Project: Remote Firework Ignition Station

[NeRd]

This is a small project that I decided to start on right after I finished my 'Mithril P4' case mod.

The RFIS is something that allows me to kill time, play around with some design ideas I had and most of all have an insane time with fire crackers!

How RFIS works...

The simplest way to explain RFIS would be:

Simply put, I plan to step-up 12vdc to around 30kV @ around 1 - 2A connected to a spark plug. My initial design was to have 12vdc applied continuously to the coil and then kill the applied voltage resulting in a single 'spark'. I have now decided to pulse the coil with 12vdc @ 2kHz resulting in a continuous spark!

The coil I plan to use will look something like this:

 

[NeRd]

Custom PCBs

For this project I decided to design each sub-system as a seperate PCB, allowing me to use them in other projects and by selling off the excess it becomes affordable. This also allows me to add and remove sub-systems as I see fit.

Voltage Regulator

The purpose of this circuit is to regulate the source voltage down to a steady 12vdc. As my source I will be using 2x 9v batteries connected in series. To calculate the resistors required I performed the following:

[img]http://www.bsodmike.com/stuff/RFIS/vreg.gif[/img]

Vo= Vref(1 + (R2/R1))

Simple algebra gives us:
R2 = R1 ((Vo/Vref) - 1)

Vref = 1.25v (voltage developed if R_adj = 0ohm),
R1 is set to 240ohms (as per datasheet)

R2 is calculated to be 2064ohm, taking 2Kphm for the resistor!

Astable 12vcd Pulser

This comprises of an NE555 timer chip configured to pulse the signal @ 2Khz by the following:

f = 1.44/(R1 + 2R2)C  --> R2 = (1.44 - fCR1)/2fC

Taking f = 2000Hz, R1 = 500ohm and C = 1x10^-6f (1uf) >

R2 = 110ohm, but I used 100ohm resulting in the pulse rate being upped to 2.1kHz

Custom Voltmeter

This has an LM3914 configured to take a reading between 10.5 and 15V in 0.5v steps.

<This will be updated later>

 

[NeRd]

Getting the show on the road...

For once I forgot to take an image of the box before I took my drill to it, but this will have to suffice:

You can see that I have drilled and mounted a couple switches here. This is still a work in progress so I will post back with a finalized design of the front and sides!

This is far from over, so keep ya eyes peeled for updates, expect a big one sometime next week

 

[mattg2k4]

30kv @ 1-2A is 30-60kW. Divide that down by the 12v input and you'll need 2500-5000A from the batteries. I don't believe that standard 9v batteries can provide that much power So hopefully you meant 1-2A input, which would result in .4 - .8 mA output.

Also, why are you regulating the 18v input back down to 12? You're down to 66% efficiency right from the start. Would 18v be too much for the coil? I personally would recomend NiCd or Nimh battery packs, which can easily be put in series for 12v, with significantly higher current draws available. You could also use a small 12v sealed lead acid battery.

 

[NeRd]

Your right, I should have been more clear here. The 2x 9V batteries are to drive the electronics, but the main power for the 'Battery - FET - diode - Coil' line I will be diverting power to an external 12v lead acid battery via some funky XLR jacks... ohh yes, 1 - 2A input. As for regulating the 18v down to 12v, I can't get 12v battery packs here easily, and as 9v batteries are readily available I thought I'll use 'em.

 

[NeRd]

Spent most of the evening today solding up the PCBs and testing a PCB from a different project (an led throbber) and well it works..yay

You can admire my mess here:

Close up of the 3 main PCBs:

More to come!

 

[Whatsisname]

I did something similar for igninting potato cannons, only i used a flyback coil jacked from an old computer moniter.

 

[d03boy]

Why not just use flint?

I suppose I could guess why... electronics are funnn yay

 

[Aristarchus]

Personal experience suggests that you're going to have some significant problems lighting fuses with a sparkplug. IF you can manage to send a high voltage charge directly thru the flash powder, you MIGHT get the things to go off, but this is going to be a hit or miss proposition as well. Your best bet would be to use some sort of resistance wire ignitor with a pyrophoric coating (such as model rocket ignitors), but this would be an expensive and/or time consuming affair...

But have fun anyway, and don't kill/injure/break anyone/thing in the process....

 

[NeRd]

Thanks for the replies! My Bulgin has been delayed for another week, but I got the final XLR jacks I need and some other stuff. The Voltmeter PCB should arrive next week as well. I'm sure a 'continuous spark' should catch the fuse alright eh? That's why I'm using the DC Pulser!

