Damaging SCRs / Electromagnetic Projectile Accelerators / Forums

Forums

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

« Previous topic | Next topic »

Damaging SCRs

1 2 next

Move Thread

LAN_403

nlbutts

Thu Jul 28 2011, 01:34PM

Registered Member #4033 Joined: Thu Jul 28 2011, 11:23AM
Location:
Posts: 6

I built my first coilgun a few weeks ago and it worked great. I used a 100VDC 11,000 uF cap and then experimented with different coil designs. My first firing circuit used a high power (or I thought high power) IGBT. It lasted for one firing. I then bought this SCR on eBay:

It is the IRK-136-14 part (1400 V reverse breakdown). This SCR should be able to handle a non-repetitive current of 3200 Amps.

It worked great for my 100 VDC gun. But when I got my caps on eBay it came as a set. So I also got two 450V 350uF caps. So I redesigned my gun to use these two caps. I charged my system up to 200 V and it worked great. Then I charged it to 400 V and I thought it worked great. But now the SCR has about 1000 ohms across it. After letting the system set overnight the SCR is now showing 700 KOhms across the anode and cathode (Yes I am measuring using the correct polarity on my multimeter). I didn't try to charge the caps and fire it again, but it looks like the SCR may have been damaged.

My question is why would it fry. Based on simulations in Spice and the tools on

, the peak current should have been around 1000-1200 Amps. This SCR should be able to handle periodic currents as high as 3200 Amps.

I've attached a poor schematic of my firing circuit.

Thanks in advance!

Forty

Thu Jul 28 2011, 07:21PM

Registered Member #3888 Joined: Sun May 15 2011, 09:50PM
Location: Erie, PA
Posts: 649

the 4th page of that datasheet shows 3 distinctly different versions of that model, so which do you have?
there's pretty much no way that those caps could of fried the scr unless it was already broken when you got it, or you fried the gate. what were you using to signal the gate?

nlbutts

Thu Jul 28 2011, 11:17PM

Registered Member #4033 Joined: Thu Jul 28 2011, 11:23AM
Location:
Posts: 6

I have the IRKT model. I have my caps connected to pin 1, the coil connected to pin 2. I have a voltage divider from wall power down to 15 VDC. I then switch this voltage through a light switch to pin 4.

I just tried it again and it keeps conducting. Maybe I need to ground pin 4 instead of leaving it floating.

Thanks for the help.

nlbutts

Fri Jul 29 2011, 02:33AM

Registered Member #4033 Joined: Thu Jul 28 2011, 11:23AM
Location:
Posts: 6

You mention frying the gate. I guess I didn't put any current limiting resistor before the gate control. I had a 4.7uF cap charged up to ~5 volts. When my switch closes, all of that energy gets dumped into the gate, which may have fried it.

Are these things really that sensitive?

Forty

Fri Jul 29 2011, 02:57AM

Registered Member #3888 Joined: Sun May 15 2011, 09:50PM
Location: Erie, PA
Posts: 649

well they are called sensitive gate thyristors lol.

from the datasheet it looks like the max gate current is 3A. minimum required to trigger is 2.5v at 150mA.

what happens when you hook it up to the bank and charge it?

nlbutts

Fri Jul 29 2011, 01:07PM

Registered Member #4033 Joined: Thu Jul 28 2011, 11:23AM
Location:
Posts: 6

Yeah, I missed that part when I was looking at other devices. When dealing with BJTs and MOSFETS I always put a current limiting resistor on the base/gate. But this thing is the size of a brick, why would one need to handle with kid gloves :(

So it has two SCR in the same package. The one I have been using "looks" like it is high impedence when I measure it with a multi-meter. But when I start charging the caps it starts to conduct and my caps do not charge above ~5 V. If I disconnect the SCR, then I can charge my caps to 450V (I have a bleed resistor that will bleed away any energy in the caps when not in use).

Last night I tried to hook up the second SCR (still not using a current limiting resistor). It behaves a bit strange. It will fire, but it must fire VERY slowly because my coil can barely move the projectile. I don't have a scope so I can't look at the actual waveform. I tried to fire it several times all with the same results.

I ordered another (although not nearly as big and fancy) SCR last night. When that one comes I WILL put a current limiting resistor on the gate before I fire it.

nlbutts

Fri Jul 29 2011, 05:09PM

Registered Member #4033 Joined: Thu Jul 28 2011, 11:23AM
Location:
Posts: 6

OK, Now I'm confused. I removed my two 450V 350uF caps and put my old 100V 11,000uF cap and the SCR works great. I did add a 10ohm current limiting resistor and now use a 9V battery to fire it. I fired it several times and it works great. It's like the SCR can't hold back higher voltages.

Forty

Fri Jul 29 2011, 05:40PM

Registered Member #3888 Joined: Sun May 15 2011, 09:50PM
Location: Erie, PA
Posts: 649

that happened to an scr i fried the gate on. it would only hold back like 70v.
keep using the second scr (that's convenient that you've got two chances in one)
what does your coil look like? (wire diameter, turns, length)

nlbutts

Fri Jul 29 2011, 08:31PM

Registered Member #4033 Joined: Thu Jul 28 2011, 11:23AM
Location:
Posts: 6

I've played around with a few different coils. My current coil is 12mm long and around 4 layers thick. I use 22gauge magnetic wire (available at Radio Shack). I should really get some thicker stuff, so if anyone has suggestions, I'm open to it.

I measured the resistance and inductance of the coil yesterday. I got 0.184 ohms and 56uH of inductance. In Spice simulations I get a peak current of 310 amps and a pulse duration of around 14ms at 100V and 11.000uF.

With my second setup at 400V and 700uF I should get a peak current of 768 amps and a pulse duration of 1.5 ms. I'm firing a 4 gram slug. I probably need to improve my coil design. I'm not using any external iron or anything fancy.

Since the second SCR looks like it is working, I'm wondering if at 400V if my pulse is too short and it doesn't accelerate the slug. What do other people aim for for the pulse duration?

Forty

Fri Jul 29 2011, 09:38PM

Registered Member #3888 Joined: Sun May 15 2011, 09:50PM
Location: Erie, PA
Posts: 649

you can adjust the initial positioning of the projectile to better fit the pulse length. finding the distance mathematically is pretty much hopeless, since there are so many factors to take into account. so just set up a little barrel with measured tick marks on it and adjust the placement until you find the best spot.

1 2 next

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