flash bulb firing current characterization?

Just have to compare the internal-resistance specs on the battery and the cap. For example, I looked into the same idea (augmenting the surge current capability) for my controller. But the battery I'm using has

Thats more what I was thinking. For GSE the limiting factor is how big a battery you can be bothered to carry around :-)

On my staging timer I had two 12v 'lighter' batteries, one for power and the other for firing, and the firing battery was connected in parallel with a

4700 uF capacitor. Overkill perhaps, but from a few quick and dirty oscilloscope measurements, peak firing current of around 7 amps - much higher than what you get from the little battery.

and that is why many new clubs build their launch systems w/o continuity check systems.

I guess the technical challenge is simply too great.

I rigged up my test circuit (which I can post on abmr as a powerpoint file if anyone is interested). It pulses a specific amount of current through the bulb for a 10 msec period, then rests for 1.27 sec.

The amount of current starts at about 35 mA and then increases by 4.4 mA each subsequent step.

So far, I've tested just 3 flashbulbs (from same box) and the range is from

426 mA to 497 mA. This is much higher than I expected, no doubt due to the short pulse time which means the total energy delivered is quite low. If one takes 0.5A at 12 Volts for 10 mSec, one has 60 mW-sec.

In the Robert Briody's experiment in 2000, the flash bulb he tested went in .3 mSec at 4 A/12V. That's 14 mW-sec. Likely the lower value is due to the speed with which the power was delivered (higher current) so less could be disappated before hitting the flashpoint.

Since this started as a question of determining what constitutes a continuity test current that is "flash bulb safe", a 10 msec test time seems far too short. I don't think that I could push a test button for only 10 msec. A 1 second test time seems much more appropriate. Note that NASA and DOD prefer ignitors that meet a 1 watt for 5 minutes no-fire test.

If you were trying to determine an all-fire current _and_ you had a time constraint on how quickly the bulb had to fire, then 10 msec would be reasonable.

If you look at the all-fire/no-fire data from

will see that the all fire pulse energy is 6mJ (the Joule being the usual way of saying watt-second). This would seem to be less than the 60 mJ you state and flash bulbs are generally considered to be more sensitive than electric matches.

Except that your 60 mJ is almost certainly far too high. If you are using a current regulated source (without seeing the schematic it is hard to be sure) then the output voltage is precisely what is required to cause the specificed current to flow and no more. In order to measure the energy, you will also have to measure the voltage at the flash bulb.

bit eimer wrote: > I rigged up my test circuit (which I can post on abmr as a powerpoint file > if anyone is interested). It pulses a specific amount of current through > the bulb for a 10 msec period, then rests for 1.27 sec. >

All true in the context of manual control over the "resistance test" timing. I'm actually using a PIC to perform the test and expect to be able to complete all measurements in substantially less than 1 msec, which is why I chose 10 msec for the pulse period.

True

Thanks for the link. I do suspect that my mesurements (in terms of joules) are significantly off as I am not using a constant current source - rather a fixed set of resistors. Of course that means that as the bridge wire heats and increases in resistance, I end up delivering less current than "expected". OTOH, my data nevertheless be valid from the standpoint of determining what an acceptable value of series resistance would be in order to constrain the delivered energy to a safe level.

Thanks for your comments. If you've interest, I've posted the circuit diagram on abmr.

So what happens when the PIC fails right after it turns on the test current? Certainly a fairly low probability event but it is something that you must think about.

I like to at least try and make things tolerant of a single point failure. But I have to deal with a club member who is still spooked by the K550 that went off a few feet away from him as the result of an ill-conceived power on self test in a wireless controller. Borrowed, not club equipment.

To get an idea of how I approach this, read the description (URL in signature) of the igniter test meter I built. I could be paranoid, but I have been around purpose built igniter testers and the builders (and users) are even more paranoid than I am. :-)

For example:

Hmmm. OK, you aren't using a constant current source but just some resistors and a constant voltage (or reasonably constant) source. If the DUT resistance is very small with respect to the resistors this works OK. But that isn't really the case here.

But it is possible to make some assumptions and come up with a better estimate of the energy going to the flash bulb than 60 mJ.

With a nominal test current of 448 mA, the selected resistance is

1/(1/39+1/78+1/156) = 22.3 ohms. If the flash bulb resistance is 1 ohm, then:

current = 12.7V/(22.3 + 1 ohm) = 0.545A

power = 0.545A^2 * 1 ohm = 0.297 W

energy = 0.297W * 0.010s = 2.97mJ

Which is likely to be low because of the heating effects on flash bulb resistance. Run the numbers for a bulb resistance of 2 ohms and you will see that the energy is 5.46mJ. The current decreases but the delivered power increases. Power delivered to the load will increase until the load resistance is the same as the source resistance. 22.3 ohms (more or less) and 1.81 W in this example.

The current path is Vcc, PFET, igniter, limiting resistor, capacitor, Vss. In the event of a PIC failure (leaving the PFET on), the delivered power will be limited by the RC time constant of the resistor/capacitor (which is about 10 msec). Of course, this is a "better case" than the straight 10 msec pulse at "fixed" current as produced by the characterization circuit I posted.

Integrity of the cap/resistor are confirmed at an earlier point in the test sequence.

I'll have a look.

Ahhhhhh! Now I understand why there was such a disconnect between my 60 mJ and the energy quoted on the Daveyfire site. My brain took a short vacation, causing me to compute the power delivered in the circuit as a whole (12 V x .5 A x 10 mS), which is obviously irrelevant. Thanks. Your numbers are correct.

replying to bit eimer, Terence Staines wrote: Anybody have any bright ideas on how to fire a flashbulb remotely, without using leads? I was toying with the idea of buying a laser engraving diode to do it, but they're pricey. I want to dangle a flashbulb from a very fine nylon thread, and either heat it or zap it without wires in some fashion, to set the bulb off (for my own amusement, to take novelty photographs).