Re: 350v & 3760uF coilgun. Are back emf diodes a must?
Camel, Sun Nov 16 2008, 04:19AM

You'll need to estimate the inductance of your coil. Most likely you will have enough back emf to destroy your caps.

Re: 350v & 3760uF coilgun. Are back emf diodes a must?
casey, Sun Nov 16 2008, 04:22AM

it is about 700 amps

Re: 350v & 3760uF coilgun. Are back emf diodes a must?
teslacoolguy, Sun Nov 16 2008, 04:37AM

Yes you will need back emf diodes.

Re: 350v & 3760uF coilgun. Are back emf diodes a must?
Barry, Sun Nov 16 2008, 03:15PM

3760uF and 350v is about 230 joules. Nice.

Inductance is millihenrys (or microhenrys) not amps.

Whether or not you need a freewheeling diode depends on the total series resistance of the discharge path. With just a little resistance, the oscillation may be damped enough that the capacitor is not reverse-charged.

To find out, fire it at low energy and no diode. Measure capacitor voltage before and after. If it's more than a dozen volts negative, then you need a protective diode.
Edit: If the remaining negative voltage is more than 10% of the positive starting voltage, then you need the diode.

Better yet watch the capacitor with an oscilloscope. This tells you both the total series resistance, and the discharge time (or resonant frequency). Scopes can really help you tune the inductance for a desired discharge time.

Let's call it a "freewheeling diode" since it is the ringing (not back emf) that damages capacitors.

I strongly suggest minimizing net resistance. It makes a big difference. Use short fat wires and soldered connections. Overbuild the copper discharge path. You only need a freewheeling diode if you don't have a spare capacitor.

Cheers, Barry
I'm trying to take one day at a time, but sometimes several days attack me at once.

Re: 350v & 3760uF coilgun. Are back emf diodes a must?
casey, Mon Nov 17 2008, 03:37PM

if i can get my hands on an oscilloscope, what kinds of numbers should i be looking for? "fire it at low energy and no diode" what would be low energy in your books? like how much voltage?

Thanks for all your help Barry

Re: 350v & 3760uF coilgun. Are back emf diodes a must?
Mattski, Mon Nov 17 2008, 06:56PM

The lower the better: you can always increase energy but once you blow something up you can't go back. I would guess something on the order of 12-20V is a good low power starting point, using E=0.5*C*V^2 your energy at those voltages will be under a joule. You're looking to see what portion of the starting voltage ends up negatively charging the capacitors, so the higher this voltage is the more you might stress them.

Re: 350v & 3760uF coilgun. Are back emf diodes a must?
casey, Tue Nov 18 2008, 05:42AM


so at 90v, i'm getting about 5v bak into my caps. I should be safe? I gotta take quite a few turns out of my coil too.

the projectile flies about 5 feet at 90v, but doesn't leave the barrel with 140v. thats with the protector diodes

Re: 350v & 3760uF coilgun. Are back emf diodes a must?
guitardude012, Tue Nov 18 2008, 08:23AM

What voltage do you get back after 140v? The protection diodes help prevent the negative capacitor charge so that makes the test Barry offered difficult.

You will probably have to move the projectile starting position after such a large change in bank energy(90v->140v). This may be one reason why your projectile doesn't shoot at higher voltages.

Re: 350v & 3760uF coilgun. Are back emf diodes a must?
Barry, Tue Nov 18 2008, 01:49PM

I use ten-percent as a general rule of engineering.

Charging your capacitors to 35v can be considered "negligible" compared to their 350v rating. You can safely dry-fire and test to your heart's content without fear of damage.

The deal here is that the flyback diode will greatly extend the time for the magnetic field to die away. It is a drastic change of timing. Think of it this way ...

• the coil will be energized in approximately a sinusoidal waveform according to the LC frequency

• But then the diode kicks in at maximum current ...

• Now the coil is essentially shorted, and it doesn't see the capacitor ...

• the current dies away at a roughly linear rate according to copper's resistive losses and the 1.4v forward-biased diode drop

You found an oscilloscope? Awesome! It will really help show you what's happening in there.
Cheers, Barry
Today is a day for firm decisions! Or is it?

Re: 350v & 3760uF coilgun. Are back emf diodes a must?
Gratitude, Wed Nov 19 2008, 01:05AM

I think Barry is right If the remaining negative voltage is more than 10% of the positive starting voltage, then you need the diode, other wise just enjoy

Re: 350v & 3760uF coilgun. Are back emf diodes a must?
Backyard Skunkworks, Wed Nov 19 2008, 02:23AM

A good thing to keep in mind is that electrolytic capacitors will undergo reverse breakdown at around 1.7V. After reverse breakdown occurs, any current flow will chemically damage the capacitors, the more current flow, the more damage.

