeviJanssen Yes, I have a video on my channel that explains the discharge curve of a capacitor somewhat. There are two components to the discharge curve, there rising curve and the falling curve, which are the inductive discharge and capacitive discharge respectively. The initial current peak is not instananeous, the time of peaking is determined by t_peak = pi*sqrt(C*sum{L_series})/2. There is a way to match the time of peaking with time it takes to reach coil center, but then you will not have utilized the full potential. I think using the first coil to get an initial velocity going is a good sacrifice, then dumping full capacity in the next coils a lot faster. Or just use a the battery for the first coil and the capacitors for the next ones.
@@H34... I think I have a link on my channel somewhere.
Could I get a link to that discord?
I want to build a coilgun like this and I've built simple coilguns before but I don't have the technical or electrical knowledge to build one this complex. Is there any way you could only have 2 or 3 longer coils instead of say 10 shorter coils?
Is there a way to dump the flyback current into the next stage?
Also, I think you're going to have thermal issues with this design. The heatsinks will become saturated from the MOSFETs and coils with nowhere to dump the heat. You'll end up with thermal runaway as the heat builds to failure.
Very well done. I also realized thinner coils were the way to go. I'm currently designing a coil gun of the same design but using ultra capacitors. You should check those out. Much larger energy densities than traditional capacitors.
Yes, I forgot to mention that in this video. I think the ideal power source is a small battery pack that charges a large super capacitor bank. The batteries I have will not last long under my loads, but a supercapacitor bank would not even flinch. Big ones can do 10,000 amps no problem, are rated for extremely high cycles, and have incredibly low ESRs for higher current. I would have opted for making this supercapacitor intermediary supply, but it just would have been too expensive.
I have made a coil using 32 gauge enameled copper wire but it does not work can somebody help fix this I have tried different batteries but still does not work do I need to perfectly coil the wires side by side? do I need to use a larger capacitor bank? please I need help
@@dabestname9624 it's really hard to tell, without you giving specifics, like type of batteries, their voltage, size of your coil, what your using for the projectile... But I'm going to say you probably need larger diameter wire (smaller numbered guage number). I'm guessing your batteries aren't pushing much voltage/current through that wire because either you have so much wound up that there is too much resistance and/or the small size of the wire will just burn out when you add a high enough current.
You don't have to perfectly wind a coil to make it just work, but you will loose some efficiency. You might need bigger capacitors? It's hard for people to guess what you have
I have to say that I was under the impression that you had way more subscribers. Keep doing the great videos, improve audio and your channel will grow in no time
this is so optimized
Wow! Thats ambitious. You and these commenters would make a great team. adding structure and efficiency is something I'm familiar with but using coils as detectors "Marcin Chojnacki" to me, is a huge alien project in itself and overloaded my brain with iterations on how to dial it in
You could also do a grated barrel cover for more heat sinking
This is incredible!
A very precise planing. I found no better one´s. But I miss the detector for the projectile that switches on the next stage.
Damn dude your dedication and focus is insane, that design is clean! I've been designing one for years on and off but this is just satisfying
Not sure if this idea has been proposed or how or if it would help, but possibly a double action hammer similar to a double action revolver could be used to energise the coils as well as push the projectile to "chamber it" and give it a small initial momentum to help it into the coil magnetic feilds. Depenting how you design the "hammer" you could have it sequentially activate each coil with a blade like switch
So did you arrive at a completed product and demonstrate it? Is there a part 6?
why not make one or more pre-accelerator series of coils with finer treads and lower current and a seperate controller(s)?
each fine tuned to the projectile velocity that the previous series of coils achieved?
its basically your first idea of a 1/x cros section coil but sliced into many pieces.
You might want to add an air gap in your aluminium as you will get inductive losses in the effective single closed coil of Al, Magnetic shielding between the coils could help too IMHO.
Alum backing plates? OK for thermal, but what about magnetic field blocking. I would use alum in the back of the mosfet, then sandwich it to a delrin plate to which the coil is mounted. I would also NOT use carbon tube for the barrel due to electrical conductivity of carbon fiber = increased losses. Love watching the project, thanks for doing it.
Yes, or course, I don’t know why I didn’t think of that. The aluminum may need to change. I don’t think the carbon fiber should be too much of a hinderance, though. The conductivity is quite a bit lower than any metal and the internal structure is largely chaotic, which could prevent eddy currents similarly to how motor stators use laminated steel sheets. I’m not sure, it will require testing.
