New Coilgun Project / Electromagnetic Projectile Accelerators / Forums

Forums

4hv.org :: Forums :: Electromagnetic Projectile Accelerators

« Previous topic | Next topic »

New Coilgun Project

first previous 8 9 10 11 next last

Move Thread

LAN_403

klugesmith

Fri Nov 30 2012, 01:56AM

Registered Member #2099 Joined: Wed Apr 29 2009, 12:22AM
Location: Los Altos, California
Posts: 1716

Strong work there!

Your performance table is nicely presented. The efficiency numbers are in the top tier, as far as I know. Shall we assume that all 3 stages have the same initial voltage, and that the second (but not third) automatically has the same capacitance as the first?

What physical positions (of projectile) represent the boundaries between stages, where you chose to partition the kinetic energy added? How did you determine the velocity at those points?

I wonder if magnetic coupling between the coils transfers any significant amount of energy between stages?

Yandersen

Fri Nov 30 2012, 04:27AM

Registered Member #6944 Joined: Fri Sept 28 2012, 04:54PM
Location: Canada
Posts: 340

It is a recuperational coilgun design, based on non-polar caps, which are repolarized by a coil, absorbing unused energy back. One cap handles two stages - first coil repolarizes cap once, second coil reverses cap to normal polarization. So first stage uses the fresh cap and second stage is fed with energy left from the first one. After a shot the cap is used twice and contains leftovers from the second stage. Coilgun can't be made simplier - it is just a basic LC oscillator. Due to a small energy recuperation (each stage dessipates over 50% of energy as heat during a pulse) it is irrational to put more stages on the same cap, so it was considered to separate caps by one per two stages. There are 4 caps, first one powerizes first two stages, second cap handles only third stage so far and leftovers are just dumped through a resistor after a shot. Third and forth caps are not connected yet. All caps are the same and are charged to the same voltage.
Velocity is measured by a time between triggering of two photogates, which is captured by my oscilloscope - tube with two photosensors is mounted on the end of a barrel and could be seen well on the first picture above. The two triggering signals are seen on the second picture (2700us for the photogates separated by 10cm from each other): period is measured as time between rising slopes (photogate lighted).
Chart was done simply by disabling triggering for subsequent stages and neasuring all parameters right after a shot.
Magnetic coupling is pretty significant but has no any impact because only one coil at a time shots.
The comparably high efficiency of my design is a result of not doing all the mistakes all you guys are doing from ancient times copying the same lame dummy techs from each other. The main mistake of yours is not considering the fact that only a little part of the magnetic energy could be IDEALLY absorbed by a projectile. Caging all the energy in a coil by dempher diode means it is doomed to be wasted as heat. See, in my design, where timing is tuned to be almost ideal, only around 10% of total energy happens to become a kinetic energy of a projectile. The rest of it, at least half, I save, this way doubling efficiency. But even if it would be just wasted, the efficiency of 10% is still much more than you guys typically have with your fetish dempher diode. The problem is that energy which is not used for acceleration is still inside the coil and will create suckback of almost the same magnitude as acceleration, up untill all energy dessipates as heat - only then negative effect will stop. So you are balancing between acceleration and suckback, and the difference is positive just because of a heat dessipation while projectile moves from one side of the coil to the other. If it would be superconductive, it will be just holding oscillating projectile forever. In my design, by the time projectile reaches the center of a coil, it is finishes dumping the leftovers of a magnetic field energy back to the cap, so nothing sucks a projectile back. Consider suckback as a major factor which is almost completely avoided in my design.

Yandersen

Fri Nov 30 2012, 04:50AM

Registered Member #6944 Joined: Fri Sept 28 2012, 04:54PM
Location: Canada
Posts: 340

Oh, and mention a short coils, of course. :)

By the way, let me explain why only a little part of a magnetic energy converts into kinetic energy. See, magnetic energy is proportional to the square of the current, while pull force is proportional to the current linearly after saturation, which is a normal mode of coilgun operation. So to double the kinetic energy you have to double the force, which is done by doubling the current, which means that magnetic field energy has to be increased by 4 times. Considering magfield energy to be equal to the cap's stored energy you got the ultimate answer for low efficiency of coilguns - for just doubling the speed of a projectile you need to store 16 times more energy in your cap's bank! Howhether, before saturation, proportion is quadring caps for doubling the speed, which is not that much scary...
So to work before saturation one need a huge amount of coils which could fit into portable-size gun only if there are no spaces between them (photogates, induction sensors and so on will not fit). My design makes it more possible due to the blind triggering, which, howether, requires standartized projectile, cap's capacitance and voltage and flaw-free barrel.