As a general rule, I'd always reccomend using (a) diode(s) in a coilgun of any signicant power (such as yours). This will really help if you want to get thousands of shots out of your caps.

Re: 350v & 3760uF coilgun. Are back emf diodes a must?
rp181, Wed Nov 19 2008, 03:20AM

Regarding reverse breakdown, this can be repaired by slowly (over a few hours) charging the capacitor. Sooner or later though, the electrolytic solution will run out.

Re: 350v & 3760uF coilgun. Are back emf diodes a must?
KLH, Thu Nov 20 2008, 07:13PM

Casey wrote...
the projectile flies about 5 feet at 90v, but doesn't leave the barrel with 140v. thats with the protector diodes

One note on protection diodes and what is going on in the inductor...

When an inductor is hooked up to a high-impedance load, its energy dissipates faster and at a higher voltage into the load because it behaves as a current source. With a diode in reverse-parallel with the main firing coil, this causes the coil to discharge into a low-impedance load (the diode). This means that the magnetic field is sustained for a lot longer, and this can cause suckback of the projectile. So the protection diodes on your 140V test are probably causing suckback.

In order to fix this, you can put a snubber resistor in series with the protection diode. To calculate the value of the resistor, use Ohm's law (V=I*R, where I is stored inductor current and R is the snubber resistance) to find the maximum resistance that will not exceed the breakdown voltage of your main power switch.

I highly recommend a snubber resistor for all spike protection circuits because the only thing that will bring the inductor's current down to zero is if the current is dissipated as heat, and I certainly wouldn't want all of that current dissipating in a diode.

Note that any back-emf that comes out of the inductor is wasted energy that was not transferred into the kinetic energy of the projectile. Therefore less back-emf before the protection diodes are added means a higher efficiency.

Re: 350v & 3760uF coilgun. Are back emf diodes a must?
buzzer, Thu Nov 27 2008, 10:17PM

"ringing, not back EMF" said Barry. Can you explain the difference between ringing and back EMF?

Re: 350v & 3760uF coilgun. Are back emf diodes a must?
rp181, Thu Nov 27 2008, 10:36PM

Ringing is voltage oscillations. This happens with too much inductance and too little damping resistance. EMF is when the magnetic field collpases, this will always happen with an inductor, you just manage the energy differently.

Re: 350v & 3760uF coilgun. Are back emf diodes a must?
Barry, Fri Nov 28 2008, 04:26PM

"Ringing" is what happens in an under-damped RLC circuit. When the total loop resistance R is less than a certain value (the critical point) then the circuit response will begin to oscillate to some degree. With progressively lower total resistance the ringing continues longer. Look here for comparison graphs.

Ringing depends on all three values of R, L and C. It is not just "too much inductance" as rp181 suggests. If resistance R < 2 sqrt(L/C) then it will ring. With any more resistance than this there will be no overshoot and no negative voltage on the capacitor.

"Back EMF" is a generated voltage in electric motors due to armature motion. In normal electric motors the back EMF is what you get when you operate it as a generator. In coilguns, as the tail end of a projectile passes through the coil it can interact with the magnetic field to generate current and therefore cause a voltage spike.

Make sense? I searched but didn't find a really good graph of a coilgun's inductor voltage, showing a small spike on the tail end of the operating cycle.

Back EMF in coilguns, by the way, is related to the DSE (dreaded suckback effect). If the magnetic field remains on too long, the projectile motion is slowed. While this is happening the kinetic energy converts back into additional coil current, causing a small voltage spike. It is not a big voltage but visible on an oscilloscope.

Cheers, Barry

Re: 350v & 3760uF coilgun. Are back emf diodes a must?
KLH, Wed Dec 03 2008, 06:23PM

Back EMF applies to anything that involves a collapsing magnetic field, and it behaves like a current source. This EMF from an inductor results whenever an applied voltage is removed. This is what causes ringing, which damages the capacitors through reverse charging.

In this case, the ringing is irrelevant; the back EMF (if strong enough) will destroy the capacitor regardless if it rings down or not.

Re: 350v & 3760uF coilgun. Are back emf diodes a must?
Backyard Skunkworks, Wed Dec 03 2008, 07:01PM



I don't think the back emf is generally strong enough to destroy a capacitor in one shot; normally the reverse charging just breaks down part of the electrolyte. Once this happens enough, the capacitor fails during chargining/discharging/charge storage.