You have to consider that eddy currents won't be created in the aluminum when the coil reaches saturation, therefore if the time the coil takes to reach saturation is much shorter than the time you intend for the coil to be energized I think the aluminum next to the coil shouldn't be a problem.
There exists a simple circuit that will force the discharge rate to increase. I suggest using capacitors. Also look into silicon anodes in the lib...
Literaly i had the same idea for my coil gun. I never knew about those batteries and with them i can make my coilgun full power. And to top it all my calculated speed for projectile is 1239km/h (2mm projectile) and if you are reading this. Could you in next coilgun video mention how you could make the projectile spin for accuracy of shooting because that is my biggest challange now.
Edit: I forgot to tell you hov mouch i love your videos. Thank you for them and stay safe.
Really cool to see your progress so far, before you spend your money though there's a few things I think you need to reconsider or look at in more detail to confirm you're making the right move.
Out of curiosity, what makes you think capacitors are inefficient, since you (I think?) stated a few times that's why you're using a battery?
Also, if you really want efficiency, have a look at brook's coil ratios. It'll give you the maximum inductance (i.e. magnetic force) for a given length of wire (i.e. for a given copper/resistance loss). Your large diameter pancake coils probably aren't super efficient at the outer edge of the coil. Super flat coils also suffer more from weakening due to end effect. You might also want to look at magnetic saturation of your projectile since you're interested in efficiency. You want to just barely saturate the projectile, but not more, as after that point you get diminishing returns from the strength of the magnetic field of your coil.
Also, wouldn't the large aluminium plates capping the front and back of each coil which is experiencing very high change in magnetic flux just lead to lots of eddy currents, weakening the magnetic field of the coil and leading to parasitic losses from resistance and eddy currents in the aluminium plates? Wouldn't a non conductive material be better? Clear acrylic or poly carbonate would probably look pretty cool, acrylic would probably also save you a decent chunk of change over the aluminium.
Also, you're planning to use a ~33V battery and hoping to discharge 300A through a module, you'd need a total series impedance of around .1 ohm which I don't know is realistic.
At the end of the day, coil guns are linear electric motors. Typical electric motors are very efficient, but I believe a big chunk of what makes coil guns inefficient is the demand for high acceleration in a short span of time and a short barrel. This forces you to work with very high current at short pules through an inductor and working past the magnetic saturation point of the projectile.
My hypothesis, and effectively what I'm trying to test with all this, is that the instantaneous current pulses from a capacitor discharge is the primary inefficiency. The idea is that you cannot discharge a capacitor at the point where there is greatest potential for force without over shooting the center point of the projectile and pulling it back. It is difficult to explain it simply, but I went more into it in a previous video. You're right about the aluminum plates, it was brought up before and is something I hadn't thought about. It may very well need to change. The coils I've designed should have a resistance aound 96 milliohms, and it's not uncommon for BLDC motor coils to have very low resistances like this. A coilgun certainly is a linear motor. Motors are very efficient, coilguns are dreadfully inefficient. It seems to me that the primary difference between these systems is how they are driven. A BLDC motor is driven by smooth currents approximating sine waves, no spikes or pulses. Capacitor based coilguns are strictly driven by current spikes, nothing is smooth or uniform. My goal is to power the coilgun as smoothly and consistently as possible, to power it more similarly to how you would power a motor. Hopefully we'll see some improvement.
@@LeviJanssen The suck back issue you're worried about with capacitors also exists with batteries if you don't switch off the coil in time, it's not an issue unique to capacitors.
Also, in a coil gun you're unlikely to fully discharge a capacitor due to the significant current rise time in a coil (seems fast, but with a fast moving projectile in a short coil, there isn't much time), so you're not going to see the full capacitor discharge curve you're thinking about, you'll see a fairly small section of it near the beginning.
Another thing is, the reason coil guns tend to operate at such high currents for such shorts amount of time (i.e. current spikes) regardless of being capacitor or battery driven is to achieve reasonable acceleration. In a short barrel like you plan is maybe 1-3 times the circumference of a hobby BLDC motor. To get one of those from 0 to max speed in 1-3 revolutions would require huge current spikes, and wouldn't work with a smooth, gentle sine wave. Even if you did generate your current in a sine wave (which wouldn't be efficient, since you want a rapid rise, to be able to make the most of the coil pulling the projectile to center, and a rapid fall, to avoid suck back) it would look like a current spike due to the high frequency it would be at in order to match the high speed of the projectile.
