233 replies to this topic

[MFX]

One pro firer I know, actually calculates the resistance of each cct then uses a meter to check each one before connecting them to the firing box.

Each cct is designed at so many ohms and if it isn't that when he starts he rechecks all connections. Especially useful with large series arrays!

We have a simple rule of thumb where :-

1 field cable drum = 1 igniter
X meters of bell wire = 1 1gniter. (x depending on type of wire in use)

Add them up along with the number of actual igniters and if greater than a specified number (depended on firing system in use) then you need to reduce the number of igniters or the length of bell wire. We also meter each circuit to look for unusually high readings.

[digger]

Now then.

I have been mulling over making a firing system for some time now. I have read this entire thread to see what others are considering / have done.

I recently borrowed a couple of firing units from a forum member (P.S. Ta very much). This has given me some food for thought, so after a few days of man flu and sitting in front of the laptop I have ordered a big pile of components to start some prototyping.

So here is my plan for a flexible firing unit.

The idea is to have a semi wireless / wired system using an embedded microprocessor on the master remote units to allow the use of cheap long distance 802.11 b/g secured networking to the field (the nature of the embedded system would also allow the use of a simple network cable connection in the event of interference).

It would theoretically be possible to use a large number of the remote wireless masters (no actual cues on the masters).

Each master would be able to control 8 active slaves (if that is the correct word) each with 32 - 64 cues. It would be possible to fire each cue on each slave individually or all at once. I am not sure of what is possible yet regarding the minimum delay between cues, but hopefully it will be below 20ms.

Each of these slaves would be able to control 7 sub slaves (32 - 64 cues) which would be able to fire independent cues or in parallel with the primary slave to allow 8 cues simultaneously at a time (depending on what is required by the program) . (All slaves and sub slaves are identical units, differing only by the a preselected function that they are assigned)

The system will use solid state outputs with inbuilt current limiters short circuit protection.

The system will be able diagnose problems with the output stages to prevent a faulty unit being used.

The system will be able to check the connection of all the ematches and report any connection problems (and compare it to the firing program / sequence).

Each unit will have a key switch to activate it, if any key switch on any unit is in the off state the system won't arm, the system with be able to identify the state of each unit.

So now for a few questions:-

1) do you think a hardwired kill switch is required back to the control point even though the remote masters will not be able to fire as soon as communication is lost with the remote firing station (with big red kill button).

2) Do you think 8 instantaneous cues is enough (each should be able to fire plenty of series wired matches)

3) What other features should I build into the hardware.

4) What are best places to get some nice plastic waterproof cases (IP67 IP68) for building the units into?

5) What good cost effective IP68 multipole connectors are available?

6) What else would be useful for safety features?

[Arthur Brown]

Having just worked on a big show (ca £15K budget ) that lost timing because of wireless links going down, Though I like wireless, the wired system must be better for security.

Cases will likely be Pelicases or Seahorse cases
Connectors that meet IP68 are rare, even then they only meet it mated or with covers attached

RS for buccaneer Get cable that will stand flexing without being so expensive that you cannot afford to use it!

[digger]

Having just worked on a big show (ca £15K budget ) that lost timing because of wireless links going down, Though I like wireless, the wired system must be better for security.

Cases will likely be Pelicases or Seahorse cases
Connectors that meet IP68 are rare, even then they only meet it mated or with covers attached

RS for buccaneer Get cable that will stand flexing without being so expensive that you cannot afford to use it!

The main reason for wanting wireless is to keep very long cables to a minimum, but I hear what you say and the proposed system will be able to be linked with cat5e if required and I suppose it is cheap enough to be disposable.

We have a several wireless links at work that seem pretty reliable, one of them even operates over a 5 mile link to one of our remote warehouses so it must be possible with the right equipment to get secure reliable communication.

I have had a look at the buccaneer stuff, it does look very good for the money.

Edited by digger, 10 November 2008 - 10:03 PM.

[Arthur Brown]

It's easy to get an installed radio link made to work by RF specialists then keep the system running. The hard bit is plonking a receiver down in the mud walking half a mile, and finding that something is interfering! This event had the radio mics go down AND the radio pyro controllers. Likely it was multipath transmission problems due to the steel structure of the location buildings, but the result was a load of missed cues til things were called manually. The hardwire fireworks controller worked perfectly once the operator got the go message but the wireless firing box was supposed to give him several cues that didn't happen!

Neutrik do an XLR size cat5 connector in both panel and cable and I'm told that Canford do a robust cat5 cable

Edited by Arthur Brown, 10 November 2008 - 10:20 PM.

[mike_au]

Personally I don't like the idea of using ethernet/802.11* for this sort of thing.

Ethernet was never meant for mission critical operation. If someone hits the "abort" button, just as the arp cache expires, you are wasting time (ok, not much time) doing an arp lookup before you send the message. With 802.11* you will get mobile phones searching and possibly trying to associate with your equipment not to mention the fact that 2.4GHz is getting pretty crowded these days.

Using embedded gear is better than using a laptop or something (at least you won't have to worry about Windows looking for updates, etc) but I would still shy away from ethernet.

There are a lot of communication systems that are designed for high reliability links between PLCs and the like, why don't you have a look at some of them? RS485 can run over cat5 cabling, and there are several off the shelf wireless adaptors for it. It's not like you need massive amounts of bandwidth.

