Yet another coil gun prototype / Electromagnetic Projectile Accelerators / Forums

Forums

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

« Previous topic | Next topic »

Yet another coil gun prototype

first previous 6 7 8 9 next

Move Thread

LAN_403

V2006

Wed Dec 12 2018, 08:35AM

Registered Member #61550 Joined: Thu Apr 06 2017, 03:23PM
Location:
Posts: 86

Great choice.

Zuckerstange47

Sat Dec 15 2018, 12:59AM

Registered Member #61926 Joined: Wed Nov 22 2017, 04:08PM
Location:
Posts: 34

I cleaned the machine up, because there were too many cables hanging about, and installed the new light barriers. I also soldered a new perfboard with MOSFET, driver, diode and zener - the previous version had an external MOSFET-driver which was responsible for lots of ugly cables on the table.
The much more stable mounting for the light barriers and coil is new, too, and the parts can finally be fixed by nuts.

Like I said before, I would like to experiment with different coil-designs before moving on. The plan is to charge the capacitors to a specific voltage, and try to get the fastest shot by varying the starting position of the projectile and of the light barrier that deactivates the coil (the first barrier moved to the other side of the coil to be able to work with longer coils, as Tomspol suggested on page 1). Also I would note the remaining voltage after the shot to bring the energy-consumption into the equation. I am not sure if this is a good approach, I am open for suggestions.

@DerAlbi: If I understood right, you indicated that my "Schutzbeschaltung" might not be the best solution. I made a sketch of my circuit containing the diode and zener. Maybe you could tell me if this is correct, or what I should change. I used the following explanation for making the circuit:

At the moment I am only testing with relatively low power, but I plan to raise it some time and then I would install multiple MOSFETs and drivers in parallel.

Photos:

Ciruit (Germon language!):

ShieldExperiments

Sat Dec 15 2018, 04:28AM

Registered Member #61689 Joined: Mon Jul 10 2017, 02:46PM
Location:
Posts: 14

Dang, you Germans always seem to make beautiful coils!

Zuckerstange, I'd reccomend just going for it with the zener + diode for a while, and run it until it burns out. That way you can fine tune your other variables while you save up for a better snubber.

That's not to say you can't get away without a proper snubber, if you're willing to sacrifice some efficiency you can even get away with just using a straight diode and a higher voltage tolerant MOSFET.

But really, I'm impressed, and listen to DerAlbi - He's the real expert here, not me, and I wish you well on your adventures into this interesting field. Just don't be surprised when I steal all of your best ideas ;)

Zuckerstange47

Sat Dec 15 2018, 01:54PM

Registered Member #61926 Joined: Wed Nov 22 2017, 04:08PM
Location:
Posts: 34

ShieldExperiments wrote ...

Dang, you Germans always seem to make beautiful coils!

Thank you very much!

DerAlbi

Sat Dec 15 2018, 05:32PM

Registered Member #2906 Joined: Sun Jun 06 2010, 02:20AM
Location: Dresden, Germany
Posts: 727

As elected King Albi of the Coilgun-Nation, i hereby declare that no coilgun shall exist with way shorter or longer coils than the projectile length; on this miserable planet.
Punishment shall be received on violation and executed by the majesty him self. A shot with no less than 2.5J to the thumb-nail must be suffered in addition to a 2 weeks arrest, strapped to the workbench where the offense was thought out.

But in seriousness, you asked for a reason the last time this was mentioned.
Go to Wolfram Alpha, and plot the function: "-x^3 * exp(-abs(x)) from -10 to 10"

This is how the force curve of a coil looks like (roughly). The left side, the force is positive, so the projectile is accelerated, at 0 (it correspondents to the the projectile center is at the coil center) there is no force, and than suckback kicks in. The time one experiences negative force, the current better already reached 0. This is not the actual force curve, but the force that you would measure if the current was constant.

If you have a too short coil, the "humps" move further apart (you have a larger area around x=0, where the force is 0, because as soon as the projectile tip exits the coil, there is pretty much no force anymore. The same is with a too long coil: as soon as the projectile is fully inside the coil, there is also no force anymore, resulting in the same force function shape.
The reason, the force-curve changes with the shape is because the force is related to dL/dx. The more the inductance changes with the projectile movement, the higher the force on the projectile.
But as soon as the projectile penetrates a too short coil completely or a projectile is completely inside a too long coil, the inductance does not change anymore, so no force.
Having the coil as long as the projectile results in the best possible force shape.

However there are positive side-effects to violating the rule:
Too short coils: they have a good material-to-force ratio (bot not better as optimal coils with projectile = coil). The timing is more forgiving because if the pulse is too long, you dont immediately run into suckback. Still, it decreases efficiency, after all you have a too long pulse, but shot performance does not suffer.
Too long coil: is the most useless design one could make in terms of material-investment (on this miserable planet). You have the same forgiving timing behavior (no immediate suckback). But all the portion of the coil that is longer than the projectile can be considered as additional resistance. Such resistance can be beneficial: on Barry's site he experimented with the (over/critical)damped LCR-circuit to shape the current waveform. This is what happens if you build a way too long coil. But you can save the copper (and weight) and use a resistor instead.

Still dont violate the rule. Remember the punishment!

Regarding the Zener:

I'd reccomend just going for it with the zener + diode for a while, and run it until it burns out.