I have made a coil using 32 gauge enameled copper wire but it does not work can somebody help fix this I have tried different batteries but still does not work do I need to perfectly coil the wires side by side? do I need to use a larger capacitor bank? please I need help
, sorry i have commented this a few times
waitnig for new video
Aluminium is paramagnetic so you might have some problems with it. Screwing the mosfet into an aluminium lump would be good for cooling considering the current you're switching, but the plates separating your coils might be better in fibreglass, abs, or carbon fibre.
While we're at it, the carbon fibre barrel isn't a very good material either, consider a rigid HDPE tube instead: low friction, non-conductive, insulating, and doesn't fail catastrophically and spray harmful particulates if it breaks.
Thanks for sharing what you have learned, I was very interested in what coil parameters are optimal in this arrangement. Still in the early stages of research before I start building some test models myself.
From what I have seen so far, a many staged design is the way to go for another reason. A key limiting factor seems to be how fast a coil can be "spun up" and "down" with timing of turn off being key.
I expect the field collapse and kickback is also a large source of inefficiency and hope that by having at least a few coils fully energised at any time that rather than collapse, the field will be coupled or migrate to neighboring coils and build on itself.
Still leaning towards a capacitor/SCR/coil repeating module to keep costs down and allow for an allotted amount of energy to be dumped into each coil in a wave like pattern. Hopefully the time taken to discharge and build field strength allows for this overlap and the many coils lessens the need to turn off so quickly.
Mechanically, I want to eliminate any spacing between coils to maximise coupling, maybe even have coils of a single wire thickness each and common ground (perhaps a conductive barrel with lengthwise cut) or have flattened "wires" exit each coil to save any long high-current runs.
This will also necessitate using all the surrounding space by winding the supporting components around in a helix or radial pattern. It will get quite crowded but at least there would be no need for bus bars. Then there are only low current charge and control lines to manage.
I'm also wondering how to close the control loop, or at least get some data back for analysis. Since a supersonic projectile is still a way off, perhaps an ultrasonic rangefinder could provide some feedback as to the whereabouts of the projectile at low cost. Even if you manage to break the sound barrier, feedback during the initial acceleration might be enough to extrapolate the trajectory.
Thanks you give me an awesome set of ideas to consider for my design I am working on.
Great video!
Fantastic work dude. I will say, I have had this thought about a battery as well and decided against it. The reason being is I don’t think you’re taking the current slowing property of an inductor into play as much? Albeit those solenoids do look lower turn numbers than most coil guns. When you pulse into a inductor the current rises, now it will rise dependent on the inductance, number of turns ect. But my point is the pulse won’t be front loaded with current as I might be thinking you think it does? I am design a coil gun as well but focusing on a very high capacitance bank with a very low ESR and switched via IGBTs and some careful circuit timing so that the center of mass of projectile does not pass the center of coil while it’s on ect. I find this approach quite fascinating though, especially having multiple coils on at one time, it’s quite novel and interesting. Also, I do not if you’ve even read this far.... but what are your thoughts on a hollow cylinder as a projectile? This way you could keep mass low (ie high velocity) and the ferromagnetic material would only be in the areas where the B field is largest. Just a thought, I will per sally be experimenting with this idea. Best of luck, I will follow this project closely your doing a great job.
I am assuming that the inductance of the coil will be low enough that current rise won't be too too big of a concern. If it is quite considerable, though, I think the timing of the coils can be adjusted to compensate just by leading the projectile a little. My goal is to accelerate the projectile as consistently and smoothly as possible, not pulses of energy like the discharge of a capacitor bank, so the current rise could actually help. maybe. we'll see. On the matter of hollow projectiles, I wouldn't be inclined to think is was really necessary. Coilguns are already so inefficient, I don't imagine the hollow core would do too much to help. I also think of coilguns as specializing in slow, massive projectiles. A more massive projectile means the steel is less likely to reach magnetic saturation and slower velocity means more time in the barrel which is more time to impart energy onto the projectile.
As far as I know hollow tube like projectiles fly like crap
I've seen people say that the reason people have such a low efficiency is because the steel projectiles reach their maximum magnetic flux... I think that is something you have considered, but might want to look into that.
