First (real) Coilgun / Electromagnetic Projectile Accelerators / Forums

Forums

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

« Previous topic | Next topic »

First (real) Coilgun

Move Thread

LAN_403

uzzors2k

Sun Jun 18 2006, 02:05PM

Registered Member #95 Joined: Thu Feb 09 2006, 04:57PM
Location: Norway
Posts: 1308

Long time since I've been in the EM projectile forums! Its taken more than a year, but I've finally gathered everything for my real coilgun. The bank is 450v @ 4700 Âµf, so a maximum of 500 Joules. The coil is wrapped around a pen body, and the projectile is anything I find laying around. Horribly large airgap, so efficiency is pretty poor so far. Unfortunately I haven't been able to charge it beyond 300 volts yet, which equals to less than half of the cap bank's joule capacity. It won't even shoot through a cardboard box yet.
So before I officially unveil some test videos, how can I charge the caps up? Until now I've been running my boost converter off of a 100W supply. The charge time is still lengthy, and it stops charging at 300 volts. What I want to do is charge the cap bank directly from mains. Can the capacitors take the current though?

Bjørn

Sun Jun 18 2006, 03:59PM

Registered Member #27 Joined: Fri Feb 03 2006, 02:20AM
Location: Hyperborea
Posts: 2058

The capacitors will take it but the fuse will not, use a lightbulb in series to limit the current. A rectifier will give you about 330V DC. If you want more you need to make a voltage doubler and some way to stop charging when it reaches 450V.

You should add a bleeding resistor that constantly discharges the capacitors since it is just a matter of statistics before you manage to touch the terminals.

uzzors2k

Sun Jun 18 2006, 07:53PM

Registered Member #95 Joined: Thu Feb 09 2006, 04:57PM
Location: Norway
Posts: 1308

Ok, good advice. The fuse blew on my first attempt at charging from mains, I didn't know the caps drew so much current to charge! I'll try the lightbulb current limiter then. How can I detect the voltage on the capacitor bank? The first thing I think of is a resistor divider, which could trigger a FET. Could it work? What is the minimum voltage a FET will trigger at?

Btw, I had a 10K bleeder in place, but it wasn't soldered on yet in that picture. Thanks for thinking of my saftey though!

Jim

Sun Jun 18 2006, 09:12PM

Dunce.
Registered Member #28 Joined: Fri Feb 03 2006, 03:29AM
Location:
Posts: 76

An empty capacitor is akin to a dead short.

Avalanche

Sun Jun 18 2006, 09:58PM

Registered Member #103 Joined: Thu Feb 09 2006, 08:16PM
Location: Derby, UK
Posts: 845

uzzors wrote ...

Ok, good advice. The fuse blew on my first attempt at charging from mains, I didn't know the caps drew so much current to charge! I'll try the lightbulb current limiter then. How can I detect the voltage on the capacitor bank? The first thing I think of is a resistor divider, which could trigger a FET. Could it work? What is the minimum voltage a FET will trigger at?

Btw, I had a 10K bleeder in place, but it wasn't soldered on yet in that picture. Thanks for thinking of my saftey though!

A resistive divider is part way correct to detect the voltage, but for more accuracy you will require a comparator, instead of relying on the FET switching on at a certain voltage.

A comparator will compare two voltages, and switch it's output high or low depending on which voltage is greater. You can use the output of the comparator to control a power mosfet, which switches the charging voltage (ie. rectified mains) on and off. So you have two resistive dividers - one for the capacitor bank, whos output will vary between say 0-12 volts, and then the other divider is on a known voltage source. By comparing these two voltages, the comparator is a very accurate way to switch off your charger when it reaches a set voltage, they are great little chips.

Hope this helps!

Bjørn

Sun Jun 18 2006, 10:00PM

Registered Member #27 Joined: Fri Feb 03 2006, 02:20AM
Location: Hyperborea
Posts: 2058

The capacitors will take as much current to charge as they give when discharging them unless you limit the current.

