6 Month Gauss Cannon Project / Electromagnetic Projectile Accelerators / Forums

Forums

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

« Previous topic | Next topic »

6 Month Gauss Cannon Project

first previous 6 7 8 9 next last

Move Thread

LAN_403

guitardude012

Sun Jan 04 2009, 01:36AM

Registered Member #968 Joined: Fri Aug 24 2007, 04:54PM
Location: Norman, Oklahoma
Posts: 104

The dv/dt problem rp was talking about is how quickly the input to the SCRs is allowed to rise. If you charge your cap bank and it is connected to the SCRs they can fail. The problem is that if the dv/dt rating is exceeded the SCR can conduct and since the gate level is low, there is a higher "internal resistance" although again I hesitate to use resistance in this way. The SCR then has to dissipate far more energy internally than normal and lets its magic smoke out.

rp181

Sun Jan 04 2009, 05:52AM

Registered Member #1062 Joined: Tue Oct 16 2007, 02:01AM
Location:
Posts: 1529

dv/dt is the maximum rate of rise of voltage. That means if the off state voltage rises more then the specified voltage in the specified time, the SCR will fail. It doesn't have to do with gate, it triggers the SCR with no current. Look at the datasheet, it will say something like 500v/us. What guitardude said is right.
When the dv/dt is exceeded, capacitive coupling causes a charge enough to trigger. This partiall triggering is bad, and causes much more losses than usual. Directly from a powerex technical document:
""dv/dt triggering" may not switch the SCR into full conduction rapidly and the partially-triggered SCR may dissipate more power than is usual, possibly harming the device."
Sorry if i come of as rude, but i am trying to save you the trouble. As i said, i have burned 3 donated SCR's trying to figure this out, now i am seeking some more SCR's from westcode.
BTW, i am in a IB program to =p

blackgrunge

Sun Jan 04 2009, 04:18PM

Registered Member #1889 Joined: Mon Dec 29 2008, 07:36AM
Location:
Posts: 55

Then you know what IB Physics incorporates, its really a great program. How are you liking it so far?

And to clarify, my SCRs are triggered by ceramic capacitors. I know that they have their own rise and fall time but they help to not keep the gate open when I'm holding down the trigger. Normally the gate would remain on because I'm holding down the trigger when the optics try to shut the SCR off but when I have the capacitor its one pulse and then it doesn't matter if i hold it down or not. Its too much trouble to try and hit the switch fast enough to trigger the gate while releasing before the projectile breaks the IR beam to shut it off.

The rise time of the SCR input is controlled by the inductance of the coil that is in line with it. I've never run into any trouble with that but what happened with yours that caused it to do that?

Here's some of the specifics that were requested:

Coil Diameter: 18.28mm
Length: 76.2mm
Gauge: 20 AWG
Inductance: .452mH
Resistance: .697 Ohms
Layers: 6

Barrel O.D. 7.51mm
Barrel I.D. 6.01mm

Capacitor Voltage: 400v (charged to 330v)
Capacitance: 31200uf
Terminal: Screw

Optics
Shielded IR sensors (unknown part number)
Hex Inverting Schmitt Trigger (74HCT14)
Opto-isolator Driven SCR Gates (Motorola H11)
SCR (2N5207)

Misc.
Push button 1st stage trigger
Serial Ports handled inputs
Semi-slotted graphite arrow shaft
Magnet wire coated with high heat tolerance bi-layer enamel coating
Voltage Multiplier output: 330v @8-10amps (adjustable)

Also, here are the pictures of my refined setup for suggestions on improvements (**Some are clickable thumbnails**):

Assembly for optical sensors:

Serial connection for sensors:

Serial Ports for SCR gate triggering:

SCR connection mess... :

Serial connection for SCR gate:

Logic Gate with sensor serial port:

Capacitor connections:

SCR wiring:

Overview of everything:

rp181

Sun Jan 04 2009, 04:32PM

Registered Member #1062 Joined: Tue Oct 16 2007, 02:01AM
Location:
Posts: 1529

Im still in MYP, no physics yet =p
The coil in line with the SCR shouldn't matter, its rate of rise of the voltage in the off state. I like to think of electricity as a pneumatic system, it makes explanation simpler. Ive messed around with spudguns some, and theres a vavle type called a burst disc valve. Theres a layer of material seaperating the chamber and barrel, think of this as a SCR. As the pressure (voltage) rises, it blocks air flow, until pressure is enough to rupture the disc (i.e. breakdown voltage). But if you take say half that pressure, but immediately dump it behind the disc, it ruptures due to the fast pressure spike, its just like that.
Commercial applications use snubber networks to limit this peak, but that would not be desirable in a coil gun.
From your charge voltage being 330v, im assuming your using a mains doubler. The same configuration caused a dv/dt triggering. Maybe you could draw up a schematic showing your power/charging section?

When dv/dt occurs, it does not damage the SCR. I just ruined them trying to debug it (i disconnected, charged, and connected the SCR, di/dt destroyed it). In the other thread you said you used a capacitor on the hot side, i didn't, maybe thats why =p
EDIT: just kidding, 8A out is way to high, my fuse for that outlet was 15A on the hot side.

As for wires,
When you have extra spots (as in serial cable) its good to parallel spots to minimize inductance. Also, just zip tie together every wire that goes to the same spot. I like to heatshrink my connections, makes it look better.
You have some extra space on the board, why not put the SCR's on it?

For capacitors, homedepot as some aluminum flat bar. I use hammered out 1/2in copper tubing.

El_Roberto

Sun Jan 04 2009, 08:46PM

Registered Member #1774 Joined: Wed Oct 22 2008, 02:51AM
Location:
Posts: 135

What are your sensors mounted in? Looks like Iron?

guitardude012

Sun Jan 04 2009, 09:13PM

Registered Member #968 Joined: Fri Aug 24 2007, 04:54PM
Location: Norman, Oklahoma
Posts: 104

It seems like your wire gauge is a little small for a little less than 1000 joules per coil. Are you planning on changing this coil or still using it as the first coil in your 5kJ multistage CG?

Edit: Any new velocity or kinetic energy measurements?

rp181

Sun Jan 04 2009, 09:20PM

Registered Member #1062 Joined: Tue Oct 16 2007, 02:01AM
Location:
Posts: 1529

Naa, it looks like wood from the grain.
I agree on the wire gauge, for your next setup with those 8 capacitors, your gonna need to go at least 16AWG(with 500j a stage)

El_Roberto

Sun Jan 04 2009, 09:30PM

Registered Member #1774 Joined: Wed Oct 22 2008, 02:51AM
Location:
Posts: 135

I think he's going to be using the whole cap bank on each stage, so he'll need thicker wire than that. Yea I think it is wood looking again, Im just trying to figure out what to make mine out of, I found some nice nylon rod for $10 a metre so I might use that.

guitardude012

Sun Jan 04 2009, 09:55PM

Registered Member #968 Joined: Fri Aug 24 2007, 04:54PM
Location: Norman, Oklahoma
Posts: 104

rp181 wrote ...

Naa, it looks like wood from the grain.
I agree on the wire gauge, for your next setup with those 8 capacitors, your gonna need to go at least 16AWG(with 500j a stage)

I have gone with 14awg (I believe the 220C enamel) at ~600J (2 of your caps) with coil params of about 0.110ohms and 0.140mH. Magnet wire gets expensive and hard to obtain after 12awg but I agree even 16awg will be fine depending on # of turns and layers.

rp181

Sun Jan 04 2009, 10:33PM

Registered Member #1062 Joined: Tue Oct 16 2007, 02:01AM
Location:
Posts: 1529

I am using 12AWG, but its not magnet wire (i didnt want to order online). Works well with 1.2kJ

first previous 6 7 8 9 next last

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