Pyrotechnic Firing Controller / Projects / Forums

Forums

4hv.org :: Forums :: Projects

« Previous topic | Next topic »

Pyrotechnic Firing Controller

Move Thread

LAN_403

Hon1nbo

Mon Sept 27 2010, 04:32PM

Registered Member #902 Joined: Sun Jul 15 2007, 08:17PM
Location: North Texas
Posts: 1040

UPDATE: Oct. 1, 2013

This project had gotten scrapped during my freshman year of college due to time, and looking back on it I had some bad design choices and could make it a lot better, so I've redesigned the unit and, thanks to a shiny new laser cutter I have access to, have started work.

The old version is left at the bottom for posterity, though I removed the well oversized images.

The controller is a mixed mode unit, which will support acting as an independent standalone firing controller, a slave in a network of units, or a preprogrammed display system.

There are a few things I've noticed about firing systems over the years, one is that they tend to have very long and high pin count cables that are more expensive, and difficult to maintain. So I want to build a system that uses something a little easier to work with, such as Cat-5e.
Secondly, there are some all in one firing controller, but they either aren't as flexible to work with or they are very expensive. I thought I might be able to build a cheaper solution for an easy to deploy, and scalable, firing system.
Lastly, scale. Firing systems tend to be limited in what the master node can control without complicated multiplexer boxes, but with standard networking technology each node can simply be addressed by the IP address. As well, networking systems are easy to make redundant in case of failure (or let's say a ground det on a network switch/multiplexer box).

The Major components:

BeagelBone Black edition: runs the OS and networking, has plenty of GPIO for controlling cues and reading continuity checks without having to use additional multiplexers or shift registers. Still debating whether to run the normal Angstrom Linux, or switch to a more common OS for real time operation such as Mona Lisa Linux (as I've found medical devices to be running at times).

Pelican 1200 case w/ panel mounting kit: The original planned 1120 case ended up not being big enough for practical use, as it would have needed external power and had little roon for controls and hardware, so I bumped it up a little bit. With the panel mount kit, it will have a control panel for manual operation (although only 3 cues with the limited number of switches) and access to the HDMI and USB ports of the Beagle Bone. The controls are able to be set and the lid closed when leaving the controller as a slave unit.

External Connectors: going with a suggestion, I am going to install PyroClips instead of the original common speaker terminals. As well there will be a waterproof RJ-45 connector for network connectivity, as well as a waterproof DC power input for external power sources.

Command and Control: The units, when operating as a network slave, will be driven over SSH. This will allow a secure control when using a wireless system such as WiFi, as well it can provide a simple interface for control on the client side GUI. Being over a LAN system, it should be very speedy for control as the beaglebone can be run without a graphical interface, or any other libraries that might eat up resources.
The use of a lightweight protocol like SSH will also allow for easier to design custom firing control panels, also based on boards such as the BeagleBone or lightweight CPUs, to control the rest of the nodes.

Current Progress:

Oct. 1, 2013: I've laser cut the manual controls onto the panel and mounted the basics, as well I laser cut the pelican case a little bit to make it more, let's say, easy to identify.

I am also currently also waiting on parts for the BeagleBone and external interfaces..

Oct. 7, 2013: Pictures (finally!) Up next is going to be mounting the Pyro Clips once i get them, followed by making some transistor circuits to drive them, then installing the BeagleBone, and final wiring. Also going to toss the unit into my bath tub underwater and make sure the seals are good before I load up and sensitive electronics.

Pelican 1200 case with the panel mount frame and a hole cut for the waterproof RJ-45 connection

1381191724 902 FT1630 20131007 182040 Copy

Test fit of the buttons on some card stock cut before the main ABS panel (also makes the lettering a little easier to read until I fill the grooves with a contrasting paint)

1381191829 902 FT1630 20131007 182046 Copy

Outside of the pelican case

1381191829 902 FT1630 20131007 182054 Copy

Parts added to the ABS panel

1381191829 902 FT1630 20131007 183206 Copy

The key switch, as required by NFPA 1123

1381191829 902 FT1630 20131007 183256 Copy

The RJ-45 connection, which is a simple pass through with a cable gland based seal

1381191829 902 FT1630 20131007 183523 Copy

1381191829 902 FT1630 20131007 183540 Copy

The panel mounted in the case

1381191829 902 FT1630 20131007 184356 Copy

An overview of the unit as it currently stands.

1381191829 902 FT1630 20131007 184928 Copy

Begin Old Post---------------------------------------------- -------------------------------------------------- --
This time around I am building a custom firing controller for pyrotechnics. This will primarily be used for C grade (consumer) fireworks until I get my license, though I will be bringing it out to a 1.3g shoot and might test it out at that level though I doubt there will be time, so I'm not betting on it.

I am building this in some spare time when I'm not in class or out with friends. Most of the parts I already owned (and the pelican 1120 case is recycled: it used to be my hard drive case but now none of my drives save one fits as I switched away from HDD toasters to enclosures). The one thing that was bought new besides small parts like terminals is the Panel insert for the pelican case.

The Project Code can be found on my temporary website, it currently is 500+ lines of code and still going!:

This project is to make an Automatic Controller for pyrotechnic displays. To start there will be only one "node" with 8 cues. In the end, there will be an infinitely expandable number of nodes each having up to over 40 cues a piece. The nodes communicate between each other and the primary control unit (a computer or a master node) over an Ethernet Connection. The basis of each node is an Arduino and a Pelican 1120 Case. The 1120 will likely need to be swapped for a 1200 when the max number of Cues is reached, unless I decide to make the max number of cues only accessible by a module attached via an umbilical.

