Multi-Stage Coilgun Project Help / Electromagnetic Projectile Accelerators / Forums

Forums

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

« Previous topic | Next topic »

Multi-Stage Coilgun Project Help

1 2 next

Move Thread

LAN_403

X14U2NVX

Tue Jan 03 2012, 03:07PM

Registered Member #4306 Joined: Tue Jan 03 2012, 02:42PM
Location:
Posts: 9

Okay, I may have gotten myself too deep here. I've made some small single-stage coilguns with the camera circuits like the ones here:

So I have a basic understand of coilgun circuits, but Iâ€™m in need of some help.

I am now working on a self-optimizing multi-stage coilgun. Basically, think of a multi-stage coilgun, but instead of using calculated resistors, capacitance, voltage ect, I was going to use a computer controlled relay in order to close and open each circuit after specific set times.

In other words, Iâ€™m using a relay to control the circuit instead of blobbing out those tedious calculations for the circuit. I realize that doing those tedious calculations would make the coilgun more efficient for the parts I currently have, but Iâ€™m doing this as a sort of programming project for myself.
EDIT: I was being a little vauge, the program is going to change the times that the curcuit is opened and closes and between firings and will optimize final momentum through recorded data taken by a force sensor. That isn't the problem, I just need help with the coilgun.

The end result (hopefully) would be a 2-3 stage coilgun that after a set number of trails, would optimize circuit timings for the first coil, then would optimize for the second coil, for the third, ect, so at the end you would have an efficient multistage coilgun.

My main concern is, having seen no other project like this on the web yet, could I even accomplish this with lower-end (cheap) USB programmable relays such as the one here?:

My main worry is that the relay would not be able to supply a good enough current to launch a projectile, because the one Iâ€™ve found is only rated for 15 Amps. My only saving grace is that I have to demonstrate how a program could interface with the relay to optimize the coilgun, it doesnâ€™t have to be a glorious coilgun of 100 Joules.

m4ge123

Tue Jan 03 2012, 05:13PM

Registered Member #4118 Joined: Mon Oct 03 2011, 04:50PM
Location: MD
Posts: 140

For varying the amount of energy going into the coils, you definitely can't use relays. They respond way too slowly and the contacts will weld themselves together after a few shots. There are two ways for you to control the amount of energy going to the coils: use MOSFETs or IGBTs and turn them on for a variable amount of time, or use SCRs and vary the cap voltage. It would probably be easier for you (and leave you more options) if you used a microcontoller instead of trying to inferface with a computer. But if you do use a USB interface, I would suggest you adjust the cap voltage because you probably won't be able to send precisely timed pulses through the interface.

X14U2NVX

Tue Jan 03 2012, 08:59PM

Registered Member #4306 Joined: Tue Jan 03 2012, 02:42PM
Location:
Posts: 9

Thank you for the feedback, I'd figured relays wouldn't work, but eh, it was worth a shot.

Now, you say use a microcontroller to interface with either a MOSFET or a IGBT.

As far as the MOSFET or IGBT goes, I'm researching that now (haven't worked with them enough to understand them thoroughly)...

But for the microcontroller, you mean something like an Arduino or other programmable board so to speed up the transmissions?
EDIT:
After Researching MOSFETs and IGBTs, I've felt like the MOSFET would do the better job for my project (IGBT have bigger time variances, which is a confounding variable in my project). But what amperage should I get them to? Assuming I could use my Arduino at this point, i've picked out:

Would these be okay? (Once again, sorry for the "beginner" questions, but I'm in a whole new zone with these MOSFETs)

Turkey9

Tue Jan 03 2012, 11:15PM

Registered Member #1451 Joined: Wed Apr 23 2008, 03:48AM
Location: Boulder, Co
Posts: 661

To know what current you need, we need to know the size of your capacitor and the inductance of your coils.

Are you using any type of feedback in this scheme? For example, an optical trigger to start the pulse and then using the micro to vary the length of the pulse? If you're trying to control all of the coils just using a fixed timing, it won't work. The problem is that because this is the real world, each shot with the first coil won't be the same. That error from shot to shot can't be predicted and will just be amplified in the next stage. Before long the coil will either fire too soon or too late, possibly shooting the projectile backwards down the barrel.