Also 154 joules is insane power if it succeeds, for certain the most powerful coilgun I've seen. More power than most crossbows, absolutely deadly. Considering you are powering directly from the battery you could probably go semi-auto or full auto, at witch point the CIA shows up at your doorstep. But good luck getting it to work tho).
It is possible that max saturation is reached, but typical systems have so many other inefficiencies that I doubt thats the main cause. This can also be overcome by simply using a more massive projectile. A system like this is most efficient with a large projectile at low speeds, especially since that gives the barrel more time to impart its energy.
You're right, so why not use an iron barrel of say 2mm wall thickness as to increase the magnetic flux directly over the projectile alike to the steel plates in electric motors which increase its strength. I've also had it in my head that you Have to use a non conductive barrel for some reason, just as long as your enameled wire doesn't arc to it. Also very important, is to find pure iron rods to make the projectiles from.
You work is amazing keep making such fascinating content
have you considered capturing the flyback through the diodes to recharge the battery or a storage capacitor for reuse later? a possible way to improve overall efficiency i would have thought...
I was thinking the same, But i don't know if it retain the charge, i think a double switching it's needed
I'm thinking that the centre Carbon fibre tube should be replaced with a steel tube so the magnetic field will be stronger to push the projectile forward harder, not sure if it's the right thought though
Why not conduct the power through the lag bolts?
Kapton tape would be a good insulator to have behind the pcbs
Usually to-220 transistors have their center pin(pin 2) connected to the metal package.Connecting them to a common plate shorts them all.Also the fact that the coils are "air coils" probably decreases their efficiency alot.Maybe add some ferrite to concentrate the field.
Hi Levi, very interesting, but how will you manage the activation timing of each module during projectile acceleration ?
And furthermore, how do you manage the duration of the current in each coil ?
I suppose that to optimize the projectile speed the current in each coil should start flowing when the head of the projectile is aproaching che coil center, and the current should sudently stop when the projectile's mass center is aproaching the coil center.
His idea is to pre suppose the behavior and run the necessary activation based on the physical model. That said in my opinion it might still be possible to add a light sensor on the pcb he just has to figure out how to make it with smd parts.
Any update on this project? Did you ever test the new coil design?
You should write a paper with the parameters that you have taken, the pros and counters of your experimental desing.
Really nice video! Looking forward for the next video (maybe the realization?😅)
You know with the TO package you can pull your load through the back plate, yes?
I am definitely not knowledgeable on this topic at all but have you thought about solid-state lithium-ion batteries? I have no idea if they are being sold but its supossedly as close as it gets to have all the advantages of standard lithium-ion and super capacitor at once
you should fine tune coil diameter and lenght and power and prohectile lenght.
i mean the partion of your coil which is radially farther away from the center of the barrel is used less and wastes power. .
and theres the other thing which is that the outer half of the magnetic field is not used.
isnt there a way to bend magnetic lines towards the center? i think that is not possible but who knows.
Why can't you just split one long multi-layer coil into smaller section-coils one after another? You could weld all of the inner coil ends to the barrel (as a common input) and all of the outer ones to separate bars, this way you could avoid having any gaps between coils thus achieving maximum efficiency.
Here are some illustrations I made to show what I mean
i.imgur.com/E9vgNAV.png
i.imgur.com/FdmOZYz.png
i.imgur.com/vintT6H.png
Intresting
bad ass, thanks for the knowledge
That's awesome, do you think you could adjust and arrange the software to work out the amount of turns... For a Brushless motor?
I use a small capacitor Bank that can output 140kW over a 2ms and this is more than enaugh for a coil. It is much saver than shorting out Lithium batteries.
However i can se your point because you do not have to reload a capacitor But for testing you rarely fire that often in a row Not only because the heat buildup in the coils that can messen up your measurements.
I would still suggest to up the voltage because it keeps the whole system a bit smaller and gets rid of the heating Problems... at lest for a moment. Your Design still has same flaws but if you get it working you will find out by your self.
If you want help, try to contact me.
You have to insert a resistor for every diode when you plan to build parallel circuit with them because they could kill them self threw the different U/I-graph (tolerances) they could have.
Not at these high currents where ESR dominates voltage drop.
@@gregthomas7328 but the problem is the ESR. They can be different and so the current could be also different. So some diodes could get more current and so they could destroy them self. If you wont use a resistor then you should use a diode that withstands much much higher currents.
