"First" Coilgun

Michael Witt, Sun Mar 05 2006, 07:16PM

I am considering building a CG, but I have a few questions before I start to go build it and spend money on it. I built a small one with a disposable camera flash unit. I know, high ESR, but that wasn't the point. The point was that I could make a coil sufficient to move a nail, even if only a quarter of an inch, with no more than $6 (USD) of parts (I know it's not a CG, but I call it one anyway). I succeeded, so now I am moving up to one that will cost me some money, but I dont want to spend a lot of money then find out that I need a bigger capacitor, transformer, etc.... So I looked at oz.net Link2 , and got a few general ideas, but had a few lingering questions, ones that I didn't find answers to, so, I looked here. These are the questions I have:
I have other questions, but I'll post them later, when I get to that stage in construction.
Re: "First" Coilgun
Sebastian, Sun Mar 05 2006, 07:37PM

Energy stored in a cap in Joules: 1/2 * C * V²

If you choose too thin magnet wire you will see that it gets too hot even after one shot. So you have to see what size caps you are gonna use and calculate from there.
Re: "First" Coilgun
Michael Witt, Sun Mar 05 2006, 08:28PM

Thanks, that helps with two questions.

I just found my schematic utility, and I'll post a draft of the circuit I was thinking about soon.

smile
Re: "First" Coilgun
the_anomaly, Sun Mar 05 2006, 08:55PM

Well I would go with the scr for this one. IGBTs' are really for performance and are usually driven with a single pulse. For the caps look for 300volts or above and a few hundred uf. Thick wire is a must to handle the high current and lower the resistance.
Re: "First" Coilgun
Michael Witt, Sun Mar 05 2006, 10:30PM

Ok, I've made a rough draft of my schematic, I know, the image quality is poor, but ym schematic prog doesn't like it when I try to print to an image.

The second attachment is a ZIP file of the image as an MDI.

Please tell me if you see any problems with this CG, I'll fix them.

Notes on the Schematic:
S1 and S2 select capacitor state; Pole 1 is default pole, and is an open circuit, used when finished charging, before discharge into coil; Pole 2 is for charging capacitor; Pole 3 is for slowly discharging capacitor, and should be used when capacitor has been charged, but will not be fired, and must be slowly discharged, CG should be stored in position 3; Position 4 is for quick discharge, in case of emergency, should not be used if possible, as it may cause damage to the capacitor.

CS1 and CS2 are for selecting which coil the charged capacitors will be discharged into, they are also how the energy is recovered, through a half wave bridge rectifier, and being charged back into capacitor.

S3 is for selecting which capacitor to monitor with the voltmeter. Set to position 3 before discharging caps. NOTE: the number of poles on the switch is one more than the number of capacitor banks used, the final pole being open, and the previous poles being conneced to one side of each of the banks.

L1-L6 are all part of a single stage, and layered on top of each other, they are not multiple stages. Yet.


1141597726 282 FT3201 Cgdraft1


Please note that I most definitely made mistakes, as I often forget polarity stuff, so feel free to correct me.
]1141597726_282_FT3201_cgdraft1.zip[/file]
Re: "First" Coilgun
Simon, Mon Mar 06 2006, 01:34AM

Whoa! That's pretty complicated for a "first" single-stage cg circuit!

As to design, I'd recommend getting the capacitors, something like 300V and whatever energy rating you feel like, and then designing the rest of the cg around them.
Re: "First" Coilgun
Bjørn, Mon Mar 06 2006, 01:49AM

Do as Simon says, simpler is better in this case. Make a simple reliable design and make an effort to measure your efficiency and try different things to improve it.

You can improve a lot on your first attempt without spending any money, a properly designed coil gun with a single photo capacitor is capable of sending a projectile accross a room at significant speed.
Re: "First" Coilgun
Michael Witt, Mon Mar 06 2006, 02:56AM

ok frown

Oh, I just remembered why I didn't like that idea angry . I dont have a switch that can handle that kind of voltage, so I have just been touching the lead of the wire to the other post of the cap, which is where I think I am loosing the energy. Is this assumption correct? Is thee a quick fix for it? I always get HUGE sparks this way.

[Edit: Fixed double post]
Re: "First" Coilgun
..., Mon Mar 06 2006, 04:43AM

First look arround and find a cap in the range of 300-500v at a like 1000uf. That would give you about 100j, which with a 1% efficiency (a resonable goal) give 1j out-40m/s for a 1g projectile if I did my math right.

From there, you need to calculate the size of the coil, for which I recommend barry's coil gun simulator. (use the calculator for transistor if you use a IGBT or SCR if you use a SCR) Aim for current in the 800a range.

As to the switch, I would recommend a SCR for your first coil. I could sell you SCR's that will handle 800a for $1/scr (sk065k's). If you want to put some more effort into the driver (you will need a pulse generator instead of a switch), then an IGBT will increase your efficiency a little.

I would recommend magnet wire around 20awg or so.

I wouldn't recommend using a transformer to charge it, but use either a voltage tripler/quadrupler or a boost converter. If you go the tripler/quadrupler path, please use a pair of 12v transformers back to back to limit the power and isolate teh cap bank from ground.

