Small Coilgun Project with Trigger Circuit / Electromagnetic Projectile Accelerators / Forums

Forums

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

« Previous topic | Next topic »

Small Coilgun Project with Trigger Circuit

1 2 3 next

Move Thread

LAN_403

Backyard Skunkworks

Fri Jan 25 2008, 05:37AM

Registered Member #1262 Joined: Fri Jan 25 2008, 05:22AM
Location: Maryland, USA
Posts: 451

Hey, I'm gonna build a small five stage coilgun, I'm getting five 18.75 joule 50 volt 15kuF caps from Mouser. The coils I'll wind with transformer wire and I've just designed a circuit that I hope should trigger everything on a varible delay of a few dozen ms.

Heres the circuit
Coilguntriggercircuitdo0

I'd like any input anyone has on this, this is the first real coilgun I've built (I've played with photoflash caps using a single stage one first). The goal of this is to demonstrate a working coilgun for my science program and I'm hoping to get atleast enough energy to punch through paper. I'd also like some feedback on projectiles and coil properties for decent muzzle energy.

EDIT:
Also I'm thinking of submerging the coils in Mineral oil chilled to about -25C to get lower resistance.

Shaun

Sat Jan 26 2008, 04:53AM

Registered Member #690 Joined: Tue May 08 2007, 03:47AM
Location: New Jersey, USA
Posts: 616

All I'm going to say before I can view your schematic is

1. You absolutely don't need them submerged in mineral oil. The resistance will be more than low enough; inductance is much more potentially problematic (plus resistance change at -25C would be negligible).

And 2., If I were you I'd have less stages, 3 at most. Plus, on multi-stage guns there is usually less capacitance on each extra stage because the projectile will be moving through the coil faster.

...

Sat Jan 26 2008, 07:54AM

Registered Member #56 Joined: Thu Feb 09 2006, 05:02AM
Location: Southern Califorina, USA
Posts: 2445

You will have a hard time making the gun work efficiently without some type of active feedback from the coils due to variations from shot to shot due to the projectiles location, etc.

Honestly, if all you want to do is punch through paper you could get away with a single 18j stage (I did it with a 800uf/200v = 16j )

If you want to add more stages I would recomend an optical trigger bases sytem. At the very least you will need some type of active gate drive for your FETs, because as drawn you are going to slowly ramp up the gate voltage which will cause the fet to be in its linear range when it is passing peak current--which will cause it to explode. You could probably get away with a simple comparator in series with the gate, although a real gate driver chip would be preferable.

Finally, have you done any math to see how large of a fet you are going to need? At such a lot voltage I would expect several thousand amps of peak current for any reasonable projectile speed... Not easy to find FETs, or even IGBTs for the matter, that can handle that kind of abuse

really SCRs are the way to go for this application.

Backyard Skunkworks

Sat Jan 26 2008, 10:46PM

Registered Member #1262 Joined: Fri Jan 25 2008, 05:22AM
Location: Maryland, USA
Posts: 451

Thanks a ton for the feedback on this; I've done a partial redesign.

Originally I wanted to get a hold of some liquid nitrogen and 5kJ capacitors. But I redesigned the thing to run on several dozen photoflash caps (I have a box of these cameras, get to know your local photo processor) and IGBTS however I learned how much IGBTs cost.

Then came the last redesign.

Anyway, now I've redesigned the circuit so it's timed right and replaced the FETs with large NPN transistors which trigger huge relays that close the 50v part of the circuit. I also added a second cap to the first stage and scrapped the fifth stage.
As far as the cooling I may still do it just for the "coolness factor" (pun intended) as this doesnâ€™t need to be a super efficient military weapon as much as a semi-glorified science project, it would show how real cryogenic cooling would be done even though I guess it won't do much.
I figure to get the timing thing right I'll tune the variable resistors so that later stages goes off a little before the projectile gets there, to avoid suck back (even though it loses some power that way).
At this point I got almost everything I need and I've started doing circuit prototyping for the timing.