The first prototype will consist of one node with 8 cues and an LCD with keypad interface so that this node will end up being a "master" node when more are built. It will accept the timing and sequencing information and fire shots accordingly. It will start out acting independently. Once I install a multiplexer I will add in the Ethernet Functionality as without it I only will have four firing cues which is unacceptable for even testing.

When I have time/resources, I will construct slave nodes and install multiplexers on the Arduinos. I am going to use the MUX shield from SparkFun, however: the shield has to be modified such that there are no pin conflicts with the Ethernet or LCD shields.

The process is simple for the first prototype:
1) User Settings Menus:
A) user inserts the total number of cues
B) User inputs the information about each shell (for more finely tuned firings)
i) time until shell fires after last fire/start of program
ii) delay in igniting fuse (for C grade pyro, which this will be tested with, using NiChrome to light the Visco Fuse)
iii) time from fuse ignition until shell launch (again for C grade pyro with which this will be tested - B grade pyro fuse burns near instantly)
iv) Time from launch of shell until shell burst (used for advanced timing of shows)
C) Allow user to test the firing sequence before arming the shells, sequencing is displayed on LCD or LEDs
D) firing system primed but programming safety engaged, user must turn it off for the system to activate
E) user arms system and it begins the sequence, also it will start music at this point if connected to an MP3 player

The Enclosure is a Pelican 1120 Case with a few add ons: first, the Panel Insert was installed which adds a panel with a dedicated O-Ring so even if the case is open the contents underneath the panel are still waterproof. However, the prototype uses an unmodified LCD shield for the Arduino so it will not retain the waterproof features, but the LCD can later be placed under a membrane. Also, there are terminal ports on the sides for connections to the pyro product. These are waterproofed such that the water cannot enter the case but they themselves might still be at risk of shorting. However, with a good connection the majority of the current should still rather take the route through the igniters than through any small amounts of water. There also are manual buttons for the initial prototype should the automation not be needed or fail. They will not be there on the 40+ cue versions due to physical constraints in size, but I may move up to a Pelican 1500 which would allow enough manual buttons and terminals easily. The case is designed to be closed or open for firing, closed being the preferred method such that it is protected from debris and low bursting shells.

The connection terminals are standard Speaker red/black terminals. There are four on each side, and when the multiplexer is installed there will be an Expansion port due to lack of space for 40+ terminals. On the front side there will be a USB port for programming. There also is a jack for 12V power for charging the internal battery or for using auxiliary power for the system.

Inside the Pelican Case lies an Arduino (ATmega 168), LCD Shield, two 12V batteries (one for the Arduino and the other for the actual firing), a Transistor Array for the firing terminals, and pull-down resistors to help ensure that no current induced from one shell firing will set off another shell.

-Jimmy

---------------------------------- -------------------------------------------------- -------------------------------------------------- -------

9/27/2010:

I have written a good portion of the code in one sitting, and I marked out the slots on the pelican panel insert but I had trouble with my dremel cutting it, might be the really cheap cutting wheels I used. will post pics once the initial cuts and fitting of parts are made.

------------------------------------------- -------------------------------------------------- -----------------------------------------------

10/4/2010

Here are some photos of the parts and expected layout before I started cutting the case (which I still have to finish cutting, I will either have to use the shop or go home as the dremel motor may also just not like my cheap inverter in my car)

An overview shot of the initial parts for the first prototype: the Pelican 1120 case with panel frame insert, a set of speaker terminals, an arduino with LCD and Ethernet shields (ethernet will be installed at later time), a Key Switch to arm the system electronics (software is armed by LCD keypad, key actually allows power from battery to igniters bypassing continuity checks), and a power jack for charging or for power off of external 12V source.
old_1

close up of misc parts.
old_2

approximate layout of the panel components
old_3

one set of terminals in approximate positions, but I may go with a 2x2 on each side instead of a 1x2 set (1x2 is shown)
old_4

quicksilver

Fri Oct 01 2010, 10:47PM

Registered Member #1408 Joined: Fri Mar 21 2008, 03:49PM
Location: Oracle, AZ
Posts: 679

Do you have any pics of the case or setup thus far? I've thought about something somewhat similar...(I had really loved rockets but drifted out of it).

With computer PSU moving higher than 1000watts and inverter designs being as small as they are now any place an automobile/truck could go a decent system could be set up.

Hon1nbo

Mon Oct 04 2010, 07:20PM

Registered Member #902 Joined: Sun Jul 15 2007, 08:17PM
Location: North Texas
Posts: 1040

thread updated with pics, still needs a lot of work. Sorry for the wait since I posted this, but I don't have as much time on my hands with college, and no, not all of it is studies

-Jimmy

Martin King

Mon Oct 25 2010, 10:40AM

Registered Member #3040 Joined: Tue Jul 27 2010, 03:15PM
Location: South of London. UK
Posts: 237

Looks nice so far I'm a big fan of Peli cases myself, just a couple of minor comments, have you considered Pyroclips? FAR superior to speaker terminals IMHO. Also I note your keyswitch has the key removable in both positions. My arm keys can only be removed in the safe position (to stop someone walking off the the key leaving the box armed) and if possible I try and arrange for the key switch to be in a position such that the key has to be removed before you can close the lid of the box.

Martin.

Hon1nbo

Tue Oct 01 2013, 07:08PM

Registered Member #902 Joined: Sun Jul 15 2007, 08:17PM
Location: North Texas
Posts: 1040

Updated last at the dawn of time, so I figured it was time to brush off the dust.
Pics of current progress will be uploaded tonight when I get home, as I was making this thread I realized I had the wrong pictures.

-Jimmy

Hon1nbo

Tue Oct 08 2013, 12:32AM

Registered Member #902 Joined: Sun Jul 15 2007, 08:17PM
Location: North Texas
Posts: 1040

Updated with images (finally!)

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