Good luck!
Re: "First" Coilgun
Bjørn, Mon Mar 06 2006, 01:44PM

A simple switching method like you describe can be very lossy. It is also not very consistent and ruins much of the science in it. As described above, a SCR is cheap and very suitable for this kind of work.

Because the coil is inductive the current will not start out very high and then ramp up to maximum current. That makes it possible to use mechanical switches quite successfully at these energy levels and voltages because they can get a reasonable contact before maximum current is reached.

If you use a SCR or a mechanical switch then you must think about how long it takes to discharge your capacitor, if not you will most likely have some problems with "suckback". This happens if there is still significant current in the coil when the projectile exits. It has happened that the projectile has been ejected in the wrong direction on particularly bad designs.
Re: "First" Coilgun
Michael Witt, Mon Mar 06 2006, 02:18PM

Ok, I'll get an SCR. But what is a "Pulse Generator"?

And how do I make/buy a voltage tripler/quadrupler?
Re: "First" Coilgun
..., Mon Mar 06 2006, 03:05PM

a pulse generator is a device that puts out a short pulse when you push the button instead of just being on. You could use a simple 555 or whatever, but it requires decent electrical knowledge, and it is just one more thing to go wrong... If you aren't looking for efficiency I would stay away from IGBT (for now). Because most of the efficiency gain from using a igbt/half bridge is that you don't fully drains the caps, so they can say they only used 82j, instead of the full 100; but the projectile speed isn't really all that much higher for a given cap.

Here are the schematics for a doubler/tripler/quadrupler. Use normal 1n4007 or whatever diodes, and caps in the 100uf range rated for 200v. I could sell you one for $20...

Good luck!
Re: "First" Coilgun
Michael Witt, Mon Mar 06 2006, 09:47PM

I wasn't planning on making a pulse generator, I just wanted to know what it was, in case I saw it in the near future.
Re: "First" Coilgun
Simon, Tue Mar 07 2006, 12:54AM

I once used a single PhFC to throw a sawn off nail (of a few grams) across a room at just over 10m/s. Experimenting with that taught me a lot about cg efficiency and was pretty fun.

As to switching, going from mechanical, touching-leads-together switching to an SCR improved efficiency by something like .5%, IIRC. That's significant when the whole cg is only a few percent efficient, and the benefit is probably much better with higher power models.

Thought you might like some actual numbers.
Re: "First" Coilgun
rupidust, Tue Mar 07 2006, 02:20AM

You present manual method of switching will still work with an SCR upgrade. But instead of the noise, heat, and light you are now manually triggering a small current in the 100mA range to the SCR gate. Further upgrading to pulse control is nothing but electronics. When the electronics know how is present, such an upgrade is routine.

Regardless of SCR or IGBT, the coil geometery and projectile position is the key foundation. You can get 4% efficiency on a first stage even with SCRs by using < 50J (low joule cg) for the first stage. Second stage at < 50 will up that even more. Low Joule CGs have good efficency levels. 10 layers, 30 turns, 8mm inner diameter 20 awg mag wire and 10 gram steel slug 30 mm long will surprise you.
Re: "First" Coilgun
Michael Witt, Tue Mar 07 2006, 10:45PM

Still a few questions.
  • Could I use the transformer from a microwave to put out the voltage necessary to charge a large cap? (is the amperage too low?)
  • Could I use multiple SCRs to pass a larger current (in parallel)?
  • and, what is a PhFC?
Re: "First" Coilgun
Self Defenestrate, Tue Mar 07 2006, 11:53PM

The microwave transformer has to be rectifyed to DC, and that would probably be too high a voltage anyways. Bear in mind they output ~2kv at ~.5A, so play carefully. Multiple SCR's are good, as well as cheap and easy if you go for TO-220 package devices. Lastly, a "PhFC" is a photo flash capacitor, like on out of a camera flash circuit out of a disposable camera.
Re: "First" Coilgun
Michael Witt, Tue Mar 07 2006, 11:58PM

Thanks.
For the transformer, I was thinking that I could use a resistor (or two) or a voltage divider to get the required power, either way, I want to be making as little as possible from scratch ('cause I have evil gremlins in my fingers).
Re: "First" Coilgun
Simon, Wed Mar 08 2006, 01:36AM

A better idea than a voltage divider would be to run the MOT (Microwave Oven Transformer smile ) from a lower voltage. Even better would to build one of the many cap charging circuits around.

The best thing would be if you can get hold of a professional flash unit cheaply. Then you have a nice capacitor with its own (very nice) charger. I've found ones with dented/cracked cases here in Australian tip shops for $5.
Re: "First" Coilgun
Michael Witt, Wed Mar 08 2006, 04:09AM

I'm a photographer in my other life, I probably have one lying around somewhere wink. My PhFC currently is one from a $4.99 walgreens disposable camera. I didn't want to spend that musvh money upfront (if it didn't so much as move the nail, I gurantee I would have given up).