Wouldn't a glass plate or carbine fiber be better help obsorb more heat to
Making a Coilgun - Part 5: Barrel Module
Add thermal paste between electronics and heat sinks in the barrel.
Regarding using projectile sensors in the barrel to trigger the next coil in sequence, would either of the below solve the sensor/firing electronics speed problem?
-use the collapsing magnetic field of the coil last used to trigger the next required coil. The circuit could be connected direct between coils with a one time setting of a pot for initial calibration.
-use the Hall effect of the beginning of the needle entering into the next coil to trigger it. And the same as above the circuit could be connected direct between coils with a one time setting of a pot for initial calibration.
-Can the system be scaled down to accommodate 1mm sewing needles?
The eye of the needle in the tail may provide aerodynamic stability.
tuning the triggering will be a nightmare. I saw some guy using light curtains for it
what about making it fully enclosed liquid sealed and add a nonconductive liquid like mineral oil for more cooling
Maybe too late to comment but did you think on providing a low-reluctance path for the magnetic circuit? probably a case around the coil to concentrate the magnetic flux on 3 sides of the coil. the inside being the projectile?
This is very interesting. Will there be an update anytime soon?
The longer the coil the more the peak field stretches out towards the end of the coil versus being concentrated in the center. These coils will be able to take the heat though no problem.
Why not fitting all the module entirely on a pcb instead of the metal plate?
1) Use Aluminum PCB's, instead of FR4 and Aluminum heatsink. 2) Integrate the coil in the multilayer PCB itself instead of using a separate hand-wound wire coil.
Any update on this project?
The back plate will cause eddy current and low the efficiency. Be careful.
why did you stop working on this project, had hoped to see a final prototype or a working model
I don’t know if it would eat away at the enamel, but you could try super gluing the coil in place
Hi, I have enjoyed watching all of these videos. Are you currently working on this design still?
Interesting video, nice to be able to follow your process! 154J is a lot of energy, like a 22 long rifle from what I can gather. Definitely enough to kill you. Better make sure to always stand behind it! :o
You need a mike, dude, it sounds like recording in a box. A cheap lapel mike is like $25, would increase the production quality by heaps and bounds almost for free! The close up parts sound waay better.
I always assumed those cheap lapels would have terrible audio. I actually have access to a high quality one, I just forget to use it.
@@LeviJanssen I'm sure they're not as good as the one you've got, but there's no way they're worse than the mic on your camera/phone due to proximity. I often find that I prefer a stable but low quality of audio over a changing but sometimes really high quality. Such as someone filming themselves with a lapel mic and then doing "b-roll" on a high end mic. It's jarring.
Remember this; a hiccup in video will go mostly unnoticed - but a hiccup in audio is waaay noticeable. If video cuts out for three seconds - no one really cares. If that happens in audio, people click off.
I hear you about forgetting - one task as you move forward is to have a process - a process where review your stuff before you commit to the effort. Hit record, tap your mic, check that audacity is picking you up and then GO.
I know I'm harping on, but I really think this would make a difference - notice how Tech Ingredients ( ua-cam.com/users/TechIngredients ) often has an audio capture visibly running in the background? It's because they've realised that people come for what they have to say, as much as the visuals. And if the audio fails, they have to start over.
Anyway, not trying to put you down, but instead increase your production value - the video was very interesting! Looking forward to give a like to the next one!
looks like if given some geometery favoring penatration, you could have yourself an expensive varmint rifle with more power than a .22short
How is this going?
so how does it decide when to turn on and off the coils? just guess and tune the timing?
yeah pretty much
Sandwiching MOSFETs isn't a good idea, the aluminium plates has to have some area to give away the heat. Also the circuit boards should be definitely isolated from the plate IMO.
utsource is another source of electronic parts.
Why do all coilgun designs use ferromagnetic projectiles rather than permanent magnets?
Cost
A bit oftopic if you would set your exposure settings to a fixed state the picture wouldn't change in lighting, if you move for example your hand.
Take an ohm meter and measure between the gate and heat sink portion. Try with each pin, never see these not connected to something.
Does "pulsed current limit" for a MOSFET (example 800A) means that in a short duration/short pulse like a capacitor discharging in a coil can attain 800A?
And does it mean that its continuous DC current limit would be the case only if the battery is passing current for 10 seconds straight without interruption ?