[digger]

Personally I don't like the idea of using ethernet/802.11* for this sort of thing.

Ethernet was never meant for mission critical operation. If someone hits the "abort" button, just as the arp cache expires, you are wasting time (ok, not much time) doing an arp lookup before you send the message. With 802.11* you will get mobile phones searching and possibly trying to associate with your equipment not to mention the fact that 2.4GHz is getting pretty crowded these days.

Using embedded gear is better than using a laptop or something (at least you won't have to worry about Windows looking for updates, etc) but I would still shy away from ethernet.

There are a lot of communication systems that are designed for high reliability links between PLCs and the like, why don't you have a look at some of them? RS485 can run over cat5 cabling, and there are several off the shelf wireless adaptors for it. It's not like you need massive amounts of bandwidth.

I agree that Ethernet was never designed to be mission critical. Hence I was considering hardwired abort buttons. The field system would only fire a shot when it has a completed command from the remote system (and all safety systems have been satisfied). The reason for picking Ethernet is ease of implementation and system programming.

Maybe cable is the way to go, I will do some extensive testing of both methods (38 degree transmission ariel, hidden secure network for wireless testing).

Yes I have considered RS485, however maybe I will have a bit more of a look at it. We use allot of this at work Profibus, Modbus etc, but we have started using a great deal of Allen Bradley / Rockwell kit with device net. So maybe I will give our control engineers a bit of hassle and quiz them a bit to see if it can be made cost effective.

Edited by digger, 11 November 2008 - 10:55 AM.

[pyrotrev]

Although the Buccaneer connectors are admirably cost effective for a waterproof type, I've always shied away from plastic shelled things from the robustness and getting melted by stray dross aspects. I went for Tuschel circular things.

[curious aardvark]

Not sure whether to start a new topic or not - so I'll start here :-)

Maplins are currently flogging off launch kontrol systems for £15 - so I bought 2 to play with :-)

I tend to not do very large displays or that many of them - so figure this system should do the trick.

But I would like reload the igniters without having to buy them every time.
Launch kontrol is one of those gadgets where the initial kit is well cheap, it's the consumables they get you on every time.
They want £10 for every 15 fuses - to me that seems a bit steep - and it's no like I have any objection to running around with portfires - 'cos that's fun too :-)

The ignition system is simply a small section of coiled wire.
Anyone know what kind of wire ? I figure I can probably work up a way to reload the fuses if I knew that :-)

Also curious to know if anyone else has played with these things.

The idea of an infra red point and shoot control really appeals to me gadget side ;-)
The first use of this thing will be in a couple of weeks at my birthday party, then there's new year and if I can relaod the fuses myself I can think of a lot of other uses for the system as well :-)

[digger]

Well I have the basic system architecture sorted out now. All the comms seem to work, with the basic set up. I just have a little more testing to check data transmission over some very long bits of cable to optimise the baud rate for the comms. I also have to build a trial output stage, but that should be fairly straight forward.

So from a practical point of view the system will consist of master units with embedded controllers. Each master will be able to control 8 channels and each of the 8 channels will be able to control 8 separate firing modules. Any number of master units can be used together theoretically and controlled from the same firing program.

So the question is for all you experienced chappies out there with your own fancy dangle firing units how many cues shall I put on each of the firing modules? I was initially thinking of 60 per module (based on cramming as many as possible into each expensive Peli case),or do you think that 30 would be enough?

[Arthur Brown]

IMO it depends!

Firstly I've used the same cue in several places before, are your boxes going to cope with this! eg cue one ten blinkers across the front - ten cues or one cue in ten places?

[digger]

IMO it depends!

Firstly I've used the same cue in several places before, are your boxes going to cope with this! eg cue one ten blinkers across the front - ten cues or one cue in ten places?

Whatever you want really. Each cue can be fired concurrently with any other on any or the same box. Each cue can drive 7 amps so it should be possible to fire multiple matches either in series or parallel from a single cue. The maximum current from each box will be in the range of 100 - 150 amps (to be decided by the safety stage). I will be doing some speed testing on the propagation delays to see what the smallest time is that can be used between sequential cues (hopefully down into low 100's of microseconds).

The idea is to cut down on any extension of cables on ematches (and use ematches with 2m leads) so that a box can be dropped next to whatever needs to be fired to keep the wiring time to an absolute minimum. Yes this will mean that there could be a large number of unused cues on any of the boxes, but the electronics on the boxes are cheap at less than a pound a cue. The expensive bits are the cases, cable and connectors.

Edited by digger, 05 December 2008 - 01:27 PM.

[pyrotrev]

Unless you're going to be lots of single shot comet/mine stuff, my advice would be to go for 32 outputs/box maximum, probably 16 or 20 is the optimum. Costs more in connectors/hardware, but more usable in the field.

Edited by pyrotrev, 05 December 2008 - 01:30 PM.

[digger]

Ta, that answers the question. I will make up a couple of versions of the board so that I can make up a few 60 cue boxes for those occasions where this is necessary, and the rest as 30 cue boxes. I suppose it is just a matter of programming the software to recognise what type of box is attached, no major difficulty really.

[Mixologist]

It sounds quite similar to the Pyromate system i use whioch has 48 cues on it.

Have you looked at seahorse cases?? Very similar to the peli ones and substantially cheaper.