I would fully agree. However, please dont be ignorant and at least know that you will damage your capacitor with this circuit. Electrolytics happily withstand -1.5V reverse polarity but you subject them to maybe -15 to -20V.
If you dont care about the life of your capacitor, then this is ok too - they will still last a while. But dont abuse capacitors, and later decide to put them into a bigger capacitor bank. While one failing capacitor (inner short circuit) is likely to be ok, if additional energy is delivered from the outside things can become messy.

Zuckerstange47

Sun Dec 16 2018, 04:53PM

Registered Member #61926 Joined: Wed Nov 22 2017, 04:08PM
Location:
Posts: 34

DerAlbi wrote ...

As elected King Albi of the Coilgun-Nation, i hereby declare that no coilgun shall exist with way shorter or longer coils than the projectile length; on this miserable planet.
Punishment shall be received on violation and executed by the majesty him self. A shot with no less than 2.5J to the thumb-nail must be suffered in addition to a 2 weeks arrest, strapped to the workbench where the offense was thought out.

I plotted the curve, but I do not understand, how the formula is derived (I nevertheless believe in the king's word). From my experience I was of the opinion, that a force should exist until the projectile's center of mass reaches the center of the coil, independend of its lengths.
I hope I can confirm your rules by experimentation, but unfortunatelly I am too stupid to make coils. Yesterday I tried to make the most simple coil (35 windings in length and 4 layers), but failed in many ways. The calculated length of wire differed very much from the real one and I could only take the center pipe away afterwards by heating the coil up and by destroying the pipe (even though I used grease). I probably have to do many more coils until I learn to not make as many mistakes...

In terms of Schutzbeschaltung I will have a look at snubber circuits someday, because I do not like to damage the capacitors. Until then, I will just keep using my diodes.

The graph:

DerAlbi

Sun Dec 16 2018, 08:58PM

Registered Member #2906 Joined: Sun Jun 06 2010, 02:20AM
Location: Dresden, Germany
Posts: 727

The formula is not derived from anywhere, it is the result of plotting a FEM analysis and finding a model function that the data can be curve-fitted to. The actual model function would be more complex, but this simple form conveys all the features that are interesting for now.

From my experience I was of the opinion, that a force should exist until the projectile's center of mass reaches the center of the coil, independend of its lengths.

This is technically correct. But just because there is a force, does not mean it is significant.
See, the function plot shows you the force. The force at any given moment is a description how much your projectile is accelerated at this point (under constant current). The velocity is the integral of the acceleration - the summed up acceleration for every moment. Even this is a bit more complicated, the basic thought is, that the area under the curve is an equivalent of how much work the coil can do. This area is maximal if the coil length matches the projectile length.

Do not obsess over coil quality. Even very ugly coils show basically the same electrical data as perfect coils. The failures I see in your coils happens if your setup is not rigid enough. The pressure of the layers force the sides apart, then the wire falls deeper into the valleys and an unrecoverable winding defect propagates. The more it escalates, the harder it is to wind further.

Zuckerstange47

Sun Dec 16 2018, 10:04PM

Registered Member #61926 Joined: Wed Nov 22 2017, 04:08PM
Location:
Posts: 34

I had a lot of problems winding the coil: At first I marked every meter of the wire to be able to measure the total length, but when I should about reach the 2m mark, I could not find it any more. I did not even find the 1m mark, which afore I had seen on the coil clearly, so I thought the glue had dissolved it somehow, so I started anew (later I found the 2m coincidentally). I then marked it with masking tape instead.
On the next try I suddently noticed that the starting piece of wire was pulled into the apparatus and got wound onto the coil, too, because the wire-tension was too high. Also I once handled the wrong turning wheel, which resulted in the three nuts (on the side of the green part) to unscrew. I think the rigidity should not be a big problem.
Maybe I will do another coil later today, perhaps I will encounter even more accidents then

DerAlbi

Sun Dec 16 2018, 11:12PM

Registered Member #2906 Joined: Sun Jun 06 2010, 02:20AM
Location: Dresden, Germany
Posts: 727

Why do you measure the length of the wire?
Measure the resistance of 10m of wire, (with a 5A current source and a voltmeter) and then do the same with the final coil. I dont see how wire length is critical to a coil.. its more turn- and layer count..
What is your thought behind that ?

Zuckerstange47

Wed Dec 19 2018, 02:10AM

Registered Member #61926 Joined: Wed Nov 22 2017, 04:08PM
Location:
Posts: 34

DerAlbi wrote ...

Why do you measure the length of the wire? Measure the resistance of 10m of wire, (with a 5A current source and a voltmeter) and then do the same with the final coil. I dont see how wire length is critical to a coil.. its more turn- and layer count.. What is your thought behind that ?

I probably did not put much thought into it. For comparing a long and thin to a short an thicker coil I thought it would be interesting to use the same wire length.
My second coil is finished now, here are some photos of them together with the projectile (which is about as long as the longer coil).
The making of the second coil was much easier than expected, due to my gain in experience and also because I used a device that automatically releases a portion of wire when needed from the spool (I constructed it from parts of a 3d printer and a pen). It is visible on the last image.

Also I attached the light barriers and started reprogramming the esp32.

[Edit: Continuity]

first previous 6 7 8 9 next

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