Yandersen

Fri Nov 30 2012, 04:59AM

Registered Member #6944 Joined: Fri Sept 28 2012, 04:54PM
Location: Canada
Posts: 340

You can try my recuperational coilgun design if you can find non-polar cap. Here is a simpliest schematic you can use to build a two-stage coilgun:

Make sure charger is disconnected before the shot, otherwise it will burn or blow out during cap's voltage inversion - that's a major drawback of the recuperational tech.

Saz43

Fri Nov 30 2012, 07:50AM

Registered Member #1525 Joined: Mon Jun 09 2008, 12:16AM
Location: America
Posts: 294

Yandersen, that's the best efficiency I've ever seen from a hobbyist coilgun. I must commend you on doing such a fantastic job, although it upsets me that you come into my thread and call my design "lame" and me a "dummy".

All of the characteristics of my coilgun are based on sound engineering and careful decision making, not because I'm unimaginative and just copied other peoples old designs. For example, I do not have a "fetish" for damping diodes and I fully understand the drawbacks to using them. I considered using diagonal half-bridges to hasten current decay and recover inductive energy left in my coils, but that necessitated the use of twice as many IGBTs and complex control circuitry, which would add weight and bulk, limiting the portability of my design.

Furthermore, the use of high voltage capacitors would require a large high voltage, high power charging circuit which would further increase the size, weight, and complexity if what is intended to be a compact portable design. It goes on- the limited power throughput of a high voltage charger would severely limit the rate of fire of my coilgun, the main purpose of which is to be fully automatic.

I instead chose to power my coils directly from Lithium Polymer batteries, giving me a theoretically unlimited rate of fire and eliminating capacitors, high voltage charging circuitry, and the associated power losses, which are significant. LiPo batteries are quite unavoidably polarized, so inductive energy recovery like your design is not possible. The inductive energy left over in the coils has to go SOMEWHERE, if I don't use diodes to commutate it, the coils will generate a voltage spike in the 10,000V range and destroy my transistors. I use 4 diodes in series for each coil to hasten the current decay, and designed my coils to be slightly longer than the projectiles to allow a "coasting zone" to minimize the effects of suckback.

I appreciate your work, but please give me some credit for the 2+ years of design work I've put into my coilgun.

P.S. I have beer cans that look like they were made of Swiss cheese after they'd been chewed up by my monster. The next beer I drink in the name of science will be dedicated to your inspiring design.

Yandersen

Fri Nov 30 2012, 08:51AM

Registered Member #6944 Joined: Fri Sept 28 2012, 04:54PM
Location: Canada
Posts: 340

Saz, please, don't let my hash words upset You! Not the first time I'm telling you that I have never seen a better quality coilgun made by anyone else. And rapid fire with 40m/s - you are one of a kind, truly. I've been dreaming about making a coilgun without caps, but just couldn't beat the halfbridge complications. Instead I come to that simple idea of oscillating gauss.
Actually, You can see that achieving 10% efficiency is possible even without recuperation. All you need is to fastly damp energy left in coil somewhere - not necessarily back in source, but at least out from the coil. Simple solution - put a zener diode or a resistor in series with dempher diode. But it will cause voltage over coil to jump for a short period of time, so make sure IGBT can handle it. It will not be hard to fit one high power zener in series with a dempher diode, right? But it may give you some additional percents of efficiency.
And I told you about the long coils. EVERY coilgun I've seen so far is burning my eyes with those inefficent long coils everyone winding and long nail projectiles as well. But there is nothing I can suggest on this by now. If you would like to build one more gauss, try short ones. Really, it works better.
High voltage is my mistake. In the beginning I've thought it will increase the efficiency, but I was wrong - voltage does not really matter. The lower the voltage, the thicker the wires should be, generaaly saying. Non-polar caps are available from 400V and I should use them instead. But, from the other hand, I prefer the coils wound with thin wires.

