Type of zener diode anti-parallel with the coil should be used? / Electromagnetic Projectile Accelerators / Forums

Forums

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

« Previous topic | Next topic »

Type of zener diode anti-parallel with the coil should be used?

Move Thread

LAN_403

OwlFowl

Sun Apr 13 2008, 04:05PM

Registered Member #1441 Joined: Sun Apr 13 2008, 03:40PM
Location:
Posts: 10

I'm building a low powered coilgun using 3x 400v 470uf caps in parallel. I'm going to use the stud type SCRs and I really don't want it to fail on me after switching it once because of the back emf. I've seen ways to eliminate this back emf using a diode / diode and resistor. But I want the current to short out quickly, so I'm considering zeners. If I put a zener anti-parallel with the coil, what rating zener am I looking for?
(I don't have a multimeter, lc meter, etc...so everything is ballpark)

Are anti-parallel diodes really that needed for the SCR I'm using, b/c I switched with a S6020L w/o a diode it didn't get hot or blew up...caps were fine too but the scr legs were just all shorted together...

pbfree

Mon Apr 14 2008, 01:16AM

Registered Member #1386 Joined: Tue Mar 11 2008, 08:16AM
Location:
Posts: 13

Confusing:

Yes, an anti-parallel diode is needed. Its rating should match the current through the coil during discharge. The current may actually be less, coil dependent, but matching them is how you know its somewhat safe. Why the search for a zener?

Your S6020L did not get how or blew up but it no longer functions? If it died gracefully, its still dead and yest needs protection. That protection is for every component in the circuit. Any diodes used to charge the bank, any mosfet, igbt, or transistors used to do the charging, the output of what ever is the gate supply, the mail power supply and regulators, all need the protection from a large reverse voltage.

OwlFowl

Mon Apr 14 2008, 04:36AM

Registered Member #1441 Joined: Sun Apr 13 2008, 03:40PM
Location:
Posts: 10

Hey thx for the input!

I'm new here and I just noticed I didn't even introduce myself....how rude. I apologize.
I'm OwlFowl

Hmm, I knew my post was going to be confusing because I rushed it... Anyway I agree with you, my S6020L died and therefore protection is needed.

I want to use a zener because I've heard that just by putting a regular diode anti-parallel will result in a longer discharge time, and might even create that dreaded suck-back effect.

My goal in this is to protect parts from back emf without elongating the discharge time.

If only there isn't such thing as back emf...

Barry

Tue Apr 15 2008, 01:59PM

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

If only there were no such thing as inductive kickback.

To be precise, "back emf" is caused by the armature traveling through a magnetic field, where inductive kickback is the high-voltage spike resulting from breaking the connection to an active coil.

Note that neither of these two effects are involved in a capacitive-discharge coilgun using SCRs. No kickback because no connection is opened; it's just normal sinusoidal current decay until the SCR turns off. And no noticeable back emf because coilguns have such low coupling between the coil and projectile (which also causes such low energy conversion efficiency).

So your SCRs should handle this usage just fine. With no voltage spike involved, the SCR's reverse-voltage rating is never at risk. Just be sure their current rating can handle your 112J capacitor bank. (Nice size by the way! I like it!)

It's those darned polarized capacitors we worry about. If you allow a negative voltage to appear on their terminals, it will damage the electrolyte. The amount of damage depends on the how much and how long reverse voltage exists. It may not fail right away, but its lifetime is shortened by some amount. And hence we look at protective diodes across the capacitor to avoid negative voltages. The problem arises because once the inductor is all charged up, it must necessarily have a negative voltage across its terminals as it discharges.

But as you've noticed, adding the diodes will greatly lengthen the decay time for the inductor to turn off. And that puts us into the dreaded suckback effect territory. As long as we use polarized caps and underdamped RLC circuits, we have to deal with this problem. I've already built a critically-damped coilgun and it sucks, so we can either (a) use nonpolarized capacitors such as motor-start caps, or (b) add something to the design to protect the caps.

Zener diodes are one idea, but they don't usually carry the kilo-amp current rating we'd like to have. Also, the zener diode's rating is still going to appear on the capacitor as a reverse voltage.

The V-switch

looks attractive for its ability to turn off the inductor so quickly. Also the H-bridge

is very attractive for its promise to recover magnetic field energy and thereby improve efficiency.

Something is needed. Keep up the good work and don't give up!

Hope this helps! Barry
Do you ever feel like that guy in the xkcd cartoon?.

OwlFowl

Wed Apr 16 2008, 01:53AM

Registered Member #1441 Joined: Sun Apr 13 2008, 03:40PM
Location:
Posts: 10

Thanks for clearing up inductive kickback and back emf for me, I thought they meant the same thing

As you can tell it's my first time dealing with coilguns.

Great! I think my 600v 85A i got from ebay can definitely handle the pulse. As soon as it arrives I'll try putting diodes with the caps and see how bad the suck-back effect is. The coil I've made is only 4cm long 6 Layers 22awg so hopefully it should have a short discharge time even with the diode. If not, I'll find out about the half bridge.

Barry, are you the person who has made the Barry's coilgun site by any chance? I got pretty much all my theoretical knowledge on coilguns from there. I used the coil simulation, RLC model, etc. They really helped a lot.

Again, thanks for your inputs! I'll upload pics and facts of my gun when it's done. Hopefully that's by the end of next month because that's when my project is due. (it's a school ISP...)

Barry

Wed Apr 16 2008, 03:14AM

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

Well as a small matter of cosmic coincidence, I really am the same Barry that publishes Barry's Coilgun Design Site. Thank you for the kind words, I appreciate it very much. And welcome to the world of coilgunning, it's great to have you here. I sure look forward to your photos and results.

OwlFowl wrote ...
The coil I've made is only 4cm long 6 Layers 22awg

It sounds like a bit thinner wire than you really want. The coil simulator suggests it may have as much as 2 ohms of wire resistance, which will dissipate a great deal of energy in the first few milliseconds when you need it the most. You might have another look at wire diameter, and see what happens with say 16 or 18 awg instead. It would be great to get the coil resistance down to a few tens of milliohms.

And oh yes, be sure your connections and hookup wire have lower resistance than the coil, if possible. This means that soldering most of the connections will pay off with better performance.

Cheers, Barry
The geek shall inherit the earth. (Maybe they already have?)

OwlFowl

Thu Apr 17 2008, 02:57AM

Registered Member #1441 Joined: Sun Apr 13 2008, 03:40PM
Location:
Posts: 10

Cosmic coincidence

I understand there's too much resistance with the diameter of the wire I use, but my budget and time is very limited because of work.. My caps do not even charge to the full 400v, they are only charged by the disposable camera circuits which is only up to around 300V. I didn't expect much efficiency from this coilgun because it's only my first; I only wanted it to shoot out a projectile at a reasonable speed. (20m/s or 1.5% efficiency is my maximum expectation)

By the way I just realized there is an inner diameter input from the Inductor Simulation... From that I've used the RLC model from Barry's Coilgun Site and got this chart
6 Layer Coilform

I think it's a very damped circuit with only 15-20A of negative voltage so I will parallel two 6A10s and pray for a short discharge time.

I'm pretty sure I'll be getting low muzzle velocity with this gun but I'm planning to make a second CG in the future which is going to be more efficient

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