So does this version of the thing look like it should at least shoot right?

http://img81.imageshack.us/img81/3169/newcoilguncircuitzr3.gif

Shaun

Sun Jan 27 2008, 04:28AM

Registered Member #690 Joined: Tue May 08 2007, 03:47AM
Location: New Jersey, USA
Posts: 616

Don't forget, you need diodes on the relays to protect the FETs as well as diodes on the coils to protect the capacitors.

Backyard Skunkworks

Sun Jan 27 2008, 04:41AM

Registered Member #1262 Joined: Fri Jan 25 2008, 05:22AM
Location: Maryland, USA
Posts: 451

So the diode to protect the NPN transistors (they replaced the FETs) should go between the collector and the relay?

Also is protecting the caps really necessary if the circuit stays closed as soon as it fires?

And if so what current rating do these diodes need to stand up to?

Addi

Sun Jan 27 2008, 09:24AM

Registered Member #1268 Joined: Sat Jan 26 2008, 06:12PM
Location: Middle Germany
Posts: 3

I Think, with Energy-Levels like this protection Diodes are absolutely not necessary.

If you really want to use them, they have to be quite fast and they have to stand quite a high Voltage an Current.

Why do you use this relay-delay-dircuit instead of something like an Lightgate or so ?

Shaun

Sun Jan 27 2008, 05:20PM

Registered Member #690 Joined: Tue May 08 2007, 03:47AM
Location: New Jersey, USA
Posts: 616

As soon as current stops flowing through the coil (relay coil or firing coil) it will produce a high-voltage pulse of opposite polarity which will fry your FETs and reverse-charge your capacitors.

For protecting the capacitors, put the diode across the capacitor so it is reverse-biased when the capacitor is charged normally. When the polarity is reversed, diode will short it out.

For the transistor, the diode would normally go between the collector and V+ (across the relay coil, same thing), so that it shorts out a reverse-polarity pulse. I've also seen some designs with the diode across the collector and emitter.

On the relay protection side, a single 1n4007 would be fine. The other one, across the capacitor, will see more current. On my 32J photoflash coilgun it took 4 in parallel before the reverse current stopped killing them. So 2 on each stage should do fine for you.

Backyard Skunkworks

Sun Jan 27 2008, 10:44PM

Registered Member #1262 Joined: Fri Jan 25 2008, 05:22AM
Location: Maryland, USA
Posts: 451

Being a member of the PETEC (People for the ethical treatment of electronic compnents) and seeing that Mouser has 1N4007s for only $0.04 (its as low as 1.7 cents if you buy 500) I'll just throw like six in each stage so there is no chance of anything bad happening to them (such as a small explosion).

As far as finding good mechanical relays to close the 50v side what does the continouis current rating need to be so they dont weld? I'm not seeing any spike current ratings for these. My guess is 30A ones rated for 12V would weld in this case?

Heres the newest schematic with the diodes:
http://img210.imageshack.us/img210/3673/newcoilgunschematicjs1.gif
Please double check I got things wired right to "maximize component life"

.

BTW @addi: I could have gone with optical tiggering but I decided the timing circuits would be fun to build and give flexability and given how cheap the diodes are I may as well "use protection"

.

EDIT:
Seems Mouser has IGBTs on the cheap too! Would replacing the relays with this one work?

It's rated for 180A@300V with 0.1ms spike rating of 450 amps and only $6.

Also, 20ms pulse time is about right, right?

Shaun

Mon Jan 28 2008, 03:43AM

Registered Member #690 Joined: Tue May 08 2007, 03:47AM
Location: New Jersey, USA
Posts: 616

You absolutely do not need 6 diodes on each relay, but for the capacitors it sounds like a good number.

As for for your switches, forget relays and IGBTs and get some SCRs. They will be WAY less than $6 apiece. And for only 18J per stage they wouldn't need to be big.

Try these, I've used em before to fire over 75J banks and I've never lost one:

At $2.55 each they're not too pricey; however you can get $1.30 ones from digikey that can do 200Apk, which may be good enough for only 18J per stage. I won't recommend them though because I haven't used them myself.

1 2 3 next

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