X14U2NVX

Wed Jan 04 2012, 12:08AM

Registered Member #4306 Joined: Tue Jan 03 2012, 02:42PM
Location:
Posts: 9

Okay well I'm using multiple 68000 microferad capacitors (I hope to use one for each coil, if possible). As far as the variability goes, I foresaw this in the project design and am taking this into account (there will be many trials for each setting). Part of my project includes accounting for this variance by taking large samples, and I plan to use statistics in the controller program itself. As far as the coils go, I'll make them to any specifications needed. At the moment, I was thinking about using 3-4 wraps around a plastic tube large enough to fire a small cylindrical iron(steel) dowell one inch long (both the projectile and the coil are going to be one inch). But these plans aren't set in stone and I'm willing to test multiple possibilities after I get the circuitry working.

Turkey9

Wed Jan 04 2012, 02:03AM

Registered Member #1451 Joined: Wed Apr 23 2008, 03:48AM
Location: Boulder, Co
Posts: 661

Using caps that big (i'm assuming 400V?) you'll be hard pressed to find a budget friendly MOSFET, that one you linked will certainly not work. IGBTs will also get pretty expensive.

Even with a large pool of samples, you won't be able to predict the error for the next shot. When your program runs, you'll get the timing correct every once and a while but most of the time the shots won't work. I'd bet that the better the construction of the coilgun, the higher the percentage of successful shots, however.

X14U2NVX

Wed Jan 04 2012, 03:00AM

Registered Member #4306 Joined: Tue Jan 03 2012, 02:42PM
Location:
Posts: 9

They're 100 volts, but I understand your point. I'm going to think this over, I'll have a solution/modification tomorrow. Thanks for all your assistance.

X14U2NVX

Wed Jan 04 2012, 01:40PM

Registered Member #4306 Joined: Tue Jan 03 2012, 02:42PM
Location:
Posts: 9

Okay, so after much discussion, I've come to a few decisions:

1. I need to fully explain myself, this project is itself a research project, the goal is, in fact, a project to prove the usefulness of software optimization in a multi-stage coilgun. So after optimization, itâ€™s still okay if it â€œmisfiresâ€ as long as it fires moderately consistently.
2. Realizing this and the low budget of this project (under 200 dollars), I realize that my current capacitors are much too great, Iâ€™ve asked around and now have access to smaller capacitors, some at 6800 microfarads and 25 volts, and some at 34000 microfarads and 25 volts. Obviously I will probably be using the smaller.
3. Because I have access to whatever else I need (both locally and without cost) in terms of wire, piping, projectiles, ect, I will be building my project solely around my choice of MOSFET.
4. However, as I have stated multiple times, I am quite unfamiliar with MOSFETs or their capabilities and limitations. Iâ€™m currently looking for MOSFETs but any suggestions would be appreciated. (Because this is a three-stage coilgun, I would have to be able to afford three of them, right?)

EDIT: Exactly how many Amps am I aiming for in a coilgun? I realize that the Magnetic field due to a solonoid is directly proportional to the current, so I want to maximize current, the only problem is I can't find high current tolerant MOSFETs for very cheap. I'm really trying to find Arduino controlled MOSFET, as I already have the Arduino, know how to program it, and don't have to pay for it. The most i've been able to find is 33 Amps.

m4ge123

Wed Jan 04 2012, 02:39PM

Registered Member #4118 Joined: Mon Oct 03 2011, 04:50PM
Location: MD
Posts: 140

Those 6.8mF 25V caps are really too small for a coilgun. I'd recommend 1-5 34mF 25V caps in series per coil.
I looked on mouser for FETs for a few minutes and found this

You should calculate the inductance of your coils and run a simulation to see if your FETs can handle the current pulse and to figure out the pulse length and stuff. If you don't have simulation software, LTSpice is good.

X14U2NVX

Wed Jan 04 2012, 02:53PM

Registered Member #4306 Joined: Tue Jan 03 2012, 02:42PM
Location:
Posts: 9

When capacitors are connected in series, their equivalent capacitance lowers (is the reciprocal of the sum of the reciprocals). Do you mean in parallel?

As far as inductance goes, I'm literally going to determine the coil type and windings, ect, based on the limit of the MOSFETs. As far as the one you linked me to, itâ€™s not only non-arduino compatible, but I can't find a compatible microcontroller at all.

EDIT: I've found a microcontroller and several MOSFETs compatible with the board. While they're only rated to 30 amps, I'm willing to compromise at this point. I'll probably be willing to push the circuit to 35 Amps based on how each is rated on continuous current and not bursting, but is this enough current to even make a coilgun?

1 2 next

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