Coilgun Project Help / Electromagnetic Projectile Accelerators / Forums

Forums

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

« Previous topic | Next topic »

Coilgun Project Help

1 2 3 next

Move Thread

LAN_403

BenB

Mon Apr 04 2011, 06:54PM

Registered Member #3785 Joined: Sat Mar 26 2011, 03:37PM
Location: Surrey, UK
Posts: 16

Hi all

I guess I would describe myself as a high voltage virgin, but I've decided to build a coilgun as an a level extended project. My current knowledge is limited to A-Level physics standard, but im very happy to read up on things. I have written up a list of the specifications I'm aiming for, but I would feel much more comfortable (/safer?) with a little guidance from any persons wiser than me!

I'm aiming for a firing energy of around 500J, and an efficiency of between 1-2percent. I'm aiming to trigger the gun with an SCR, and I was planning to construct the coil from enamelled 15AWG copper wire, using approx 100 feet (slightly unsure here!) In the name of simplicity, I'm only going to use a single stage.

Charging wise, I'm using a boost converter, and I have already constructed the circuit from the schematics I found on uzzors website. I'm running the circuit off 2 (very dead) 9 volt batteries in series, but it will only charge my current cap bank (1000uf, rated 330V) to around 240 volts. Will this improve with a better power supply?
I'm currently monitoring the voltage with a multimeter, could anyone direct me towards a (cheap) LCD style display to wire into the gun?

The gun itself will be mounted on a mdf or perspex frame which will also form the grip and trigger, and I plan to clad the gun in either plastic, brass sheet (eddy current problems?) or more perspex

Really I'm looking for component advice, mainly what sort of SCR I will need, where I need to make modifications/additions that I would probably overlook, and any general tips people have!

If you read this far, thanks very much and excuse my awful grammar and writing!

Ben

Turkey9

Tue Apr 05 2011, 03:05AM

Registered Member #1451 Joined: Wed Apr 23 2008, 03:48AM
Location: Boulder, Co
Posts: 661

Sounds like you have more of a grip on what you're doing than most people who ask for general construction advice, good job!

As for a cheap lcd voltage display, I'd get a multimeter off eBay for a couple bucks and strip it down to the bare bones circuit board and display. Then solder the lead connections right into your cap bank.

I'm not 100% sure but I'd be willing to bet that your boost converter will work great if you have a real power supply behind it. If you don't have one, convert an ATX supply as they are great general purpose supplies. There are plenty of tutorials online about how to do this.

BenB

Tue Apr 05 2011, 10:09AM

Registered Member #3785 Joined: Sat Mar 26 2011, 03:37PM
Location: Surrey, UK
Posts: 16

Thanks Turkey, Yeah I guess I'm a prior planning sort of guy, I've read around the topic a little bit :).

I forgot to mention, I'm hoping to make this gun entirely portable. My plan was to use a li-poly battery at ~ 14V to power the boost converter, but I'm unsure whether this would damage the battery. (Li-pos are fairly unpredictable and dangerous) Would I be able to just wire a battery straight into the boost converter with a switch, or would I need to add some sort of charging resistor? I presume an ATX supply needs to be connected to the mains?

Capacitor wise, I think that a higher cap voltage gives a greater current running through the coil and subsequently a greater magnetic field? How important is this, as I might consider replacing my current photo-cap bank with some computer grade 450V caps.

What sort of rating do I need for the diode anti-parralel to the coil?

Many Thanks

Ben

Barry

Tue Apr 05 2011, 12:21PM

Registered Member #90 Joined: Thu Feb 09 2006, 02:44PM
Location: Seattle, Washington
Posts: 301

BenB wrote ...

I'm aiming for a firing energy of around 500J, and an efficiency of between 1-2percent.

The gun itself will be mounted on a mdf or perspex frame which will also form the grip and trigger, and I plan to clad the gun in either plastic, brass sheet (eddy current problems?) or more perspex

Welcome to the forum!

I hope you mean 500 joules of stored energy. That's a really nice size to design around and you'll definitely get some very exciting results. Otoh, if you mean 500J of kinetic energy then you're gonna have all sorts of problems with a 25KJ capacitor bank.

Be sure to choose a non-conductive firing tube to avoid eddy current losses. The coilgun's outer shell or cladding material is not as important. In fact, you can use anything you want around the coil as long as there is an insulating break in the skin that interrupts current from flowing in a complete path in a loop outside the coil. Think of the outer shell as the secondary of a transformer; if you open the loop, then no current can flow, and no resistive losses can occur.

When searching for SCRs, look for the pulse rating instead of the rms current rating. The pulse rating is usually specified as the maximum one-time current for 8 msec. If the pulse rating isn't provided (and it's often not available) then it's generally accepted to be 10x the rms current rating. For example, if you find a 80-amp SCR on eBay then you can use it in a coilgun that has an 800-amp peak current. Your mileage will vary, and yours will probably peak at a couple thousand amps or more.

Can't find or afford an SCR big enough? One design alternative to consider in order to reduce your peak firing current is choose a heavier or longer projectile. This will lead you into a design that spreads out the same 500J over a longer period of time, thereby reducing the peak current. Short pulse = longer time = lower peak. Think of the 500J as the area under the curve, where the waveform is the first half of a sinusoid. Just an idea as a backup plan, go for the big SCR first, lol.

Cheers, Barry
Why is it called tourist season if I can't shoot at them?