Please comment on this, should I run the 12vdc pulse @ 1Khz or 2Khz...or should I put a rheo to control this? I don't want to clutter the front tho! I could just end up wiring in a pot.

 

[mattg2k4]

Different coils work best at different frequencies. A pot would be the best way to go.

 

[Aristarchus]

Originally posted by NeRd
I'm sure a 'continuous spark' should catch the fuse alright eh?

Don't be too sure about that. Most fuses I've come across require quite a bit of heat to catch, particularly the heavy green, coated type. THe cheap tissue-paper fuses might work, but I'm guessing that you'll end up burnign a tiny hole trhough the fuse without ever lighting it.

...I could be wrong though. let us know when you try it.

 

[NeRd]

Spent today soldering up the PCBs and getting everything hooked up. The relay swapping 'battery' input works perfectly. I did a 'mock' test run and not having a signal gen or croc clips is a really pain! Seems that out the FET comes 9.7V with a 12.7V out from the Vreg (down from 20V supply...I found a -10v line on my semi-dead ATX *test* PSU, and as the 12v line was doing 10v voila, 20V!)

so a 2K7 or 2K8 should get me closer to 15-16v through the Vreg which is the safe max for the NE55 and LM3914, and this should solve that problem. I used my fluke DMM and on AC -hz mode it was showing something close to 1Khz signal. I think it feels the signal is 1Khz as it is pulsed DC signal, so it is really a 2Khz signal. Of course I'll need a scope to check this out. Yea I'm lacking quite a few essentials...

Pot for freq control: Good idea really, if I feel like I must do this I will later on. I'll first test it out how it works the way it is, else I'll lop off the 500ohm resistor on the DCPulse and wire in a Pot...'tis a tight fit inside the box now....

So far so good. I've started work on a small job that has come my way. Here is the finished PCB layout of my latest work:

Will probably give that more priority in the next couple days...

More images to come as this project is starting to take shape!

 

[fat-tony]

Originally posted by Whatsisname
I did something similar for igninting potato cannons, only i used a flyback coil jacked from an old computer moniter.

[epiphany]

Wow! I've been looking for a cheap easy to find coil for my cannon for a while. YOU ARE AMAZING

[/epiphany]

 

[mattg2k4]

If you're interested in launching spuds you might check this design using a salvaged camera flash system.

Or read the same guy's tutorial on flybacks.

 

[Whatsisname]

You can also be 10-1,000+ feet away if you need to be.

Originally posted by d03boy
Why not just use flint?

I suppose I could guess why... electronics are funnn yay

 

[Whatsisname]

It will be difficult and unnecessary to vary the frequency. Also, "continuous spark" is going to be virtually impossible. Virtual spark usually means a 1/10 of a second or so. Anything more will blow up your circuitry/ignition coil. I only gave it a quick glance but I have a feeling that using the IC to control the frequency is going to result in a burnt IC.

Its quite possible you dont need a spark at all for when you are lighting fuses. If you had some sort of insulated clamp on the fuse, and discharged a good amount of juice through that, that alone might be sufficiant to light the fuse.....you'd have to have the clamp release of course..but it's just an idea.

Originally posted by NeRd
Please comment on this, should I run the 12vdc pulse @ 1Khz or 2Khz...or should I put a rheo to control this? I don't want to clutter the front tho! I could just end up wiring in a pot.

 

[Whatsisname]

Those things are dangerous. I feared that device more than the potato cannon itself. Be extremely cautious in its operation. Put a relay into your driver circuit to only allow operation for a 1/10 of a second MAX! Anything more = death!

Originally posted by fat-tony
[epiphany]

Wow! I've been looking for a cheap easy to find coil for my cannon for a while. YOU ARE AMAZING

[/epiphany]

 

[Whatsisname]

i have a bad feeling that thing will get blown up during its first usage.

Originally posted by NeRd
Spent today soldering up the PCBs and getting everything hooked up. The relay swapping 'battery' input works perfectly. I did a 'mock' test run and not having a signal gen or croc clips is a really pain! Seems that out the FET comes 9.7V with a 12.7V out from the Vreg (down from 20V supply...I found a -10v line on my semi-dead ATX *test* PSU, and as the 12v line was doing 10v voila, 20V!)

so a 2K7 or 2K8 should get me closer to 15-16v through the Vreg which is the safe max for the NE55 and LM3914, and this should solve that problem. I used my fluke DMM and on AC -hz mode it was showing something close to 1Khz signal. I think it feels the signal is 1Khz as it is pulsed DC signal, so it is really a 2Khz signal. Of course I'll need a scope to check this out. Yea I'm lacking quite a few essentials...