I would LOVE to be able to build one of the cap charging circuits I saw on one of the older fourms, but if I don't have explicit part numbers for everything (and a PCB layout doesn't hurt), then my gremlins go to work. frown
Re: "First" Coilgun
..., Wed Mar 08 2006, 05:54AM

WilliamN gave a pretty detailed description of what he used... There is a schematic in the thread somewhere. I wouldn't worry about the mosfet driver, just use the 10ohm resistor and 12v supply for the 555.

For a project like this I would stay away from a mot. You could get a 240v->something wall transformer and plug the output into a corresponding 120v transformer to give about 340v when full wave rectified... Add a few 1n4007's and 5uf caps and you have 700w to play with. And it will be limited to a few watts, so it won't be as dangerous as a mot (but still should be considered deadly, especially when combined with a cap)


Good Luck!
Re: "First" Coilgun
Michael Witt, Wed Mar 08 2006, 01:35PM

The only reason I was thinking about a mot is because I can talk to a guy at my dad's work who can probably get me some BIG caps.
(The company he works for is G. T. Schmidt, a company that makes Nd:YAG and (recently started making) CO2 pump lasers for marking, they are the ones that make the lasers that mark most GM vin tags visible through the car window) and they usually have a few extra caps. He's already given me 5 PC-104 computers, and a dozen good stepper motor controllers; so I'm sure he'd be willing to give me a few extra caps, or a PS from one of the lasers.

[Added:]
I'm working on a PS now, I should have it posted in a little while.
1141865581 282 FT1630 Ps
Re: "First" Coilgun
Simon, Thu Mar 09 2006, 01:25AM

That circuit looks okay, except using a resistive divider for any major power control is inefficient and inelegant.

If you can design a circuit to disconnect the charger when the cap is at the right voltage (op-amp + relay, for example) then using the undivided output from the PSU will be a much better solution.
Re: "First" Coilgun
Michael Witt, Thu Mar 09 2006, 01:38AM

I guess I've been obsessed with resistive dividers recently....
I was also thinking it would be cheaper than anything else...
and I don't know any other ways sad

How would the Op amp work? As an inverter, controlling a relay, the NC side being charging, the NO side being done charging? Then how would I get the voltage down to the right voltage so that it doesn't blow up the chip? (resistive divider? amazed)

What capicatance/voltage should the cap be?

One other thing, know any way to make a light / LED indicating thet the cap is done charging? Like the way you see in a flash unit in a disposable camera?
Re: "First" Coilgun
..., Thu Mar 09 2006, 02:18AM

what is the device in the center of your schematic? A camera charger?

The easiest way to make this work would be to get a logic gate that has hysteresis on the input (a mosfet driver would e great), and use a resistive divider from the cap to the input of the ucc chip. Use the bleeder resistor for the cap(s) as one half the divider, and use a variable resistor about 1/5 of the bleeder cap for the other half. Hook the mosfet driver in the middle, and use the output either to control the charger directly, or use it to control a relay (make sure to have a diode antiparallel to the coil so that inductive spikes don't kill the driver), or a mosfet to control the charger. Adjust the put to get it to stop ate the right voltage.

What cap are you referring to?

To make a neon light that flashes when there is power put a small (like 1ufd) cap in paralel with a neon light, and hook it across the bank with a 1meg resistor. It will start to flash arround 90v, and flash faster as the voltge rises. Adjust the cap to change the brightness and the resistor to change rep rate.

good luck!
Re: "First" Coilgun
Michael Witt, Thu Mar 09 2006, 04:11AM

I'm trying to avoid mosfets, because I don't really get them. And what is hysteresis?

The cap I am referring to is the only cap in the PS schematic, it is just before the transformer (what will be used is an arc suppressor/snubber network cap from Mallory). I was just wondering if you would reccomend a capicatance and resistance.

The thing in the middle is just a transformer. The transformer would probably be just an isolation transformer, put after the rectifier.

I'm sorry the diagram is so small, I just don't know of any easy to print directly from my schematic program to an image.
Re: "First" Coilgun
..., Thu Mar 09 2006, 05:09AM

Hysteresis has to do with how the gate is triggered. I gate with it will only change state if the incoming voltage gets a certain voltage from the trigger point. So it a gate has 100mv of hysteresis and would normally trigger at 2.5v it would only change if the voltage dropped below 2.4v or go above 2.6v. It is is anywhere in between it just keeps it old state.

in order for a transformer to work it needs AC, in your schematic I only see it getting DC...

I don't have any suggestions on the cap.
Re: "First" Coilgun
Michael Witt, Thu Mar 09 2006, 11:58AM

Oh yeah, I forgot that I need to put the rectifier after the transformer. Thanks.

[Added:]
Anyone know about Jameco ValuePro Caps? Those caps work out to be something like $0.13 per joule, but I don't know anything about ESR/ESL on them. Anyoneone had experience? The P/N for the one I'm looking at is 203246CB on Jameco's site (jameco.com).
Re: "First" Coilgun
Effilcdar, Sun Apr 22 2007, 08:19AM

Trust me get a variac. (variable auto transformer)and an up step transformer. That way you can fine tune your voltage going to the rectifier. Its very easy. I even managed to get it to work.