I don't totally understand if that's only a question of duration, for the mosfet to be able to take much current.
Thank you for sharing this very interesting research anyway !
Hey man awesome new Video was really interesting and I'm curious how the coilgun will turn out.
A bit constructive criticism for your video I like that you go in detail and explain everything as we sit right next to you it`s awesome but I would maybe consider telling your plan and what you did at the beginning in a short version like "here is how my system works for dummies" then people like me that don't have that much theoretical knowledge have it easier to follow up in the rest of the video.
Keep it up!
Why has there been no video since this one? ❤
12:12 ‘’It’s good shite’’
You better open the aluminium plate, cause the changing magnetic field wil induce a current into it agaist the growing magnetic field....
Levi: I'm interested in your conclusion about power source? I always thought that high amp/voltage caps would give more instant power? coupled with the power supply in parallel, as caps emit their power almost instantly, is this not correct? As it happens I would in my project, prefer to use a constant battery source it would save money and bulk. Would you be so kind as to give me a hint of what kind of coil and power I would need to drive a ball bearing of 30mm weighing 110 grams at a very fast speed. I'm not great at electronics but more suck it and see lol. I see that you are going for large gauge and short coil that would take a very large current, is this better than a many turn coil with lower current? In my project I was going to use 100 volts at 2 Amps with a thousand turn coil. Sorry loads of questions here. Thanks for your informative uploads
Its cool to see another update. About your choice of silicone. Im not am expert on the matter, but you mention that silicones have a high thermal conductivity. A quick search of wikipedia says that silicones in general have low thermal conductivity. Do you have a link to a particular silicone pad that you intend to use? What about a thin layer of thermal paste/grease or thermal epoxy instead? Although before settling on anything you will probably just test the specific mosfet that you ordered. No point solving problema that dont exist. Looking forward to the next update video!
Silicone pads are commonly used to thermally couple mosfets to heat sinks. There are also used as a clean replacement to thermal paste when connecting pc components to heat sinks.
@@LeviJanssen and coming from the pc building aspect of it, where i have the most experience, they are beaten consistently by even low grade thermal paste. Im sure it doesn't matter too much tho.
@@mblenczewski But CPU heatsinks and packaging aren't electrically active, MOSFET tabs are connected to the drain. They conduct during operation. Silicone pads are an industry standard for providing isolation and thermal conductivity.
@@Azagro fair enough. In the video levi mentions this, but then says that the datasheet doesnt mention this explicitly. Thats why i said that the mosfets should probably be measured first in my original post. Im a complete novice in electronics, so you are going to have to excuse my lack of knowledge :)
Great depth the design approach, Subbed! Point of Rail G history; Was Tesla's several million volts dumped through a cascade of H2 sparkgaps? Thanks
Respectable amount of energy
As in, you better respect it💀
Impressive design. You are clearly a talented electrical and mechanical engineer with good theoretical background. Your design is also very neat and clean. But I see a lot of practical problems with your design.
Firstly - those coils are going to be a PAIN to wind. Hand-wound coils won't sit neatly like that. Timing will be very difficult with all those modules especially with no optical gates. Also you should do a simulation of pulse decay time though those coils, especially the later ones. As Azagro mentioned - with your diode and mosfet configuration you will begin to encounter HEAVY suckback in your later stages. I think you'd be lucky to get 40 m/s with this system.
just found your channel and like what your doing. I'm putting a supply list together for doing some testing myself, I'm living outside the states at the moment so can only do small scale for the moment. I do this as a hobby and ignorant to the math/ formulas involved but wanted to get your thoughts on a few ideas before i waste my time with something i haven't built yet. Can reversing the polarity of the coil (at some point while the slug is in the coil) increase the speed of the projectile. second is is there any benefit in accelerating the slug prior to entering to coil? appreciate your time.
Your audio is super quiet.
The aluminium plate will both generate heat and dempen the magnetic field, why not use glass and glye it to that, the threaded rods also needs to be plastic as not as to again dampen the magnetic field, else i really like your design and your overall good conciderations that is a good deal better than all the copy paste endeavours here on youtube, thanks for your good efforts and please make at least one of these guns, you might be able to sell it if it becomes better than all the lackluster chinese kits, it is very cheap to get pcbs made, that and material components collection and a bit of glass work for the cooling plates instead of aluminium and you would have a killer kit you could sell.....why not?
I would order a bit here for little denmark for sure!!