I'm working on my gauss for half a year. It's my first one, so I'm still learning and understanding how it all works. Believe it or not, but just only a week ago I've finally understood why subsequent stages have higher efficiency. :)

Hope my design will inspire you to build at least one more gauss - I will share with you all my knowledge and info I've collected while investigating operation of recuperational design and help you build one. With your quality build it will be Something - my hands, as you can see by the look of my monster, are not that straight. :)

P.S.: Saz, I didn't call your design lame and I really don't think you are dummy - I was saying in general. It is just a reaction on dempher diode and polar caps design I see everywhere (and yours is different) and it upsets me that most people still following that dead-end solution, get few percents of efficiency and then call gauss guns unperspective crap which doesn't have a future and doesn't worth time spent, after which they announcing that they switching to something else. It is not fair. At least you don't give up on gausses, Saz - here is some more for you to try.

Saz43

Fri Nov 30 2012, 10:23PM

Registered Member #1525 Joined: Mon Jun 09 2008, 12:16AM
Location: America
Posts: 294

Do I understand correctly that your circuit feeds the leftover energy from the first coil through the capacitor and right into the second coil? That's very smart, and I'm suprised no one has done it before (at least on 4Hv).

Using a Zener diode to quench the current works on the same principal as using multiple diodes in series, as I've done. It allows the voltage across the coil to rise to a higher point before the diodes conduct, which results in faster commutation and reduced suckback. Using a resistor in series has a similar effect, but research suggests the resistor method is not as effective. As I said, I've studied this extensively and made my decisions based on science, not precedent or ignorance. With a single diode, current in my middle coils would take about .1ms to commutate. With four in series, itâ€™s more like 0.01ms for the current to reach zero. I would add more in series to enhance the effect, but the inductive voltage spike is already nearing the 600V limit of my IGBTs.

Saying that a shorter coil is more efficienct because it produces a stronger force than a longer coil with a weaker force (if both are supplied the same energy) is like saying a tall rectangle is bigger than a short, long rectangle when both have the same area. I don't doubt your coils are highly efficient, I'm just saying it's not only because they're short.

Have we set the record for the longest, most verbose coilgun thread yet??

Yandersen

Sat Dec 01 2012, 07:16AM

Registered Member #6944 Joined: Fri Sept 28 2012, 04:54PM
Location: Canada
Posts: 340

You understood correctly - every second coil is fed from the same cap in which previous coil has discharged. BTW, as I added forth stage it gave me 31% of efficiency increasing projectile kinetic energy by 1.5J reaching 43.4m/s. Sorry, just can't hold this info! :)

Yeah, a lot of talking and not too much progress. Actually, it is only yours coilgun I'm interested to see finished. As there were no updates for a long time I've decided to inspire you a little by showing what coilguns can, in case you got out of mana. Hope it will push you a little towards finishing your coilgun - seeing things made by others have always been giving me some powers to work on mine. :)

ben5017

Sun Dec 02 2012, 08:56PM

Registered Member #3315 Joined: Thu Oct 14 2010, 04:23PM
Location:
Posts: 156

Yandersen,

I am intrigued with your re-coup solution, I would like to discuss the fesability of adding a similar version of this into my design. As to not clutter up saz thread I would like to discuss it in myown. X3CC

Yandersen

Sun Dec 09 2012, 06:33AM

Registered Member #6944 Joined: Fri Sept 28 2012, 04:54PM
Location: Canada
Posts: 340

Saz, Ben, anyone who has exp with half-bridged coilguns, what do you think, will it work if instead of IGBTs I'll try SCR and use cap as power source? As I tried out this on paper, I expect thyristors to close when voltage on cap will drop to 0 volts causing coil to charge the cap back. Am I right?
Simultaneous triggering of two SCR is not a problem (transformer with two secondaries).

Nope. After cap drops to 0V, current keeps running in two diode-thyristor loops. Maybe, extra third thyristor for V-switch in parallel to the coil will close everything up?..

first previous 8 9 10 11 next last

Moderator(s): Chris Russell, Noelle, Alex, Tesladownunder, Dave Marshall, Dave Billington, Bjørn, Steve Conner, Wolfram, Kizmo, Mads Barnkob