BenB

Tue Apr 05 2011, 05:18PM

Registered Member #3785 Joined: Sat Mar 26 2011, 03:37PM
Location: Surrey, UK
Posts: 16

Thanks Barry

Yes, I'm talking 500J stored energy. 25Kj is perhaps a wee bit ambitious for a first project! (understatement of the week?)

Would a pvc tube suffice? I know the coils contract slightly when the gun fires, how thick do the walls of the pipe need to be?

I just had a quick look on farnell and I came across this SCR: VS-181RKI100PBF http://uk.farnell.com/vishay-formerly-i-r/vs-181rki100pbf/thyristor-285a-1000v-to-93/dp/9104623 Would that take the current? Would I get away with anything less beefy, the budget is tight here :P If an SCR has 4 leads, where does current need to be applied to trigger it?

Does the diode across the coil need to be equally hefty, does an equally strong backwards current develop after the magnetic field collapses? (knowledge is limited here :P)

Many Thanks for such detailed replies (and sorry for all the questions!)
Ben

Turkey9

Wed Apr 06 2011, 02:10AM

Registered Member #1451 Joined: Wed Apr 23 2008, 03:48AM
Location: Boulder, Co
Posts: 661

You don't need an SCR anywhere near that large. I'd say it'll only need to handle around 600 - 1000 amp peak, which can be had in small to-247 package devices that cost a couple bucks each.

Edit: Thanks barry, my quick estimations on peak current were waaaayyyy off. Go for the stud, it will last for a long time and be used in many projects.

Barry

Wed Apr 06 2011, 01:49PM

Registered Member #90 Joined: Thu Feb 09 2006, 02:44PM
Location: Seattle, Washington
Posts: 301

BenB wrote ...

I just had a quick look on farnell and I came across this SCR: VS-181RKI100PBF http://uk.farnell.com/vishay-formerly-i-r/vs-181rki100pbf/thyristor-285a-1000v-to-93/dp/9104623 Would that take the current? Would I get away with anything less beefy, the budget is tight here :P

If an SCR has 4 leads, where does current need to be applied to trigger it?

The stud SCR looks about perfect to me! I'm not sure a tabbed device like the TO-247 can handle it. But it's easy enough to check by estimating your peak current.

Let's try it on the RLC Simulator page. I used your 500J bank with some typical assumptions of 300 volts and low resistance. To get 500 joules that means the capacitor is 11,000 ufd. Assuming a 2 msec discharge, that means the inductance is 32 uH.

Try it. For starters dial in V=300 volts, R=40.0 mohm, C=11.0 mf (= 11,000 uf), L=36 uH. The peak current is around 3,500 amps.

Play with the sliders to use your expected components. Try other values to see if you get a combination that might fit a smaller SCR in the range of 600 - 1000 amps.

For connecting a stud SCR, follow your link and open the datasheet. Scroll down to the pictorial diagram. The high current path is the obvious big anode and cathode connections. The Gate is the white wire where you trigger it. The small red wire goes unused; it's another cathode connection for convenience to let you measure the exact voltage.

For the firing tube, anything nonconductive will work. PVC is okay but something with thinner walls is better. It doesn't need strength, so even a big drinking straw can work. Don't worry about contraction, just be sure the projectile has a loose sliding fit.

I fully expect your stud SCR will work fine and last a long time here, even though it's a bit pricey. But I just don't expect any small tabbed SCR can survive a 500J discharge in 2 msec.

Cheers, Barry
Whatever happened to SCR prices in the last five years? I used to be able to get a nice stud SCR on eBay for $5 each and now they're going for $50. Bastards.

BenB

Fri Apr 08 2011, 01:12PM

Registered Member #3785 Joined: Sat Mar 26 2011, 03:37PM
Location: Surrey, UK
Posts: 16

Thanks Barry, just ordered the final load of parts for the project.

I'm still having trouble with the charging circuit. I've checked and double checked the circuit against the schematic, also tried driving it with every 1.5V increment from 9 to 15 to make sure it wasn't a power issue.

The caps (1000uf @ 330v at the mo) charge in ~20 seconds to 280 volts, then my multimeter (rated to 600v) starts giving off crazy voltage reading. It darts rapidly from values between 0 and 1932 (and every random value in between :S). If I turn off the circuit, the reading settles back down to 280 volts, and turning it on again begins sending the multimeter mad. I'm measuring from one terminal of the bank to another.

I just hope I'm not being stupid :P Edit: More than happy to supply photos/video if it helps clarify anything
Edit #2. I wouldn't happen to need a charging resistor would I? Just put 2 1/4w 68r resistors in series between the leads and the cap and got a stable reading (And a lovely burning smell too!)

Turkey9

Sat Apr 09 2011, 07:50PM

Registered Member #1451 Joined: Wed Apr 23 2008, 03:48AM
Location: Boulder, Co
Posts: 661

I'd try different frequencies on your charger. I only have a basic understanding of boost converters, but it sounds like as the current drawn decreases, your converter destabilizes. This might mean that you need to change the duty cycle or frequency of your charger.

BenB

Sun Apr 10 2011, 08:14PM

Registered Member #3785 Joined: Sat Mar 26 2011, 03:37PM
Location: Surrey, UK
Posts: 16

Cheers, I think I'm going to put some pot's in place of the resistors on the 555 and tune it in.
On another note, I machined up a barrel and bolt system today out of brass. Just need to slot the brass to reduce eddy currents, the barrel is <1mm thick and the projectile fits within 1mm, so hoping for some decent efficiency with this gun :)

1 2 3 next

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