Pot for freq control: Good idea really, if I feel like I must do this I will later on. I'll first test it out how it works the way it is, else I'll lop off the 500ohm resistor on the DCPulse and wire in a Pot...'tis a tight fit inside the box now....

So far so good. I've started work on a small job that has come my way. Here is the finished PCB layout of my latest work:

[ing]http://www.bsodmike.com/stuff/PumpController/PCv3.0_brd.gif[/img]

Will probably give that more priority in the next couple days...

More images to come as this project is starting to take shape!

 

[FLECOM]

why not just use model rocket starters?

they are like a little tiny wire inside of that stuff they put on the end of a matchstick... you can fire with straight up 12v with a few amps for a few seconds...

 

[Whatsisname]

because you constantly would have to replace those. You can jump spark gaps virtually indefinatly.

Originally posted by FLECOM
why not just use model rocket starters?

they are like a little tiny wire inside of that stuff they put on the end of a matchstick... you can fire with straight up 12v with a few amps for a few seconds...

 

[FLECOM]

Originally posted by Whatsisname
because you constantly would have to replace those. You can jump spark gaps virtually indefinatly.

right, but can you get it to work

and you would have to be replacing the fireworks constantly anyway so im not sure why this would be such an issue... i believe that actuall fireworks shows use something similar to what i suggested

 

[Whatsisname]

I've gotten a flyback powered potato cannon ignition working. You can switch where the high voltage is directed from afar so you can setup a large number of fireworks and fire them all individually.

Originally posted by FLECOM
right, but can you get it to work

and you would have to be replacing the fireworks constantly anyway so im not sure why this would be such an issue... i believe that actuall fireworks shows use something similar to what i suggested

 

[Party2go9820]

Originally posted by FLECOM
right, but can you get it to work

and you would have to be replacing the fireworks constantly anyway so im not sure why this would be such an issue... i believe that actuall fireworks shows use something similar to what i suggested

You're exactly right. Well, all the smart pyrotechnicians use them. Its commonly refered to as an ignitor or "squib". The squib is packaged into the fuse (usually just paper coated in black powder) so that when it fires, the fuse lights which starts the rest of the process - lift charge -> dispersment charge -> color/display burn -> ohhs, and ahhs.

It doesn't take much juice to "ignite" the squib, but its better to have too much than not enough. We usually use custum batteries - I think its just a bunch of D cell's bundled together - but I know for fact you can do it from a recharable drill battery in a pinch </voice of experience>

The beauty of this system is you can wire the shells/squibs together and really fill the sky..

 

[TEK NINJA]

I have done an ignition system before, based on a simpler (but not re-usable) design.

Basically, I would wrap tiny copper wires in sulphur, which was patially submerged in fuel. I would then apply a 12 volt current, which would cause the wires to heat, the sulphur to ignite and burn quickly, and light up the gas. It worked quite well, ignition is guaranteed if mounted properly.

 

[Whatsisname]

From the sounds of though, such a system is impossible to shut down if something goes wrong.

Originally posted by TEK NINJA
I have done an ignition system before, based on a simpler (but not re-usable) design.

Basically, I would wrap tiny copper wires in sulphur, which was patially submerged in fuel. I would then apply a 12 volt current, which would cause the wires to heat, the sulphur to ignite and burn quickly, and light up the gas. It worked quite well, ignition is guaranteed if mounted properly.

 

[TEK NINJA]

Originally posted by Whatsisname
From the sounds of though, such a system is impossible to shut down if something goes wrong.

Once its on fire, its on fire... How would it be any easier to "shut down" gas you just lighted on with a spark plug?

 

[Whatsisname]

With fireworks you wouldnt light any gas. You'd position the fuse between a spark gap and when the spark gets fired it will light the fuse on its own.

Originally posted by TEK NINJA
Once its on fire, its on fire... How would it be any easier to "shut down" gas you just lighted on with a spark plug?

 

[TEK NINJA]

Originally posted by Whatsisname
With fireworks you wouldnt light any gas. You'd position the fuse between a spark gap and when the spark gets fired it will light the fuse on its own.

Well if just a small spark can light it on, it could work with the ignition system I proposed, without the gas.

 

[GlobalFear]

Shove a m-atx mbo and a wireless card in it, write a simple app and control it from your laptop. ammobox pc that lights off fireworks!