Single stage halfbridge coilgun / Electromagnetic Projectile Accelerators / Forums

Forums

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

« Previous topic | Next topic »

Single stage halfbridge coilgun

Move Thread

LAN_403

Lekky1

Mon Dec 22 2008, 10:42AM

Registered Member #1592 Joined: Thu Jul 17 2008, 06:58AM
Location:
Posts: 11

Hi everybody!

I have been studying these pages for some time & have finally got a project to post. Electronics is a hobby of mine, so please excuse me if I don't use all the correct terminology.

After studying other peoples coil guns (thank you all), I decided to use half bridge switching because I liked the concept.

My first version used two FGL60N100BNTD igbt's per side. These are 1000v devices, rated for a peak current of 200 amps.
I'm using an IR2113 halfbridge driver to switch the igbt's & RHRG75120 diodes to redirect the back emf from the coil.

Capacitors are a bank of 16 x 120uf flash capacitors.

The boost converter is a circuit I found on the internet. It's based on a uc3843.

Working voltage is 300V.

The projectile is a 4.5mm ball bearing.
This is a very innefficient shape, but it's very convenient!!

Barrel is a pen casing machined down to 0.3 mm wall thickness.
Coils are wound directly onto the barrel.

Pulse timing uses a 555 timer. I had designed this to be optically triggered, but during testing, I fired it manually with a switch.

First coil was 300 turns of .8mm wire in 6 layers.
Using this coil I could punch a hole through two sheets of paper at 18 inches & still bounce the bearing off the wall!

My second coil was 200 turns of 1mm wire in 6 layers.

Second coil was much more effective & could fire the bearing through a manilla folder & still bounce off the wall.
Unfortunately after I increased the pulse length past 500uS the igbt's shorted.

This wasn't unexpected. I knew that I had overestimated the devices, but decided I would see how far they could be pushed.

I'm sorry I haven't any velocity readings, but I am working on a speed trap, so maybe I can measure version two's speeds.

Version two uses 3 igbt's per side & after some more research (thank you WmHarriss) I learnt that they can be forced to current share by placing a .01 ohm resistor in the emitter circuit of each device. Hopefully this will help extend their lifespan!

Version 2 also uses a picaxe micro to control firing.

The circuit diagram for version2 (attached) is preliminary, the boost converter works, but I have yet to double check IR2113 circuit for accuracy.

Photo is of version 1 with pulse timer board seperate. igbt's have been removed.

The third pic I have attached is the pc board for version 2 & I would like some advice on this if possible.

I am concerned about the tracks I have indicated with coloured lines.

Red is the trigger from the picaxe to the IR2113. Is it likely that voltages induced into this track may cause problems?
Could the "spikes be controlled by using a zener diode?

The blue is an earth track. I have taken this earth from the battery lead & run it all the way to the IR2113 to help prevent high currents from causing earthing problems at the ic. Is this good policy, or is the long track going to be a potential problem from induced spikes?

I would greatly appreciate any input.

Thank you all once more for posting your projects & results. Without your information, I could not have gotten this far.

rp181

Mon Dec 22 2008, 04:47PM

Registered Member #1062 Joined: Tue Oct 16 2007, 02:01AM
Location:
Posts: 1529

Very nice setup, and very neat. For a chronograph, you can make a simple one with 2 photo gates. The two receivers are connected in parallel, and connected to a PC microphone cord. If you use a program like Audacity, and record the "audio" during the firing, you will see two spikes on the recording. If you measure the time between the spikes (audacity does this too), then you can use the distance of the 2 photo gate pairs to calculate speed.
I agree with you on pushing the IGBT's. You would probably be better off with some minibrick's (ISOTOP package).
Did you etch those boards yourself? there very neat.

BTW, i made a java applet to make calculating speed and efficiency easier.

Hope this helps =)

Backyard Skunkworks

Tue Dec 23 2008, 02:36AM

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

Another easy way of getting velocity is simply firing the gun straight up and measuring how long until the projectile comes down. This works pretty well unless the gun has enough ballistic energy to become dangerous, or clearly fires at over 40 m/s.

I've had a lot of sucess with rp181's method too, it just depends on which you find easier...

And of course you can also measure the deflection of the projectile when fired perfectly level (requires a bit more number crunching).

rp181

Tue Dec 23 2008, 03:34AM

Registered Member #1062 Joined: Tue Oct 16 2007, 02:01AM
Location:
Posts: 1529

With the method of firing straight up, results will be off unless the projectile falls straight back into the gun barrel =/
Of course with the way i posted, your not going to get the exact beginning of the peak, but you can zoom in alot on audacity. Im working on a program that automates it, giving high accuracy by using the absolute peak of each spike.

Lekky1

Tue Dec 23 2008, 09:11AM

Registered Member #1592 Joined: Thu Jul 17 2008, 06:58AM
Location:
Posts: 11

Thank you gentlemen!

RP181, I have downloaded your applet, but I get a message sying "this browser does not support applets".

I'm guessing I have to tick a box to allow it to run, but haven't found where yet.

Thank you both for the velocity measurement ideas.

Unfortunately the bearing is so small, I would probably lose it if I shot it straight up!

I have some opto switches that would work well with the audacity program, so will download it & setup when I have version 2 finished.

The boards I etch myself using a "press & Peel" iron on film. The pattern is printed onto the film using a laser printer & then ironed onto the board.
I haven't graduated to double sided boards yet, but its going to have to happen soon because I'm finding myself limited with single side.

Received .01 ohm smt resistors today, so will forge ahead over the next week or so to finish version2.

rp181

Tue Dec 23 2008, 03:48PM

Registered Member #1062 Joined: Tue Oct 16 2007, 02:01AM
Location:
Posts: 1529

Are you opperating it in the window or downloading it? either way, java needs to be enabled (somewhere in tools, internet options, advanced, there should be a java category).

You should use a metal rod for projectile (home depot and Lowes have them). It helps ALOT, and also makes detection with a photogate easier.

Lekky1

Wed Dec 24 2008, 04:20AM

Registered Member #1592 Joined: Thu Jul 17 2008, 06:58AM
Location:
Posts: 11

Thanks Rp181, I'll have a look at my Java settings.

I have opted for ball bearings as opposed to a steel rod because they are readily available in a uniform size.

Once I get the coil gun electronics sorted out, I have plans to add an injector/pre-accelerator solenoid.

Ball bearings will be a lot easier to load.

Any thoughts on induced voltages in the tracks I have marked on the pcb?

I have the output of the micro connected to this track clamped using a 5.6v zener.

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