Put a small resistor in series with your bleeder resistor, use Ohms law to find the value to give the voltage you need. Driving a FET directly is not a good idea because it will make it very hot when it is conducting less than 100%. Use a comparator circuit with a bit of hysteresis.

rupidust

Sun Jun 18 2006, 10:33PM

Banned
Registered Member #110 Joined: Fri Feb 10 2006, 12:23AM
Location: Banned City
Posts: 85

If you have a 100W equivalent output from your boost converter, the charge time should be close to 5 seconds. Not being able to charge pass 300v indicates another issue. To test, remove 1/2 the caps to make the bank 211 joules. If failure to charge past 300v persist, either the charger is dying early on or a cap is dying like a bad apple in the apple barrel. Question, are you feeding the booster only 100 watts from a power supply or is the supply a bank of batteries? Can you show the boost converter?

Another issue I see is it not even shooting through a cardboard box yet. At 300v the bank holds around 211 joules. 211 joules more than sufficies to penetrate a cardboard box assumming 2% output for a 4 joule kinetic slug. Once the cap bank is near fully charged, the performance just may decrease because the system already shows signs of excess capacitance and surpluss of suck back.

Your setup accomodates small, short, and light weighted projectiles. I myself will not go over 100 joules per stage for slugs twice as big as I see will fit that pen barrel. Try using only 50 joules with or withouth charging issues.

uzzors2k

Mon Jun 19 2006, 11:30AM

Registered Member #95 Joined: Thu Feb 09 2006, 04:57PM
Location: Norway
Posts: 1308

This is the boost converter design I'm using

except I'm using a K1507 MOSFET instead of an IGBT. Nothing was tuned on it, since I built it over a year ago when I was pretty newb in electronics.

I messed up with the power consumption of the boost converter, 100W is what the PSU can deliver. Turns out it draws 30 watts max. It draws 2.5 amps when I first begin charging, but it tapers off rapidly, and trickle charges to around 300. So 30 W to about 10W. I'm running it off a PSU at 12volts. Funny thing is the same thing happens with the mains charger, it draws enough current to light the lightbulb, but as I charge I can watch the lightbulb dim. The charging always stops around 300 volts. The output from the rectified mains is 320volts, but I could only get the cap bank to 310. Is that a sign of a faulty capacitor in my bank?

What seems wierd to me is that the boost converter draws so little, even when faced with a virtual full short. The charging problem must lie there. I'll try building a different design, but what could be limiting the current so much?

The projectiles I use don't fit very snug, so I'm sure a lot of efficiency is lost there. In fact I think there is room for almost twice as much projectile in the barrel. Currently the coilgun can penetrate a cardboard box at about 30 cm distance. That required about 200 joules. So other than the charging problem and the projectile size, the efficiency is ok so far. After I get the bank fully charged I'll begin tweaking. I don't think the bank is too large yet, as the coilgun has shot stronger for every volt so far.

Btw, I made a pretty retarded mistake while measuring the current. I'll let the timer tell the story.

EDIT: After digging in the archives I found the problem with the boost-converter, not enough voltage input.

TheMerovingian said...

The second thing to remember is that for a boost converter Voutmax < Vin/(1-max duty) (and < max switch voltage, but this is not a real problem).

So if I can resurrect my boost-converter, and my PSU, I'll have my coilgun running.

Thx for the help so far!

rupidust

Tue Jun 20 2006, 02:11AM

Banned
Registered Member #110 Joined: Fri Feb 10 2006, 12:23AM
Location: Banned City
Posts: 85

You need to feed that booster 24+ volts for it to put out an equivalent 100 watts. That is a 30 watter booster when ran from 12v and 2khz and will deliver an equivalent 25 watts after minusing mosfet/igbt watts. As for output, the booster will pass 1200v from a 12v input, 2khz at > 50% duty cycle signal to IGBT, 1200v IGBT and 1200v diode, I know because I designed it, but of course should be limited to 600v when switched with a 600v Mosfet/IGBT. 2.5 amps is the limit because of the inductor's inductive reactance. Run then booster at a low 100 hertz and it will smoke the igbt and coil in a second or so because the inductive reactance is now very low.

Try charging only 1 cap at a time till bad one is found. If you run a doubler from 120vac mains, each of your 450v caps will charge to 340vdc. If the whole lot of caps do not charge then the error exist because mains is bad :), double is bad, the whole lot of caps are bad, or builder is bad, or combination thereof.

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